» America 2050’s report produces a long-term, phased rail system proposal.
Over the past six months, there’s been much talk of high-speed rail development in the United States, most of it revolving around how the federal government will choose to award the $8 billion included in the stimulus for the program. The discussion has been frustrating, because it has skirted around the incredibly important issue of which corridors are right for fast trains with under-researched over-the-top plans or out-of-date, politically-motivated proposals; meanwhile, the term high-speed rail is thrown around with gumption whenever any train improvement project, no matter how fast, is mentioned.
As far as I know, perhaps with the exception of the high-speed rail investment scheme I produced in February, there has been no serious attempt to evaluate the nation’s rail corridors using an objective, repeatable standard. Fortunately, America 2050 has done just that in a new study released today. It is a quality product that outlines a future American rail network whose shape and proposed phased implementation are the result of a considered study of the country’s cities and metropolitan areas. It is a report whose methods and conclusions should be examined by the Department of Transportation before it releases its National Rail Plan and begins awarding stimulus funds later this fall.
The report’s authors, Yoav Hagler and Petra Todorovich, evaluated more than 27,000 corridors — that is, every pair of cities with respective populations of more than 50,000 and distanced between 100 and 500 miles apart. The rating system, based on six criteria — city size, distance between cities, transit connections, per capita GDP, traffic congestion, and location in a megaregion — isn’t perfect (I’ll return to it later in this post), but at least it attempts to use relevant information to prioritize corridors. This is far more than can be said for virtually every other organization that has discussed high-speed rail development.
The highlights of the study are summarized in the map at the top of this page. After considering the 27,000 corridors, America 2050 notes that the corridors with the highest ridership potential would be in the Northeast (from Washington to Boston, with a spur through Springfield), in the Midwest (radiating from Chicago to Minneapolis, St. Louis, and Detroit), and in California (from Los Angeles to San Francisco and Sacramento). The report advocates their construction first, and I largely agree with the choice of corridors: they’re the most heavily and densely populated areas in the U.S.
In general, the second phase is equally reasonable — proposed are lines between Vancouver and Portland; Los Angeles and San Diego, Phoenix, and Las Vegas; Dallas and Houston; Toledo and Cleveland; Washington and Atlanta; Tampa and Miami; and New York and Albany. All of these lines are currently under consideration by state authorities, and they’re worthy of investment after those suggested in the first phase. Most of the third phase lines deserve to be the last under consideration, though I would suggest that Ohio’s big cities and Pittsburgh get the short shrift here.
The overall reasonableness of route selection presented, however, is what’s important: namely, this study presents an objective manner by which to compare different corridors and it reaches a number of conclusions. We should expect the federal government to present something similar in the National Rail Plan next month.
That said, there are some fundamental problems with the equation used by the report to evaluate city pairs. Most troublesome, it considers 250 miles an “ideal” distance — rating such corridors 2.5 times more valuable than 100 mile routes — and does not consider at all any city pairs less than 100 miles apart. Today, the Philadelphia-New York corridor is the top-ridership route on Amtrak’s Regional and second-highest ridership route on Acela — and yet, because the cities are just under 100 miles apart, travel between them is not considered in the study. This doesn’t diminish the top ranking of the Washington-Boston corridor, but it puts the study’s fundamentals into question.
Was the Florida line delayed to the proposed second phase because travel between Tampa and Orlando wasn’t considered? Was the route between Hampton Roads and Richmond — a potentially high-ridership route — eliminated because travel between those cities was ruled out? Do the study’s authors really think that there would be more traffic between Hartford and Philadelphia than between Hartford and New York, because the latter route is shorter?
Meanwhile, the insistence on rating local transit connections in the equation seems, as I’ve expressed before, superfluous, and it’s unclear to me how congestion on local highways — another criterion — has anything much to do with intercity rail travel.
Nonetheless, these objections are somewhat besides the point, because America 2050’s report isn’t the final U.S. high-speed rail route network. The assessment tools by which the report methodically appraises potential rail corridors are exactly the kind of system the federal government should be using when deciding how to allocate grants. It remains to be seen whether the DOT will make such a commitment to a similar level of objective evaluation.
Image above: America 2050 high-speed rail network phasing map, from America 2050
42 replies on “Establishing Objective, Realistic Assessment Tools for Planning an Effective High-Speed Rail Network”
This is only mildly better than several of the other semi-braindead national HSR network plans that have surfaced recently. Why? None of the criteria above have anything to do with pre-existing levels of preparation or willingness to proceed.
As a North Carolina resident, this factor being ignored is increasingly frustrating. When groups like America 2050 continue to lump Washington to Atlanta together as one corridor, it shows that they are mostly engaged in fantasy planning.
