National Network Discussion (split from Hoosier State discussion)

Amtrak Unlimited Discussion Forum

Help Support Amtrak Unlimited Discussion Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Status
Not open for further replies.

Anderson

Engineer
Joined
Nov 16, 2010
Messages
10,407
Location
Virginia
I would agree with Jis broadly, but I do think there are five counter-points:
(1) The National System has wound up being a bit too thin. There's a strong case that there are several routes that should be added (the Pioneer, Sunset East, and a few other routes come to mind...most of which were cut in the 1990s or early 2000s).

(2) There are some routes wherein the addition of said route would likely be an incremental improvement to Amtrak. Some expansion of the Florida service is a good example; so, too, would be expansions of service within Virginia. This might be mainly an east coast phenomenon, but it still stands as something worthy of note.

(3) Trying to get multiple states to coordinate on a project runs the same risks that trying to get multiple towns to cooperate on the Hoosier State runs: If one pulls out a project may become tricky, if not impossible, unless another state is willing to pick up the slack. For example, while it might be concievable that NY, PA, and IL might cooperate on some sort of expansion of east-west service, either IN and OH would need to get on board or the others would have to pick up the cost. Same thing with additional service to Florida: South Carolina and Georgia stand in the way in many respects. Getting one or two states on a project is one thing; trying to get more than that is likely to be a major issue.

(4) There's a difference between having to "pay something" (which I believe that all of the states that currently have corridor trains, save for IN, were prior to PRIIA 209) and having to pick up the full tab...especially in the face of probable intransigence from other states. There's also the issue of states not even being able to clearly get help with starting up a service, and without federal grants to kickstart things, the Heartland Flyer and (if I'm not mistaken) Lynchburger would not exist today.

(5) Finally, there's the issue that a lot of the losses out there are down to chronic, decades-long equipment shortfalls. Amtrak has been incessantly short on equipment, and in the case of the eastern states there has been no opportunity put in front of them to actually get their own equipment and no attempt to organize or fund anything like that. Thus where in a few years most states west of the Appalachians will likely only be paying equipment charges on their locomotives (if on that) while running fleets of new cars mostly paid for with federal money, those on the extended NEC will be stuck paying Amtrak to use 40-year-old Amfleets out of their operating budgets.
 
I would agree with Jis broadly, but I do think there are five counter-points:

(1) The National System has wound up being a bit too thin. There's a strong case that there are several routes that should be added (the Pioneer, Sunset East, and a few other routes come to mind...most of which were cut in the 1990s or early 2000s).
Eh? Equipment intensive routes with low ridership and cost recovery shouldn't be coming back. Maybe adding something so that Ohio actually has a useful train, but the Pioneer, Desert Wind, and Sunset East died worthy deaths.

(3) Trying to get multiple states to coordinate on a project runs the same risks that trying to get multiple towns to cooperate on the Hoosier State runs: If one pulls out a project may become tricky, if not impossible, unless another state is willing to pick up the slack. For example, while it might be concievable that NY, PA, and IL might cooperate on some sort of expansion of east-west service, either IN and OH would need to get on board or the others would have to pick up the cost. Same thing with additional service to Florida: South Carolina and Georgia stand in the way in many respects. Getting one or two states on a project is one thing; trying to get more than that is likely to be a major issue.
Sounds to me like an argument in favor of keeping corridors short and frequent rather than long overnight trips spanning multiple states.

(5) Finally, there's the issue that a lot of the losses out there are down to chronic, decades-long equipment shortfalls. Amtrak has been incessantly short on equipment, and in the case of the eastern states there has been no opportunity put in front of them to actually get their own equipment and no attempt to organize or fund anything like that. Thus where in a few years most states west of the Appalachians will likely only be paying equipment charges on their locomotives (if on that) while running fleets of new cars mostly paid for with federal money, those on the extended NEC will be stuck paying Amtrak to use 40-year-old Amfleets out of their operating budgets.
There's no reason that the Eastern states couldn't have done as the Western states, or North Carolina for that matter, did and at least start planning on purchasing their own equipment. The Western states were prepared for when opportunity knocked. Did the Eastern states even apply for rolling stock under the HSIPR program?
 
The Pioneer and Desert Wind were both rather mismanaged by Amtrak. As I recall, the biggest problem with the Pioneer was awful arrival/departure times at the main destinations along the route, while the Desert Wind could have done a lot better if more space had been available from LA to Vegas (I recall seeing on here that there were regular space lockouts on that segment like there are west of Reno on the Zephyr). If you give a train an atrocious timetable, don't expect good performance. Same if you under-equip a train, though that was at least more excusable.

The Sunset East was screwed by being 3x daily...much as the Cardinal and Sunset West have been. I also don't think that too many of us put much faith in the Sunset East report that was produced a few years back.

And of course, there's the fact that those trains were all axed in an era when LD ridership was somewhere around 1/4 lower than it is now, and ridership overall was over 1/3 lower. Perhaps the best way of underlining this is the fact that in FY04 (the last whole year the train ran through), ridership on the whole Sunset was 96,426. In FY14, ridership was 105,041 (which beat out FY03 by 8 riders for the best ridership on my chart). Just to be clear, this isn't a case of the train being cut back and improving as a result (something which actually makes some sense on that route); this is a function of overall ridership growth in the system.

-------------------------

With respect to "keeping routes short and frequent" for that reason, in the best of all worlds for your case you're doing the right thing for the wrong reason. I strongly suspect that you're doing the wrong thing for the wrong reason. Making them frequent, I agree with; keeping them short, however, I disagree with since you both restrict network expansion and drop links between regions in the process. I believe that it is more likely than not that in a decade there will be a robust, growing network in Florida run by FEC and at least some form of SEHSR approaching completion...and no ability to improve links between them. The same can be said of the Midwest and the Northeast. I can dig out a good chart that explains the logic here, but suffice it to say that there's a reason the LSL does so well (and as much as we can dicker over Neroden's calculations, I suspect we can agree that the "fully allocated" numbers aren't useful in this respect).

Moreover, CHI-MSP is not a "long overnight train", but WI has been playing merry hell with any efforts here, while IA's internal issues forcibly truncated a new service at the Quad Cities. Just because a train doesn't run on and on and on doesn't mean it isn't susceptible to this sort of nonsense.