Here in our neck of the woods, NC/VA are investing lots of money ($200+ million over 10 years) in existing track improvements in the Washington-Richmond segment and the Raleigh-Charlotte segment. Trains in these corridors are getting faster and more reliable all the time. Linking Richmond to Raleigh on a fast track where trains run mostly around 110 mph will create a track faster than the Acela in terms of average speed. But most importantly, NCDOT and VDOT have not been waiting for funding but designing track and bridges and retaining walls and so on.
South of Charlotte, South Carolina has not put a single PENNY into intercity rail in years. Members of both the SC and GA state legislatures refer to intercity rail as “socialism.” There is one study on rail from Charlotte to Atlanta, but there is no consensus on how to build, what to build, what speeds to use, etc.
Yet the America 2050 study treats these two corridors as if they are no different in substance.
Here’s NC/VA: http://www.sehsr.org/
Here’s GA/SC: http://atlanta.creativeloafing.com/gyrobase/georgia_s_rail_future_lags_behind_rest_of_southeast/Content?oid=798238
If wishes generated ponies, then yes, these are all good corridors to invest in. But in the real world, you have to do hard work to get these projects built. Who’s ready? Who’s spending their own money independent of stimulus funds? Who’s not?
Pass this plan through the filter of the questions above, and there’s your HSR investment priorities.
While transit connectivity is not crucial for an origin city, transit connectivity in the destination city is certainly a factor increasing the competitive advantage of HSR of all classes in both capturing existing air and car trips and creating new trips, so while it is easy to overstate its important, it merits inclusion as a factor.
Congestion would mostly be relevant to the extent that there is a transit system in the destination city that allows bypassing the congested streets, so it would seem to be at best an amplifier of the competitive benefit of transit connectivity in the destination city.
An omitted factor seems to be network economies. It would be, of course, a foolish transport policy that emulated the highest speed component of other nation’s rail systems and left dawdling Amtrak speed as the next rung down. But the above approach in effect presumes just that.
For example, an Express HSR corridor from NYC through northern PA and Northern Ohio and with appropriate junctions with the Ohio Hub would host Express HSR services NYC/Cleveland, NYC/Columbus/Cincinnati, NYC/Columbus/Indianapolis, and NYC/Toledo/Detroit. An Express HSR corridor from Chicago connecting from the other side would host CHI/Columbus/Pittsburgh and CHI/Cleveland/Buffalo. However, since the route itself is NYC/Canton/Fort Wayne/Chicago, it doesn’t even merit consideration as a third stage system.
Indeed, since at their current state of development the Acela services themselves do not provide effectively faster services than a Regional Higher Speed corridor would do in a less built-up right of way, their success provides every bit as much support for Regional HSR as for Express HSR.
As a stand-alone corridor, city pairs less than 100 miles apart would not justify the increased expanse of a bullet train Express High Speed Rail corridor over a Regional Higher Speed Rail system – but where there are corridors in the frame, such as Miami / Tampa, Miami / Orlando, the effectiveness of a single corridor including the three should include the two closer cities.
There’s obviously something wrong with their methodology since they think that Dallas-Houston should get priority over Dallas-Austin. Probably it’s the insistence on city pairs, which makes sense for an airline, but not so much sense for a railroad. Intermediate stops count. There aren’t any between Dallas and Houston. There are a lot between Dallas and Austin.
And Patrick M gets it exactly right. Until something like NatTrack is established (and I’m not holding my breath), infrastructure will continue to be built by states. To the extent that the federal government provides funding, it can put constraints on what it’s willing to fund, but the initiative has to come from states or groups of states. What struck me about the HSIPR preapplications was not so much that they added to over $100B, but that so few of them were for frequent service at speeds higher than 79mph that could actually start construction within the next three years: Midwest, NC/VA, California and extensions to the NEC. Most states aren’t even thinking about this stuff, they don’t have any staff prepared to think about this stuff, they aren’t going out and hiring staff to think about this stuff. Any national plan that ignores the massive differences between states’ readiness to participate is doomed to failure.
I’ll throw out my dream network: http://img22.imageshack.us/img22/300/hsrmapfrom2050base.jpg
The problem that I always with things like this is that they make it sound like you can’t do any other corridors, or you have to build certain ones before others. That’s not the case at all. We need to build all of these corridors, and more (some corridors like Houston-New Orleans and New Orleans-Atlanta. And as you said, there are a lot of corridors under a hundred miles that aren’t even considered here. There’s really no point in making studies like this that seem to restrict our options.