-------------------------

As to the Eastern states' situation, there is a very valid reason that the states would have had a problem there without Amtrak's involvement, which I've discussed at great length with people both at DRPT and VHSR. With the exception of New York (the Empire Service/Adirondack/Ethan Allen being effectively off the NEC...and that mix is probably not big enough to support a full new car order on its own), there is a great deal of equipment interchanging among the states' routes and Amtrak's routes. For example, a Keystone set can be put on the back of a BC-QC-Coach-Cafe set to make a Regional, and the Carolinian, VA Regionals, and NE Regionals all interchange very straightforwardly. The only true oddballs are the Vermonter (a 5-car set IIRC) and the Pennsylvanian (which still does an interchange, this time with the Palmetto, likely dating back to the Three Rivers/Silver Palm situation).

Anybody going out on their own would lose their interchangability with Amtrak, which is of significant benefit to both parties. They'd also have to deal with their own maintenance issues. North Carolina basically did this with their Pullman Standard cars, and it has been a royal pain in the arse for them.

Amtrak is a necessary partner in any new equipment order on the East Coast in a way it frankly did not matter in the Midwest or on the West Coast. Truth be told, with MI, IL, and MO on board you're changing all of the corridor equipment in Chicago except the Hoosier State (which Amtrak would probably dump a few Horizons on and call it a day) and the Hiawathas (two equipment sets, and Amtrak would probably lean on them to get into similar equipment as the others eventually). Ditto California (MSBLs) and WA/OR (Talgos).

The East Coast /is/ different from these areas because if nothing else there's a risk that the states do one thing, Amtrak does another, and suddenly you end up with two sets of cars that don't play well with one another somehow and/or wildly variant equipment. The BC inconsistency is bad enough; I do not want to imagine what would happen if, just to use an example, Virginia decided they wanted to buy a bunch of Talgos. Just as these states are operationally tied to Amtrak in a way fewer and fewer states outside the region are, there's an equipment situation to consider as well.

There's actually a specific example of this cropping up: For some time, VA has wanted to buy dual-mode locomotives to use on its Regionals. VA has talked with Amtrak about this, and they'd need to order a decent number of extras to allow smooth equipment exchanges in NYS and BOS, since a few of VA's trains either don't go to the same locations on all days (a few terminate in New York) or don't exchange equipment with one another directly.

Even if Amtrak did not need to play a role, it would behoove Amtrak to be active in such a pursuit: The MSBL orders are going to allow Amtrak to move around a bunch of Superliner coaches onto other trains (perhaps allowing some things such as the much-desired "Sparks Cars" on the Zephyrs); replacing the LD Amfleets used on the Adirondack would allow a fifth coach on the Meteor when demand would allow it while replacing SD Amfleets would allow Amtrak to add those to Regionals and other trains (an SD Amfleet on an LD train for short-haul passengers would be a decent use of equipment, and if worse comes to worse...couldn't an Amfleet I have an Amfleet II seating pattern put in it (or vice-versa)?
 
Anybody going out on their own would lose their interchangability with Amtrak, which is of significant benefit to both parties. They'd also have to deal with their own maintenance issues. North Carolina basically did this with their Pullman Standard cars, and it has been a royal pain in the arse for them.
Please elucidate about North Carolina's pain.

jb
 
All of the much maligned (by the likes of URPA, RailPAC and assorted other detractors) north eastern states have huge internal state service separately funded from the national system.
Except New Hampshire, which is just freeloading. (I wish NH would get off its duff, frankly.) Delaware and Rhode Island are piggybacking on Pennsylvania and Massachusetts, respectively.

There's no reason that the Eastern states couldn't have done as the Western states, or North Carolina for that matter, did and at least start planning on purchasing their own equipment. The Western states were prepared for when opportunity knocked. Did the Eastern states even apply for rolling stock under the HSIPR program?
There were *much* higher priorities for New York -- buying the Empire Corridor as far as Schenectady, improving the tracks, and rectifying the substandard stations which dated to the NY Central's "destroy all passenger stations" era. California, Washington, and NC did all that station stuff a long time ago, and the stations were never as dismal as the NYC/PC shacks. Similarly, Connecticut's priority was rehabilitating the Hartford Line tracks & stations -- they are planning to buy rolling stock, but the tracks & stations come first. Pennsylvania's focus was, again, on the substandard station conditions and deferred maintenance on the SEPTA lines. You get the picture.
New York got burned once by buying rolling stock before buying track improvements ("Turboliners" may ring a bell).
 
Last edited by a moderator:
Anybody going out on their own would lose their interchangability with Amtrak, which is of significant benefit to both parties. They'd also have to deal with their own maintenance issues. North Carolina basically did this with their Pullman Standard cars, and it has been a royal pain in the arse for them.
Please elucidate about North Carolina's pain.

jb
North Carolina has a stand-alone fleet of something like a dozen passenger cars which are basically the last non-Superliner Pullman Standard cars out there, so they're basically running a small PV fleet of equipment. When parts break (as parts do on occasion), they're stuck doing the same sorts of things that Amtrak has to do for their Heritage fleet with fabricating or acquiring specialist replacement stuff. The equipment is in better shape than Amtrak's by virtue of having been basically new in the early 90s, but that doesn't necessarily mean that much.
 
The Pioneer and the Desert Wind were discontinued in order make a few trans-cons daily from less than daily. AFAIR there was a significant equipment issue possibly caused by the usual insufficient maintenance budget. Something would have been cut. It was just a matter of which.
 
Since the Desert Wind and Pioneer came out of CHI as part of a Super Train with the Zephyr, it wasn't possible to have better calling times in the West after the train was split so the three trains could head for their Western destinations!
 
I don't think the Pioneer or Desert Wind ever had much potential. Lack of intermediate population.

Sunset East is another matter; the advocates have presented solid data showing that 41% of the revenue of the train came from the eastern segment, which is less than 28% of the route-miles (on the version which went all the way to Miami). Arguably the western half should have been cancelled instead, leaving a daily train from San Antonio to Miami. Anyway, a topic for another day...
 