How did the federal government phase the interstate highway system? Some of it was obviously built long before other parts, but there was definitely an overall plan. I think a big picture view is critical to these types of projects.
It is important to take note of what each state is willing to contribute, but the federal government should plan for an interconnected interstate system. They built the interstate highways almost entirely with federal dollars. The same should be true for train lines. Atlanta to Charlotte is just as important as Richmod to DC, and the attitudes of state governments don’t really cahnge this fact.
The St Louis to Chicago route has gotten a lot of attention because of the Illinois’ efforts to speed up that route, but if you think about it, a Chicago to Cincinnati rotue would be the same distance and serve a lot more people. There are at least an extra million people along a route of the same length. If each line were expanded, the Cincinnati route could quickly reach millions more customers via Dayton, then Columbus, and eventually Pittsburgh. The St Louis route would have to go all the way to Kansas City or Memphis, hundred of miles, to reach the next major population center.
What jim notes about Dallas/Houston is another example of not considering network economies, which is certainly a far more important factor than congestion – for example, <i.existence of a 125mph Regional HSR corridor from Washington through to North Carolina substantially increases the benefit of an Express HSR from NYC to Washington DC that bypassed the NEC, and raising the maximum speed on the NEC with upgraded catenary increases the benefit of an Express HSR south of DC.
And, of course, even if the Regional HSR services are rolled out incrementally, first as Emerging HSR corridors, with operating surpluses funding the upgrade to Regional HSR corridors, they can still be fully rolled out before the first train can run on the first Express HSR corridor that begins construction.
For Texas, the impact on ridership of the extra travel time required for the T-Bone rather than a direct alignment is quite modest, since at 220mph its not very much additional time … and it is more than offset by the additional ridership between Houston and the junction with the Dallas/San Antonio alignment.
Here in Ohio, the Cleveland to Toledo leg of an Express HSR corridor from Cleveland to Chicago would cost more than the Emerging HSR option for the entire Ohio portion of the Ohio Hub, and likely more than the entire Ohio Hub upgraded to 125mph … and the entire Ohio Hub, all the way to Toronto, Buffalo, Pittsburgh and connections to Chicago via Indianapolis and Fort Wayne, would be finished before any segment of Cleveland/Chicago could be finished. And Chicago/Cleveland is slated into Stage 2, so the Ohio Hub could actually be finished before construction on Chicago/Cleveland would likely start.
Working out a multi-decade roll-out of Express HSR corridors through Ohio while pretending that no other interurban rail system will exist in Ohio is just plain silly.
Can someone succinctly summarize the reasons why a Gulf Coast corridor (in one form or other) almost never makes it into even the top three priority tiers in any of these studies? There is a huge amount of air traffic between Houston and NOLA for the petro-industrial complex, and the distances are ideal for a Houston–NOLA–Mobile–Atlanta/Jacksonville corridor.
Similarly, why do all high-speed proposals never consider St. Louis–Memphis–Jackson–NOLA via the Illinois-Central; is there really not the population there to support HSR?
The reason the study shouldn’t look at state preparedness is that no state, except California, has done any work for HSR. North Carolina’s preparing to increase speeds to 110 mph, which was never called HSR, anywhere. The minimum speed at which service could be called high-speed was always 125 mph, and in newer systems, such as south Korea, they consider 186 to be the minimum.
Intermediate stops count, but less than you think. In Texas, Houston has a higher metro population than all cities on the I-35 corridor except Dallas, combined. About the only problem in the study with not considering intermediate cities is that it doesn’t evaluate alternate routes when necessary, such as T-Bone versus a triangle routing, or Chicago-Detroit via Kalamazoo and Ann Arbor versus via South Bend and Toledo.
My personal pet peeve with the study is that it doesn’t consider construction costs. In flat land with existing straight ROW – e.g. Chicago-St. Louis, Chicago-Cleveland – construction costs will be low. On the NEC, with its HSR-grade tracks that just need new catenary and curve straightening, construction costs will be even lower. Conversely, in mountainous terrain, such as Pennsylvania and California, HSR will need many tunnels and new ROW, and that will inflate construction costs.
Jeff,
The Gulf Coast is well behind the other megaregions in terms of economic standing. The oil industry you speak of is the only large industry that uses a lot of human capital. Engineering, Finance, Research, and other “smart” industries are much more prevalent in other places.
It is also not a corridor that has a lot of support from the states in connects.
The answer is the same for the lower Mississippi River Valley.
Look at the metro GDP figure on page 5 of America 2050’s pdf linked in the original post. (“new study”)
Jeff: NOLA only has a metro population of about a million. Pre-Katrina, it was at 1.3 million.