Last edited by a moderator:
The Pioneer and Desert Wind were both rather mismanaged by Amtrak. As I recall, the biggest problem with the Pioneer was awful arrival/departure times at the main destinations along the route, while the Desert Wind could have done a lot better if more space had been available from LA to Vegas (I recall seeing on here that there were regular space lockouts on that segment like there are west of Reno on the Zephyr). If you give a train an atrocious timetable, don't expect good performance. Same if you under-equip a train, though that was at least more excusable.
The restoration study on the Pioneer showed it having only a hundred thousand passengers and direct losses of $25-35 million. Most of us would call that a complete waste of money. As for the Desert Wind, unless you managed to significantly increase its speed, it wouldn't garner anything more than marginal ridership Los Angeles-Las Vegas (going from 5.5 hours to 6.5 hours in the corridor study resulted in a 36% ridership drop).

The Sunset East was screwed by being 3x daily...much as the Cardinal and Sunset West have been. I also don't think that too many of us put much faith in the Sunset East report that was produced a few years back.
Sunset East was screwed by going over terrible track, having terrible times, being terribly slow, and having absolutely atrocious on time performance.

And of course, there's the fact that those trains were all axed in an era when LD ridership was somewhere around 1/4 lower than it is now, and ridership overall was over 1/3 lower. Perhaps the best way of underlining this is the fact that in FY04 (the last whole year the train ran through), ridership on the whole Sunset was 96,426. In FY14, ridership was 105,041 (which beat out FY03 by 8 riders for the best ridership on my chart). Just to be clear, this isn't a case of the train being cut back and improving as a result (something which actually makes some sense on that route); this is a function of overall ridership growth in the system.
Add 25% and...you still have a terrible performance. Per Amtrak, Sunset East ridership dropped from 53K to 29K between 2000 and 2005. Even adding 25% to it doesn’t make for a break even extension in a best case scenario. Adding millions to Amtrak’s deficit, plus tens of millions in capital expenses (and where are you going to get the equipment from nowadays anyhow?), isn’t a sound idea, especially when there’s no possible way that you can claim with a straight face that the social benefits exceed the cost of service provision.

With respect to "keeping routes short and frequent" for that reason, in the best of all worlds for your case you're doing the right thing for the wrong reason. I strongly suspect that you're doing the wrong thing for the wrong reason. Making them frequent, I agree with; keeping them short, however, I disagree with since you both restrict network expansion and drop links between regions in the process. I believe that it is more likely than not that in a decade there will be a robust, growing network in Florida run by FEC and at least some form of SEHSR approaching completion...and no ability to improve links between them. The same can be said of the Midwest and the Northeast. I can dig out a good chart that explains the logic here, but suffice it to say that there's a reason the LSL does so well (and as much as we can dicker over Neroden's calculations, I suspect we can agree that the "fully allocated" numbers aren't useful in this respect).
If you don't keep them short, you greatly increase the expense of multiple frequencies, both in terms of crew expense and equipment. The twice daily Silvers require as much equipment as the hourly FEC for example. You also need to start introducing mid-corridor starts (as the San Joaquins are doing) to have good timings for intermediate cities (and intermediates going to terminal) unless you want to run empty trains half the day. The network linkages just aren't overly meaningful because your market will be terrible and the costs will increase faster than revenue.

On the topic of allocated costs: I think they're broadly accurate, though they certainly aren't the marginal costs (nor are they meant to be understood as such). Railroading has been somewhat perennially plagued by the idea that variable costs are actually fixed.

Moreover, CHI-MSP is not a "long overnight train", but WI has been playing merry hell with any efforts here, while IA's internal issues forcibly truncated a new service at the Quad Cities. Just because a train doesn't run on and on and on doesn't mean it isn't susceptible to this sort of nonsense.
How does Wisconsin prevent Illinois and Minnesota from paying for the necessary track improvements and equipment? As far as I'm aware, the issue is that they don't want to pony up for a rail line which is entirely their prerogative. As for Iowa, while I think they're mistaken in their rejection of the train, I don't believe that some sort of Federal overruling of them is appropriate.

As to the Eastern states' situation, there is a very valid reason that the states would have had a problem there without Amtrak's involvement, which I've discussed at great length with people both at DRPT and VHSR. With the exception of New York (the Empire Service/Adirondack/Ethan Allen being effectively off the NEC...and that mix is probably not big enough to support a full new car order on its own), there is a great deal of equipment interchanging among the states' routes and Amtrak's routes. For example, a Keystone set can be put on the back of a BC-QC-Coach-Cafe set to make a Regional, and the Carolinian, VA Regionals, and NE Regionals all interchange very straightforwardly. The only true oddballs are the Vermonter (a 5-car set IIRC) and the Pennsylvanian (which still does an interchange, this time with the Palmetto, likely dating back to the Three Rivers/Silver Palm situation).
Empire Service is, what, 61 cars or so? That’s certainly enough to support an order, especially if they partner with somebody else. As for the interchanging: Switch around the cars and credit or charge appropriately as need be. Though I’m a bit doubtful that it’s all that common of an occurrence. But I think you missed my main point which was that there was no planning done for new cars, despite their necessity (something I, quite frankly, find a bit odd given Virginia’s service expansion plans). Meanwhile California had plans and specifications already drawn up (“Surfliner III”) which allowed them to take advantage of the opportunity for Federal funds when it showed up. And sure, involve Amtrak in it. But as far as I can tell the Eastern states weren’t doing a thing and hence why they get to pay for old equipment in the operating budgets.

There were *much* higher priorities for New York -- buying the Empire Corridor as far as Schenectady, improving the tracks, and rectifying the substandard stations which dated to the NY Central's "destroy all passenger stations" era. California, Washington, and NC did all that station stuff a long time ago, and the stations were never as dismal as the NYC/PC shacks. Similarly, Connecticut's priority was rehabilitating the Hartford Line tracks & stations -- they are planning to buy rolling stock, but the tracks & stations come first. Pennsylvania's focus was, again, on the substandard station conditions and deferred maintenance on the SEPTA lines. You get the picture.

New York got burned once by buying rolling stock before buying track improvements ("Turboliners" may ring a bell).
Higher priorities don’t mean completely ignoring it. Plans and specifications can be created pending future opportunities.

I don't think the Pioneer or Desert Wind ever had much potential. Lack of intermediate population.

Sunset East is another matter; the advocates have presented solid data showing that 41% of the revenue of the train came from the eastern segment, which is less than 28% of the route-miles (on the version which went all the way to Miami). Arguably the western half should have been cancelled instead, leaving a daily train from San Antonio to Miami. Anyway, a topic for another day...
Where the track was cut was the deciding factor.
 
I am surprised to see folks fall for the fallacy of train by train analysis when they also seemingly believe in the creation of an overall viable LD network. The question to ask should be whether a Pioneer and a Desert Wind together with the rest of the network produces an overall better network or not IMHO. But I realize I may be in a hopeless minority.
 
The restoration study on the Pioneer showed it having only a hundred thousand passengers and direct losses of $25-35 million. Most of us would call that a complete waste of money.
The Pioneer study was seriously flawed. Please see the discussion and response here.
Honestly, I think that response is more flawed (and not just because they forgot to attach the timetables that they refer to in it). It's hand wavy, wants to ignore costs, not exactly well sourced, disingenuous about population increases, and so forth. I see absolutely no reason to credit their claim that ridership should be in the realm of 160-200K. Though that said, even granting them the benefit of the doubt and assuming it were accurate, it would still not make it a good route to restore. Seriously, the most cheerfully optimistic projection that they can come up with has it underperforming every single daily long distance train despite a two night journey. This particular statement, I think, dooms the entire thing:

One coach car - an Amfleet I coach, with its existing seat configuration - would provide budget transportation for persons of limited means who would take a bus if it weren't for the fact that the bus service is no longer available
If bus service cannot be profitably sustained along a corridor, there is no potential for rail service.
 
I would disagree with that generalization, Paulus. Even with the free roads, bus service is essentially unprofitable in a lot of corridors -- because unless there are bus lanes, it's always slower than driving, which makes it a very hard sell. In most places, it's really a choice of government-sponsored buses or government-sponsored trains, not "private profitable" buses, which are a mirage.

I am surprised to see folks fall for the fallacy of train by train analysis when they also seemingly believe in the creation of an overall viable LD network. The question to ask should be whether a Pioneer and a Desert Wind together with the rest of the network produces an overall better network or not IMHO. But I realize I may be in a hopeless minority.
That's not quite the right question. The right question is: what expansion would *most improve* the overall network? And I can think of several dozen which are better than the Pioneer or the Desert Wind and better than both put together. Basically anything you could come up with east of the Mississippi is more useful for the overall network. Sunset East is better, for example. So is the Broadway Limited. So is the Floridian. So is the daily Cardinal. So is the National Limited. So is the Inter-American....
Heck, so is the North Coast Hiawatha.
 
Last edited by a moderator:
I am surprised to see folks fall for the fallacy of train by train analysis when they also seemingly believe in the creation of an overall viable LD network. The question to ask should be whether a Pioneer and a Desert Wind together with the rest of the network produces an overall better network or not IMHO. But I realize I may be in a hopeless minority.
The problem is actually estimating those effects...especially when a bunch of trains are already running slam full a good share of the time. To back this out a bit, let's say that we were to add a new train SEA-SLC or SEA-DEN and bump up the capacity of the Zephyr the rest of the way into Chicago. Well, that should generate some through traffic onto the Cap and LSL...but the Cap and LSL are pretty darned full. So you expand one (or both) of them...and if the Cap gets expanded you're trying to spill even more traffic onto the Meteor (and possibly eventually the Star)...and it's not like the Meteor has lots and lots of space to spare (I'm frequently having trouble finding space on the Meteor WAS-RVR, which almost never used to be the case).

Although I often talk about making the network at least self-sustaining in terms of operating costs, I never see it as a case of Amtrak only running trains that make money. Rather, I see it as a case where the LD trains should be at least partly subsidized by routes that go well above their operating cost in terms of cost recovery. For example, the trains along the NEC make money as a whole (let's save the dispute about capital cost allocations for another day...the Regionals and Acelas clearly cover a lot more than their direct operating costs and if they're also shouldering a share of overhead in their calculations the numbers simply get more massive than the MPRs even indicate). On the existing formulas, the NEC is over $100m in the black for the first two months of FY15. Should all of that go elsewhere? Probably not; at least some share of it should be dedicated to NEC reinvestment. However, a share of it should also be reserved for offsetting other losses, if only on the general principle that the rest of the country has been covering (and will likely continue to cover) major capital investments in the NEC for a long time.

Back to the national system, part of the problem is that the system is exceedingly thin: There are only three general sections that you can argue have better than once-daily coverage (Chicago-NEC (17x weekly), Chicago-LAX (10x weekly), and NEC-Florida (14x weekly excluding the Auto Train), and in two of those cases the number is only that high because of various clunky routings (Chicago-NEC includes the Cardinal while Chicago-LAX includes the TE-SL connection) and completely disconnects a large number of major cities from even once-daily service (running down the list, of the top 30 cities in the US by population, 7 lack daily service (Houston, Phoenix, Columbus, El Paso, Nashville, Louisville, and Las Vegas) while several more have bad service (Detroit doesn't have anything running through to the east, Indianapolis is served at awful hours, and in general San Antonio has suffered from bad times on the Sunset Limited). The number with no more than a single train per day brings you to 14 of 30, with San Antonio and Fort Worth only avoiding the list by virtue of being at the intersection of two routes. Outside of that top 25, Cincinnati, Des Moines, Salt Lake City, Boise, Spokane, Cleveland, Toledo, Little Rock, Roanoke, Knoxville, and Chattanooga all either lack service or have it at bad times while it should be noted that Minneapolis-St. Paul, Tucson, Tampa-St. Petersburg, and Atlanta either have a single daily train that is swamped (traffic at ATL, TPA, and MSP is a disproportionate share of the traffic on their daily trains) or don't even have a daily train (Tucson). And so on...