Correction, Yonah: if the map is to be believed thats a Vancouver-Seattle-Portland corridor, not a Vancouver-Seattle corridor.
Thanks, good spotting Alexander.
In the areas south of St Louis and Cincinnati, east of Houston and west of Atlanta, there are no cities with metropolitan populations over 2 million. The distance from St Louis to NOLA or Houston to Atlanta is simply too far to make a worthwhile route. Stringing together a series of cities whose populations are far apart and rarely approach or exceed a million people doesn’t work. Houston and Dallas each have 5 or 6 million. St Louis and Kansas City each have over 2 million. Atlanta and DC each have 5 million, and there are three 1 million + cities in between. NOLA has about 1 million, and only Birmingham (1 million) is between Atlanta and NOLA.
Well, despite the methodology problems they seem to have a decent set of corridors. I disagree with their prioritization in some ways but not in others.
The top priority, nationally, should be Chicago-Eastward. There is an appropriate ROW with very little use from Chicago to Fort Wayne and Fort Wayne to Toledo, though it’ll need a lot of grade separations. Working incrementally from Chicago east gives you benefits at every stage: first you bypass the Chicago terminal congestion with Englewood Flyover; then you knock 10 minutes off southbound runs with the Grand Crossing connection; then you speed up the slowest tracks in the entire Chicago-eastward Amtrak runs with new tracks on the “NYC” right-of-way through Chicago; then you bypass the Western Indiana congestion with flyovers and new ROW around Gary; then you speed up runs and give service to Fort Wayne with the Chicago-Fort Wayne-Toledo build; restore Detroit eastbound service with a Toledo-Detroit connection; eliminate slowdowns in the Toledo terminal area; each step has independent utility and it adds up to an exclusive, very straight passenger ROW from Chicago to Toledo at least.
The second priority should be Chicago northward; Milwaukee-Madison is a huge untapped market.
The third priority should be Chicago-St. Louis, but they should totally do it on the Midwest HSR’s proposal, coming south on exclusive tracks next to the CN/IC tracks; it gives more bang for the buck.
North Carolina/Virginia, California, and the Pacific Northwest projects are all great too, but the Chicago ones give the most “national benefit” and are largely very straightforward. The “Inland Route” via Springfield in Connecticut and Massachusetts is great, but the work needed to straighten, speed and widen those antique ROWs is HUGE, compared to the straightforward ROWs around Chicago.
As another point, I think the priority should be to build ROWs which are “high speed ready”. Build the geometry for 125 mph and set it up for future non-disruptive installation of catenary. But don’t necessarily build the catenary now or maintain the tracks to the top track class, particularly around Chicago; speeding up the existing trains to 110 mph will build the ridership and therefore the demand for faster trains. And leaving out the occasional expensive grade separation will reduce capital costs in the short run as long as the geometry is constructed so grade-separating later won’t involve a complete reroute.
On the contrary, building new tracks without electrification is stupid, whereas electrifying existing tracks can be useful. Diesel trains are heavier than electric trains, increasing track wear; they’re unlikely to be able to get FRA waivers, increasing weight and track wear even more. On many lines, which are more than double-tracked and which are reasonably straight, electrification and track rebuilding can significantly increase speeds and reduce maintenance costs, without adding new track.
In addition, electrification means that trains can run for part of the way at high speed and for part of the way at lower speed; this is the model used in France and Korea. It’s also the strategy pursued for the NEC and Keystone Corridor: the slow, undermaintained, curvy tracks in Connecticut and on the PRR Main Line allow through-routing of trains from faster lines, such as NY-Philly.
The Inland Route is completely overrated for getting from New Haven to Boston. The Shore Line is actually pretty good – it’s curvy as hell, but there’s room for bypass tracks on I-95 from New Haven to Rhode Island, where the tracks are as straight as an arrow. The real problem for New York-Boston is the tracks between Port Chester and New Haven, which require multiple elevated and underground bypasses, including 1-2 underwater tunnels (see concept here).
@ Alon: “On the contrary, building new tracks without electrification is stupid, whereas electrifying existing tracks can be useful.”
You included all sorts of reasons for targeting an all-electric system as the final result, but did not include anything to support the thesis that “building new tracks without electricification is stupid”.
It would be axiomatic if the capital costs were the same, but since they clearly aren’t, it does not follow.
On corridors that can generate operating surpluses at 110mph, getting the services running with the capital investment required for diesel 110mph services is part of the process of funding the network. Revenue bonds can then fund the state match for the electrification improvements.