The result of this is that you get a lot of trains having to, for example, connect in both Chicago and on the East Coast (or in Chicago and on the West Coast), which simply does not work well for a ton of reasons (I can offer an excellent story of being stuck with such a connection last year...driving STL-IND was not fun).

You also get no alternative to Chicago for a same-day connection east-west (nothing goes through St. Louis while New Orleans only connects with an overnight on one's own dime) and you get messy service into a number of cities so as to maintain those rather thin links. Amtrak also probably ends up kicking out at least $1m/yr on missed connections in Chicago in a good year, which is another line to consider in this, and when things went haywire this year on the Builder and Capitol Limited you basically saw system connectivity disintegrate.

-------------------------

I figure it's time to throw another point out there, in that I'm looking over the most recent Budget/Business Plan document and getting some interesting results. Per these numbers, the direct losses of the LD system come to about $120m/yr, dropping to $100m/yr over the next 5 years. Indirect/shared costs take another $500m/yr. Some of the indirect expenses do attach reasonably directly (stations come to mind) while some of those expenses simply don't belong there as part of any reasonable business calculation, either in part (policing, casualty/FELA, communication) or at all (data processing, probably adminstrative). Some of the expenses seem to have been just split evenly between the non-NEC business lines (MoW support...for FY14 you get $37.5m to LD trains, $38.3m to State trains, and $93.6m for the NEC). For example, I really don't see how you even end up offloading 20% of the MoW support to the LD trains...only 7 of them even touch Amtrak-owned tracks (one, the Cap, negligibly so...and I forget the actual ownership of the Empire route here) and none of them touch any significant part of NEC-North (19 r/t daily) while the remaining ones make up 5.5x daily round-trips out of 42 on the NEC-South (the Acela makes up about 16, the Regionals about 19, and you have two state trains as well)...which gets you to a whopping 9% of the route-miles on the NEC. Granted you have a few other Amtrak-owned segments out there, but given that one of the largest is in MI and that I believe NY is more or less billed straight for the Poughkeepsie-Schenectady line (of which the LSL makes up <10% of the traffic), I do not see how you get 22% of MoW allocation out of this. Doing the same for state trains is more complex since a lot of that can be put down to the VA Regionals in some fashion, not to mention the NY and MI situations.

So moving on from those, if we allow direct losses of $120m and allow for shared station expenses (since that seems like a fair allocation), overall losses are still around $170m/yr (again, dropping towards $150m/yr with no new equipment). That number rather strongly implies that, given the reputed money holes that a few trains are, a decent share of the LD system is either posting negligible losses or outright making money. Frankly, capacity improvements and equipment reallocation (Sparks Cars, for example) ought to be able to improve that number quite a bit (again, capacity jams all over the system come to mind), as would daily operation on the Cardinal and Sunset Limited.

Moreover, this does suggest that adding another LD train (or two) out west would increase the direct losses by relatively little. I think Amtrak came to the conclusion that reinstating the Desert Wind would increase the deficit by about $15m (this could probably be lowered significantly by adding cut-off cars in Las Vegas)...but on this basis even adding the Pioneer as a full-blown daily train CHI-SEA (i.e. a second frequency CHI-DEN on the existing route) would likely not hit Amtrak's bottom line for more than $20m or so, less if Amtrak could re-route at least one of the two trains over IAIS without a major loss of time (and even some loss there might be worth the hit to directly serve Des Moines). On the basis of the "fully allocated" losses projected for the North Coast Hiawatha and the above-mentioned allocation issues, the NCH would probably actually improve the direct loss situation (not to mention backing up the Builder in a very real way; if one melts down but not the other, equipment could be switched between trains).

Lest I seem to be playing down these numbers, I do believe there would be knock-on benefits for such expansions that would offset these losses, as well as a decent probability that if you carried out "sympathetic" expansions east of Chicago (i.e. doubling up the LSL, Cap, and/or Cardinal, and/or adding even more capacity to/from Florida) you'd improve things more. Moreover, I do see significant political gains to be had from pushing such an expansion (since there's a funny thing...I've almost never heard a politician fight for cutting their area's trains).

TL;DR: I do think there are major benefits to be had from expanding the National Network, either by "daily doubles" on some routes or from adding new ones. I also think that posting what by historical terms would be pretty modest losses on some routes (a serious expansion of the LD system by adding three new LD trains, doubling the LSL and Cap, and making the Card and SL daily would probably cost about $30m/yr in direct losses) would (A) be swallowed up in Amtrak's overall operating picture and (B) would yield reasonable political and practical benefits to offset those costs.
 
Last edited by a moderator:
The marginal or direct cost of a long distance train is something around $45 per train mile on average (interestingly, it's very close to the 60% of ICC fully allocated costs that either SP or SF ended up figuring as the marginal in the 1920-30s). A restored additional frequency Pioneer would be ~87 million in direct costs alone, plus hundreds of millions in capital expenses. In FY 14 the Empire Builder had $60.5 million in revenue, the Zephyr $55.8 million, and the Southwest Chief $49.4 million. To hit a loss of only $20 million you'll need to completely outperform every existing transcontinental train and while there will likely be some increased traffic Chicago-Denver for the Zephyr as well, there's going to be hardly anything generated between Denver-Seattle on the Pioneer. The Zephyr's ridership gains will also have to be balanced by revenue losses on the Zephyr and Builder from increased lower bucket fares due to higher room availability. With regards to the NCH: The study projected $39.1 million in direct losses from it, not allocated losses. That would make for a 33% increase in the LD direct loss category.

But, screw it, the operational cost is a distraction from the real issue of capital cost: Where are you getting all the money for these expansions? It's not like Amtrak's rolling in the dough here and every dollar spent on long distance train expansion is a dollar that could've been spent elsewhere, like the vastly more important Gateway Project. I do not believe you can make the argument that long distance expansion has a better return on investment, including social benefits, than does restoring the NEC to a good state of repair and expanding it.
 