Alon,
“North Carolina’s preparing to increase speeds to 110 mph, which was never called HSR, anywhere. The minimum speed at which service could be called high-speed was always 125 mph, and in newer systems, such as south Korea, they consider 186 to be the minimum.”
True. But NC plans to engineer the route (at least between Raleigh and Richmond) so that electrification will be possible down the road.
In response to Patrick M. from North Carolina, the reason we focused this study on measures of potential ridership instead of project readiness criteria, is because we understand the FRA will give ample attention to issues like preliminary engineering, local support, matching funding, and a regional vision. What we felt was missing from the FRA’s process for judging the proposals were criteria that compare the corridors to one another in terms of potential ridership. So the analysis in the study focuses on what we think those criteria would be, in the absence of any “on the ground” factors. Ideally, the FRA will use both the “on the ground” factors and estimations of ridership to make their decisions.
Thanks for all the good feedback.
The capital cost of electrification is higher than the cost of no-build. However, it’s not higher than the cost of installing signals capable of 110 mph top speed. Wikipedia writes of the Ohio Hub,
The cost of electrification is about $1 million per mile, and allows shorter runtimes merely by improving acceleration and deceleration rates. It allows even shorter runtimes by allowing through-routing on higher-speed lines. In principle timed transfers work nearly as well, but in practice Amtrak’s ability to time anything is weak, and will remain so without extra signaling like PTC (i.e. guaranteed time separation from freight).
Electrification is always possible down the road. That’s not the problem; the fact that the route would never support high speeds is. One of the things Japan and France learned by accident is that there’s never a good reason to try to get more mileage out of an existing low-speed route that’s in demand when you can build a parallel high-speed line.
Alon- the track geometries from Richmond to Raleigh, while presently aiming for 110, are such that if a diesel-powered engine could hit and maintain 125, the trains would be able to take advantage of the straight track sections and go faster. Of course, electrification is another way to do this but it is very costly.
Also, Charlotte to DC is more important than Charlotte to Atlanta, for two reasons. First, you connect to the NEC in DC. You connect to no other rail network of any substance in Atlanta. Second, the population sizes of the intermediate cities between Charlotte and Raleigh are much bigger than those between Charlotte and Atlanta. NC, in addition to planning for interstate HSR, also plans for several intrastate frequencies at 89 mph. In short, NC is planning for services that will act as feeders to HSR stations. SC is planning to well, sit on its hands.
Of course everyone would love french-style TGV speeds but the biggest deterrent to this in the US is the cost of ROW acquisition and sprawling metro areas. France’s more compact settlement patterns made it much easier to find stretches for straight track. I firmly believe that the best way to expand HSR longterm in this country is to get a few good corridors up and running.
Petra- thanks for coming over and responding. I hear what you are saying on the vision piece not being part of the FRA worldview, and that’s a legitimate need you are trying to fill. Any quibbles I have with the plan are still outweighed by the fact that people care enough to make a plan.
In closing, to make on more suggestion, I would recommend that you de-emphasize the “other corridors” even further, or convert them all to bus. HSR is fundamentally an intercity rail travel technology.
My take on phasing is that politics and local priorities will heavily influence any national plan, so there’s no point in getting to worked up about the precise order. On the other hand, I do think it’s important determine early on what routing and levels of service (180mph vs. 79mph, etc.) are ultimately desired. That way when states do act, whether in 2010 or 2050, they will put their efforts into something that fits with the national network.
Click on my name to see my HSR network proposal.
@Alon: “The capital cost of electrification is higher than the cost of no-build. However, it’s not higher than the cost of installing signals capable of 110 mph top speed.”
An electric 110mph line requires the cost of installing signaling, new track and/or upgrading track including passing track, grade separations and 110mph standard level crossings, and the electrical infrastructure. The 110mph diesel system requires that except not the electrical infrastructure.
And of course, the constraint on the pace of expansion is availability of capital funding. With sufficient funding, the first phase of the Ohio Hub could be completed in two years and each phase after that in one year. Meanwhile, it has taken a number of years to secure funding for building the first stage for Amtrak speeds, and once the service is running it might take a year to secure funding to upgrading the Berea to Columbus stretch to 110mph to get it to a positive operating ratio in another year or two, to secure funding to get the trunk corridor to 110mph.
And of course, the upgrade to 110mph can be done segment by segment, since on a non-upgraded segment, the train simply observes the 79mph speed limit. The upgrade to electric cannot, it must be done a full corridor at a time, since an electric train needs electric supply from go to whoa.
So the quickest way to an all-electric Ohio Hub is to get the 110mph services started as soon as possible, so that the services can begin contributing to funding their own expansion as soon as possible, so capital funding is not held hostage to the Republican controlled State Senate.
If it is first built as a 110mph diesel system, then upgrading it to electric and upgrading to 125mph are projects for which the system itself can provide state funding, which is much harder for State Senators from areas receiving the benefit to object to.
Patrick — I don’t get it. What is gained if you “de-emphasize the ‘other corridors’ even further, or convert them all to bus”?
And how does this statement “HSR is fundamentally an intercity rail travel technology” argue for dumping conventional rain for buses?
I’m always in favor of both: Markedly speed up trains to HSR on routes where it makes sense to do so. Add frequencies, new or restored routes, new passenger cars, and incremental speed improvements throughout the national system.
Your comment seems to suggest it’s somehow either/or. Why?
This problem is unique to the Southeast Corridor. Not even the Interstate alignments there are straight enough to avoid needing eminent domain. On the other hand, due to the fast growth of cities on that corridor, it’s better to bite that bullet now than in 2029, when those sprawling metro areas have doubled in size. Today I-95 + I-85 still only needs moderate amounts of eminent domains, especially if Amtrak’s willing to serve intermediate cities TGV-style, on the edge of urban development instead of at the center.
Yes, the 110 mph diesel system requires slightly less infrastructure, which works out to a cost reduction of 17%. However, in practice, operating costs are lower with electrification, and there is more competition worldwide among rolling stock vendors, since the largest markets are almost entirely electrified. This greater competition means that less contingency is needed – the technology is more reliable, and if one vendor screws up, Amtrak can switch to another. Denmark is having a lot of problems with its diesel trains, which, due to lack of expertise among train manufacturers, had large cost overruns and service disruptions. The best way of avoiding such problems is to electrify and repeal the FRA’s crash safety regulations.
Furthermore, electrification by itself improves performance, which means not just lower operating costs but also higher ridership. The rationale is the same as for increasing speeds.
But electrification costs one fifth as much per route-km as a 110 mph upgrade…
Or, if it’s first built as a 110mph diesel system, it will bleed money, emboldening Republican politicians who want to kill the project on budgetary grounds.
Remember: health care, education, and urban mobility are fundamental rights. Intercity travel is a business investment.
Because in practice, Amtrak’s insistence on serving Montana and Mississippi is causing it to deemphasize experience with high-volume passenger railroads in its management. It has no mechanism for promoting people who know what they’re doing with the NEC or with supervising HSR construction. Instead, it denigrates such high-priority projects: it even uses misleading statistics to argue that all trains worldwide are subsidized (it counts commuter rail in France, which is operated under contract by SNCF, but not in the US, where Amtrak stays away from it).
Amtrak is ironically the only body in the US that’s not paid off by the highway industry that believes high-speed trains are losing money. If it admitted that trains could in principle make money, it would have to explain why its own high-speed trains are a flop. So instead of expertise, it encourages rent-seeking.
Regarding Amtrak service to places like Montana and Mississippi–it is interesting to explore why:
Is it because
* an internal organizational determination that service ought to be provided to remote locations, as part of Amtrak’s organizational mission, despite the financial losses?
* an external determination (either explicit direction from lawmakers, or as a form of patronage to lawmakers from particular states) that service ought to be provided to these locations, or else have funding withdrawn?
* These places happen to be on the way to somewhere else? If you want a train from Seattle to the Great Lakes region, the most direct route takes you through places as Idaho, Montana or Wyoming, and the Dakotas; similarily, the most direct route from Texas to Florida passes through Mississippi.
While Amtrak’s management can be obstinate, and probably ought to be re-organized some day–I’m not sure how else they would act given the current funding climate. As an organization, the hostility they face in DC is tremendous.
EngineerScotty: it’s all of those issues, together. Montana gets service because it’s between Chicago and Seattle, Mississippi because it’s between New York and New Orleans.
The real issue is that Amtrak shouldn’t be operating such trains in the first place. It operates those trains because of the two other reasons you mention. Even the NEC isn’t consistently profitable, which means Amtrak needs to provide pork to legislators in Montana and Mississippi to keep drawing money from Congress. This situation is contributing to a culture that supports long-distance trains for two reasons:
1. Rent-seeking – because Amtrak’s current business model is unprofitable, the company thinks in terms of metrics like passenger-miles and expenditures instead of revenue or profit.
2. Nostalgia – the lack of modern train service in the US encourages a “What was good for my grandfather is good for me” view among railfans. National rail advocates still talk of the California Zephyr and Empire Builder, which were premium train services in the 1940s and 50s. It’s no different from New Yorkers’ 1930s nostalgia, or auto fans’ GM nostalgia.