Back to the national system, part of the problem is that the system is exceedingly thin: There are only three general sections that you can argue have better than once-daily coverage (Chicago-NEC (17x weekly), Chicago-LAX (10x weekly), and NEC-Florida (14x weekly excluding the Auto Train), and in two of those cases the number is only that high because of various clunky routings (Chicago-NEC includes the Cardinal while Chicago-LAX includes the TE-SL connection) and completely disconnects a large number of major cities from even once-daily service (running down the list, of the top 30 cities in the US by population, 7 lack daily service (Houston, Phoenix, Columbus, El Paso, Nashville, Louisville, and Las Vegas) while several more have bad service (Detroit doesn't have anything running through to the east, Indianapolis is served at awful hours, and in general San Antonio has suffered from bad times on the Sunset Limited). The number with no more than a single train per day brings you to 14 of 30, with San Antonio and Fort Worth only avoiding the list by virtue of being at the intersection of two routes. Outside of that top 25, Cincinnati, Des Moines, Salt Lake City, Boise, Spokane, Cleveland, Toledo, Little Rock, Roanoke, Knoxville, and Chattanooga all either lack service or have it at bad times while it should be noted that Minneapolis-St. Paul, Tucson, Tampa-St. Petersburg, and Atlanta either have a single daily train that is swamped (traffic at ATL, TPA, and MSP is a disproportionate share of the traffic on their daily trains) or don't even have a daily train (Tucson). And so on...
I cannot emphasize this point enough. Amtrak's mission is to run a national system. From here in the Northwest, it takes five days to get to the east coast. To me, that says that Amtrak is failing in its mission, thanks to poor support from our elected representatives. The 21st century cannot depend on increasingly-expensive publicly-subsidized highways and air travel, as we face crumbling infrastructure, dwindling fossil fuels, and climate change. Does this country want to return to the 19th century, when travel was a luxury for the rich? I don't think so. We need to support efficient, affordable transportation alternatives that will benefit all citizens, no matter where they live.
 
I figure it's time to throw another point out there, in that I'm looking over the most recent Budget/Business Plan document and getting some interesting results. Per these numbers, the direct losses of the LD system come to about $120m/yr, dropping to $100m/yr over the next 5 years. Indirect/shared costs take another $500m/yr. Some of the indirect expenses do attach reasonably directly (stations come to mind) while some of those expenses simply don't belong there as part of any reasonable business calculation, either in part (policing, casualty/FELA, communication) or at all (data processing, probably adminstrative). Some of the expenses seem to have been just split evenly between the non-NEC business lines (MoW support...for FY14 you get $37.5m to LD trains, $38.3m to State trains, and $93.6m for the NEC). For example, I really don't see how you even end up offloading 20% of the MoW support to the LD trains...only 7 of them even touch Amtrak-owned tracks (one, the Cap, negligibly so...and I forget the actual ownership of the Empire route here) and none of them touch any significant part of NEC-North (19 r/t daily) while the remaining ones make up 5.5x daily round-trips out of 42 on the NEC-South (the Acela makes up about 16, the Regionals about 19, and you have two state trains as well)...which gets you to a whopping 9% of the route-miles on the NEC. Granted you have a few other Amtrak-owned segments out there, but given that one of the largest is in MI and that I believe NY is more or less billed straight for the Poughkeepsie-Schenectady line (of which the LSL makes up <10% of the traffic), I do not see how you get 22% of MoW allocation out of this. Doing the same for state trains is more complex since a lot of that can be put down to the VA Regionals in some fashion, not to mention the NY and MI situations.
You've spotted what I spotted. :) The long-distance trains are getting unreasonable overhead allocation in several categories. A fair amount of the overhead really shouldn't be allocated to routes at all (like the data processing). From a business decisions perspective, it really only makes sense to look at the direct losses.

It's worth noting that the 2014 budget doesn't match up with the September 2014 results (they're better than budget). The budget predicted $1227.6 million in "direct and shared costs" for the long-distance division, while the September 2014 results show $1093.7 million. The budget predicted $507.0 million in ticket revenue; actual result was $510.7 million. I can't tell what the variance is on F&B revenue. (I will tell you right now that F&B revenue will be dropping in 2015 due to idiotic food service cuts; however, for the sleeper passengers, this will simply translate into higher ticket revenue.)

The breakdown of the direct losses is something I've been looking at for a while, ever since Boardman put that bar graph in his presentation. While it's hard to get exact numbers for any year except 2012 (sigh), it really looks like the following is true:

-- all the Florida trains are profitable

-- the Lake Shore Limited is roughly break-even

-- most of the other trains have operational subsidies of less than $10 million / year

-- The California Zephyr, Southwest Chief, and Sunset Limited require the bulk of the operational subsidy -- well over half of the entire long-distance division's subsidy in 2012. My calculations indicate that the percentage is even larger now.

I do think there is value in maintaining a connection between the east-of-the-Rockies network to the West Coast. The Empire Builder is the best connection financially. I think there's value in maintaining a second connection further south -- but it costs a *lot*. Maintaining three of them is really quite expensive. Expanding this type of service (with trains like the Desert Wind / Pioneer) should not be considered until service is expanded in the much more promising areas to the east.

My initial-action proposal would be the following:

-- daily Cardinal

-- run the Broadway Limited, probably on the "other schedule" from the LSL (evening departure from NY, morning arrival in NY, midday at Chicago)

-- add Wolverine runs which continue from Dearborn to Toledo to connect with the LSL (first priority), the Broadway Limited (second priority), and the CL

-- run a Denver Zephyr on the "other schedule" from the CZ (evening departure from Chicago, morning arrival). Extend up the Front Range to Fort Collins if & when BNSF & equipment allow.

-- run a day train to Minneapolis, and move the Empire Builder onto a night schedule from MSP-CHI (yes, I know, cascading effects all down the line)

-- run the Silver Palm to Florida. It may not be obvious how much connection value this has; this would have saved my bacon on a recent cancelled trip (I planned to get into Florida the night before, and the next run was 24 hours later and missed the event -- if the next run was *12* hours later I would have made it). It provides morning arrivals in Miami and evening departures.

I think most of this would actually be profitable (good for Amtrak's bottom line).

My current estimates say that making the Cardinal daily would be *profitable*, and would in fact swing the train from requiring a subsidy to generating money to cover overhead -- I'm not accounting for any ransom demands from CSX, of course, so it might not be that good, but it would be good.