Reorganization wouldn’t achieve much. Even outsiders adapt to the Amtrak culture easily – David Gunn, a modernizer by Amtrak standards whose background is in running rapid transit systems, insisted that a national train system is a national asset like the national parks.
If you want to make Amtrak work, beg your representatives in Congress to press the FRA to repeal its crash safety rules, or pass a High-Speed Rail Act authorizing $5-10 billion for 220 mph HSR between Boston and Washington. Even replacing the catenary from New York to Washington and increasing superelevation on curves between New York and Boston might be enough to make the Acela profitable.
There was a suggestion a few years ago to split up Amtrak: create a new entity (possibly controlled by a consortium of the NEC states rather than the federal govt.) to own the NEC and run the NEC trains liberating Amtrak to concentrate on the long distance trains — the national assets like the national parks.
That might be revived, perhaps modified into a three-way split: the NEC, the Midwest and the long distance trains.
Then the NEC entity could look at what it needed to do to make the NEC profitable and its states could work out a deal with the FRA or Congress.
FWIW, my impression is the long distance trains exist purely because of political pressure. There are only three states Amtrak doesn’t run a train through: Hawaii, Alaska and Wyoming. The Cardinal would surely have been put out of its misery long ago if it weren’t that it serves West Virginia and Kentucky.
Having Northeastern states control the NEC wouldn’t improve it too much. The two largest Northeastern states, New York and Pennsylvania, both have complex upstate/downstate politics that in practice preclude any investment in New York and Philadelphia unless it comes together with equal investment in Rochester and Harrisburg.
I don’t remember any more whether the proposal to split off the NEC left it in federal control or not. Nor do I remember why it didn’t happen. I can think of a number of arguments against giving it to the NEC states. But intrastate regional rivalries aren’t one of them. The Washington Metro is controlled by a consortium of Maryland, Virginia and the District. Maryland has a well-known Baltimore vs. Washington suburbs rivalry, but it mostly doesn’t spill over into Metro. The other consortium members would, I think, become very impatient were it to do so.
But Amtrak has a clear conflict of interest here. Between Washington and New York it runs almost as many long distance trains as Acelas. It runs them on the same tracks, only a little slower. It timetables to make this possible. Were Amtrak to upgrade the Washington-New York track to enable 220mph, it would no longer be able to run the long distance trains on that track. It would be put to a choice: start, say, the Crescent at Washington or run it on the local track shared with commuter trains — MARC, SEPTA and NJT — which run slower even than the freights it encounters outside the corridor. Neither solution is satisfactory. As long as Amtrak doesn’t upgrade the track that much, it isn’t faced with this problem. Leave the track with Amtrak and it will never be upgraded.
Alon Levy:
“EngineerScotty: it’s all of those issues, together. Montana gets service because it’s between Chicago and Seattle, Mississippi because it’s between New York and New Orleans.
The real issue is that Amtrak shouldn’t be operating such trains in the first place.”
Of course, Andrew Seldon argues (YouTube) that its the short corridor services that lag the long corridor services in terms of the actual measures of performance – load factor, passenger miles, revenue – which implies that if they were to shut down services based on cost per passenger mile, its the regional corridor services that would get shut down first, not the long haul services.
Which is not at all surprising – shorter regional corridors will not be able to fully tap their other competitive advantages until the trains are as fast as driving, and will not gain a time of travel competitive advantage, obviously, until they are faster than driving.
Bruce, Andrew Seldon is wrong. You can see how Amtrak performs here – the relevant route performance data is on page 60. In FY 2008, long-haul services lost an average of 13.9 cents per seat-mile and 24.7 per passenger-mile; the comparable numbers for non-NEC short haul services were 5.5 and 13.
In fact, of the top 10 money losers both per passenger-mile and per seat-mile, only two are short-distance, the Hoosier State and the New Haven-Springfield Shuttle.
Jim, Amtrak could run long-distance trains at 220 mph on the NEC, and then do an engine change at Washington and run them at lower speed further south. Such trains would be faster than today’s trains even on the lower-speed section, since they’d have to feature tilting carriages to run at high speed on the NEC. This would allow higher speed limits on curves, permitting trains to maintain 79 mph speeds for longer stretches of track.
Alternatively, there are enough passing sidings on the NEC to permit both 125 mph Regionals and 220 mph Acelas; the long-distance trains would run as Regionals. Much of the route in New Jersey is six-tracked. Pennsylvania, Delaware, and Maryland have light enough commuter volumes (3 tph, peak) that the Regionals could share track with commuter rail. All that’s needed is schedule discipline, and maybe one six-track section located halfway between Philadelphia and Wilmington.