Other major network improvements would probably require more money, probably mostly from states:

-- Cleveland/Columbus/Cincinnati

-- Iowa Interstate route across Iowa (CZ and Denver Zephyr should be rerouted here upon completion)

-- Daily Sunset/Eagle plan...

-- ... and Phoenix line restoration

-- SWC over the Transcon through Amarillo & Wichita, + Heartland Flyer extension to Wichita

-- Dallas/Houston service restoration

-- Mobile-New Orleans corridor service

-- Atlanta-Montgomery-Mobile-New Orleans section of the Crescent

-- Jacksonville-Atlanta train of some variety

-- some sort of service to Louisville & Nashville. Chicago-Indianapolis-Louisville-Nashville-Chattanooga-Atlanta-Jacksonville-Orlando would make sense if CSX would cooperate.

Yeah, I'm leaving out Las Vegas. The casinos can build their own high-speed line to Los Angeles and they probably will.

Florida needs a serious service improvement and restructuring but the core element it needs is the cancelled Tampa-Orlando high speed line, so there's not much point in trying to straighten things out there until that gets built.

At the moment, the state payment requirements are making things hard because some states are run by anti-rail nuts. However, the Class Is, those ungrateful recipients of corporate welfare, remain a larger obstruction; the way around this seems to be through the states purchasing ROW, which means that there was no real way around the state governments anyway.

Here's hoping that Hoosier State support starts to translate into a rail-curious Indiana state government. Next step is to somehow get the anti-rail nuts out of power in Ohio, and then we can start making real progress.
 
It's not like Amtrak's rolling in the dough here and every dollar spent on long distance train expansion is a dollar that could've been spent elsewhere, like the vastly more important Gateway Project.
Bluntly, Gateway's largely a subsidy for New Jersey. Whose deranged governor Christie decided not to spend money on the tunnels, and to spend it on the Pulaski Skyway instead. As a New Yorker, I have zero sympathy for New Jersey at this point. If Amtrak can get NJ money (or NJ-Senator-earmarked federal money) for the Gateway tunnels, go for it. Otherwise, Amtrak should spend money on things which benefit Amtrak.

Two tunnels, or even one, are plenty for Amtrak; it's just that without Gateway, New Jersey Transit would have to shut down. So sorry, folks who benefit from New York City salaries but refuse to pay taxes to New York -- not.

In the meantime, I'd rather have enhanced service between states which bothered to spend their own money, like New York and Illinois. Better value for money than Gateway would have been doubling the Empire Corridor tunnels, rather than enhancing service to the Unappreciative Grifter State.

Apologies to all decent people who live in New Jersey; I know you're a significant minority. But you don't seem capable of getting control of the state government. And you don't even have the excuses we do in New York, which are malapportionment and elected officials switching parties due to bribes.

I do not believe you can make the argument that long distance expansion has a better return on investment, including social benefits, than does restoring the NEC to a good state of repair and expanding it.
It's actually an easy argument to make -- as long as "long distance expansion" means Michigan, Indiana, Ohio, Florida, Pennsylvania, Georgia, etc. -- not Idaho, Wyoming, and Nevada.
Which is actually an argument I keep having, because there are, strangely, people who somehow think that Idaho, Wyoming, and Nevada service should be prioritized over Ohio and Indiana service. And I can't see *any* argument for that.
 
Last edited by a moderator:
"nothing goes through St. Louis"

As several routes should. Which I won't spend a bunch of time arguing about - I've advocated St Louis being redeveloped brought back to it's previous rail glory to alleviate pressures on Chicago and from a common sense standpoint many times. Unfortunately everyone just sees negative and cost of creating this reality. Not overall benefits when completed. Not to mention bringing jobs and vitality back to a city that needs it.
 
One thing to consider with respect to the California Zephyr in particular: At A-Day, there was no daily service SLC-RNO. You had a 3x weekly run as part of the City of Everywhere, but the City of Los Angeles (i.e. Desert Wind) and City of Portland (i.e. Pioneer) ran daily, not the City of San Francisco. Denver-Salt Lake still effectively had 10x weekly trains (3x weekly on the Rio Grande and daily on UP) while I believe that Chicago-Denver was still up at 3-4x daily.

Additionally, I don't have load information on the train, but I believe that until the Builder meltdown ridership was exceedingly thin on that section of the Zephyr. This raises a serious (if awkward) question as to whether it wouldn't make sense to split the Zephyr at SLC and run a section to Seattle and a section to Los Angeles and no section to San Francisco. I'm not a fan of this option (not least because it creates a mess for Reno-San Francisco), but other alternatives (such as running a stub connection with a sleeper, coach-cafe, and coach SLC-RNO and adding several coaches and some improved food service RNO-EMY) also come to mind here. The Zephyr is very likely the biggest hole in Amtrak's bottom line out there, and most of that hole is between Reno and Salt Lake City.

As to those costs...as I said, a few lines do need to be allocated (station costs come to mind...since those are direct-but-divided, if you will), but with the exception of adding new routes those costs are close to flat no matter what you do (since in a lot of cases, adding a train or two to a given line would only incrementally add a few hours to a given station's operating times while many of the station maintenance expenses are fixed).

My initial action in the East would be as follows:
-Make the Capitol Limited twice-daily, with an agreement with PA to roll up the Pennsylvanian as part of the National System if the state properly funds a second frequency.

-Extend a second train from New York to Chicago on the LSL's route; I'd look at using the Maple Leaf for this.

-Make the Cardinal daily.

-Add a Cardinal section to Kansas City via St. Louis.

-Extend the Silver Palm.

-Work out a deal with AAF/FEC to get Silver sections operating down FEC to Miami. Include an informal proviso that if FEC extends to Tampa, Amtrak would be interested in restoring a joint crew base and station in Tampa and not competing with FEC on the Tampa-Miami markets (where Amtrak is going to get killed anyway).

Out West:

-Run a Denver Zephyr, but I would extend the train all the way to Seattle via Boise. Depending on the scheduling, it might nominally be two separate trains with a large hold at Denver (but with no change of train at Denver).

-Run a section of either the CZ or DZ/Pioneer to LA via Las Vegas.

-Make the Sunset daily.

I include the Las Vegas project because at least until they get a bullet train funded I think that's a necessary line. Even once one is funded, I still think you'd have a decent train there, especially if you could get the two stations to be close to one another (since the bullet train would inherently not be running into the eastern end of the LA basin...and the bullet train would probably help out Amtrak ridership north of Vegas in any case).

I will say that I see expansion on both sides of things as necessary. However, expansion east of the Rockies should help at least eat up some of the overhead from any western expansions and contribute to the overall situation. Additionally, as I've said, I see there as being a political need (which for Amtrak is a practical need) to make sure that the West isn't left out of any service expansions.

-------------------------

One quick thought on NJ: There are a lot of reasons for folks not wanting to live in NYC, and it's not just the taxes.
 
I cannot emphasize this point enough. Amtrak's mission is to run a national system. From here in the Northwest, it takes five days to get to the east coast. To me, that says that Amtrak is failing in its mission, thanks to poor support from our elected representatives. The 21st century cannot depend on increasingly-expensive publicly-subsidized highways and air travel, as we face crumbling infrastructure, dwindling fossil fuels, and climate change. Does this country want to return to the 19th century, when travel was a luxury for the rich? I don't think so. We need to support efficient, affordable transportation alternatives that will benefit all citizens, no matter where they live.
If air travel is somehow not possible for long distance travel, then travel will of necessity be a luxury for the rich. Amtrak simply is not an efficient or affordable transportation alternative for such long distances as across the entire continent.

Bluntly, Gateway's largely a subsidy for New Jersey. Whose deranged governor Christie decided not to spend money on the tunnels, and to spend it on the Pulaski Skyway instead. As a New Yorker, I have zero sympathy for New Jersey at this point. If Amtrak can get NJ money (or NJ-Senator-earmarked federal money) for the Gateway tunnels, go for it. Otherwise, Amtrak should spend money on things which benefit Amtrak.

Two tunnels, or even one, are plenty for Amtrak; it's just that without Gateway, New Jersey Transit would have to shut down. So sorry, folks who benefit from New York City salaries but refuse to pay taxes to New York -- not.
Eh, there were issues with the ARC tunnel design. Personally I'd favor charging hefty track fees to NJT for tunnel use if they don't pony up for Gateway. As for the salary issue: New York collects income tax regardless of where you live (and used to have a commuter tax for NYC), so I'd not consider that a big issue.

I do not believe you can make the argument that long distance expansion has a better return on investment, including social benefits, than does restoring the NEC to a good state of repair and expanding it.
It's actually an easy argument to make -- as long as "long distance expansion" means Michigan, Indiana, Ohio, Florida, Pennsylvania, Georgia, etc. -- not Idaho, Wyoming, and Nevada.
Which is actually an argument I keep having, because there are, strangely, people who somehow think that Idaho, Wyoming, and Nevada service should be prioritized over Ohio and Indiana service. And I can't see *any* argument for that.
NEC is probably still the best place for return, but I've no problem with adding long distance Palmetto-style coach and business trains out east.

Additionally, I don't have load information on the train, but I believe that until the Builder meltdown ridership was exceedingly thin on that section of the Zephyr. This raises a serious (if awkward) question as to whether it wouldn't make sense to split the Zephyr at SLC and run a section to Seattle and a section to Los Angeles and no section to San Francisco. I'm not a fan of this option (not least because it creates a mess for Reno-San Francisco), but other alternatives (such as running a stub connection with a sleeper, coach-cafe, and coach SLC-RNO and adding several coaches and some improved food service RNO-EMY) also come to mind here. The Zephyr is very likely the biggest hole in Amtrak's bottom line out there, and most of that hole is between Reno and Salt Lake City.
Neither of those sections make any sense however. There's pretty much nothing to get ridership from for the Pioneer section and the Desert Wind section is tremendously uncompetitive. The two of them combined only had 260K riders in FY1994 when running daily and I'd not be surprised if a good chunk of that was never even on the relevant section (similar to how you can book a Los Angeles-Maricopa ride on the Texas Eagle).

-Make the Capitol Limited twice-daily, with an agreement with PA to roll up the Pennsylvanian as part of the National System if the state properly funds a second frequency.
I'm not sure what this is supposed to mean?

-Extend the Silver Palm.
Direct costs of $8.4 million just for extending the Palmetto, but if you're restoring sleeper service to it as well that will jump to $25.6 million (based on the difference between the Silvers and the Palmetto for direct costs). Sleeper revenue on the Silver Service only amounted to $19.6 million in FY14.

-Work out a deal with AAF/FEC to get Silver sections operating down FEC to Miami. Include an informal proviso that if FEC extends to Tampa, Amtrak would be interested in restoring a joint crew base and station in Tampa and not competing with FEC on the Tampa-Miami markets (where Amtrak is going to get killed anyway).
Honestly, I don't see how the FEC has anything to gain with allowing Silvers down it.

I include the Las Vegas project because at least until they get a bullet train funded I think that's a necessary line. Even once one is funded, I still think you'd have a decent train there, especially if you could get the two stations to be close to one another (since the bullet train would inherently not be running into the eastern end of the LA basin...and the bullet train would probably help out Amtrak ridership north of Vegas in any case).

I will say that I see expansion on both sides of things as necessary. However, expansion east of the Rockies should help at least eat up some of the overhead from any western expansions and contribute to the overall situation. Additionally, as I've said, I see there as being a political need (which for Amtrak is a practical need) to make sure that the West isn't left out of any service expansions.
Bullet train is never going to get built unless Nevada decides it wants to put their own public funds into it (doubtful); geography actually makes an HSR line LA-LV take too long to be a good proposition. And I'm not seeing how a single daily train LA-SLC via Las Vegas is somehow a necessary line. It's not going to do a single solitary thing for road or air traffic and it doesn't have the typical excuse of serving rural folks. Barstow, Victorville, and San Bernardino are already served (San Bernardino with a far superior Metrolink line). Caliente and Milford are both so tiny that there's no reason to actually stop a train at them. Might as well have the Surfliners stop at my old high school: the track runs right next to it and it's got more people than either of those cities combined. Seriously, on what grounds is it necessary?
 
Status
Not open for further replies.
Back
Top