In Texas in the late ’80’s to early ’90’s, a private consortium, under the name “Texas TGV” paid for and finished Environmental Impact Studies for a TGV-based HSR in Texas. A “star” arrangement was the decided upon alignment (similar, I suspect to the “t-bone” arrangement mentioned above), serving the DFW region, the Austin/San-Antonio Region, and the Houston region.
I believe they even went so far as to lay out plans for acquisition of right-of-way for the rails.
The project was shelved when the state refused to provide loan guarantees to the private consortium. But all the ground-work was FINISHED. TWENTY years ago.
Think about it a little bit…
(I also heard unsubstantiated noise at the time that Southwest Airlines was heavily opposed to the plan, and fought it vociferously, since it was going to be in direct competition with the then still-fledgling airline, whose major cities were still DFW-Austin-Houston.)
-blaine
“Do the study’s authors really think that there would be more traffic between Hartford and Philadelphia than between Hartford and New York, because the latter route is shorter?”
That would indeed be a strange conclusion, given the extremely heavy volumes of bus, car and rail traffic that already exists between Hartford, NYC, and points in between.
PeakVY@24 —
Very nice map making, thanks. I especially liked having the rank of the major metro areas inside their dots. It’s good to be reminded, for instance, that Pittsburgh is still a top 20 metro area, while Norfolk falls in the mid-30s. Looks like Norfolk will soon have better train service than Pittsburgh, and why is that?
And superimposing the French TGV routes on the US map was sobering. We have a long way to go and a lot of ground to cover.
It’s easy to upgrade the rail service to Norfolk – just run trains on an existing freight line, which is as straight as an arrow. Serving Pittsburgh from any direction requires either long tunnels, or winding, low-speed lines.
Patrick — I don’t get it. What is gained if you “de-emphasize the ‘other corridors’ even further, or convert them all to bus”?
And how does this statement “HSR is fundamentally an intercity rail travel technology” argue for dumping conventional rain for buses?
I’m always in favor of both: Markedly speed up trains to HSR on routes where it makes sense to do so. Add frequencies, new or restored routes, new passenger cars, and incremental speed improvements throughout the national system.
Your comment seems to suggest it’s somehow either/or. Why?
Woody- I also support the expansion of conventional rail networks (and bus networks!) throughout the US. But I am also concerned that if we consider a rail line from Denver to Chicago as a “other potential” high-speed service we will be diluting the definition of HSR in the way that Bus Rapid Transit has been cheapened in the USA to mean new shiny buses and better-than-miserable amenities.
Another way to achieve what I am talking about with this map is to relabel the other corridors “possible conventional rail corridors.”
As we pour another $14 million into worthless maglev studies from Atlanta to Chattanooga, I’d like visions for high-speed rail to focus on where HSR makes the most sense and can be built with the greatest benefit- and that is in metropolitan America.
The Brookings Institute wisely points out “the largest 100 metros represent just 12 percent of the nation’s land area and 65 percent of its population but account for 74 percent of the nation’s college graduates, contain 77 percent of all good-paying “knowledge” jobs, originate 78 percent of all patent activity; and handle 79 percent of the nation’s air cargo in this global trade economy.”
We have recently achieved a breakthrough spending initiative on High Speed Rail. It will be partially squandered if we spend this money improving 59 or 79 mph tracks across large stretches of rural America rather than linking metropolitan areas or large and medium size.
Well, this map does contain several errors. For example, Montreal and Toronto are 3 million + metro areas. These centers are moving towards 4 actually.
We hope that the US can convince the Canadian to allow Boston and New-York to be linked to Montreal which could be become the HST hub in eastern Canada. This would make sense since Bombardier, the world’s largest train manufaturer, could supply the Zefiro (rated 320 km/h) powered by Hydro-power (the cleanest form of energy). This would have a tremendous impact for the economy in Quebec, but the actual government in Ottawa never favoured a Montreal based choice which could help its economy and will always be protecting the car market based in Toronto (reason why it saved GMC) and securing Air Canada’s monopoly, sadly offering very poor service to Canadians.
Maglev is not a valid choice for long distance. You have to understand that precise track surface finish over very long distances is almost impossible. These might lead to major accidents. Maglev could be used to link airports to mega-city centers. But the recent cancellation in Munich for the Maglev project showed that it is not an economic solution for even a large city such as Munich.
Even the Shanghai project was never completed by Siemens, but instead Bombardier. This shows how difficult these technologies are.
The alternative would be to consider high speed monorail systems such as the Trensquebec: http://www.trensquebec.qc.ca/
These could be installed over highways at just twice the cost per km of highways, meaning much cheaper then HSR.