Will Americans ever take sleepers again?

[Anderson]

How is that level of penetration "not unrealistic" given the average one-way airfares?

DC-Orlando: $149

Philadelphia-Orlando: $171

New York-Orlando: $175

Chicago-Orlando: $188

Albany-Orlando: $204

Boston-Orlando: $145

All numbers Q3 2013

I'm going to submit a couple of reasons:

(1) If I'm not mistaken, those fares are generally exclusive of non-tax fees (i.e. baggage fees). In most of those cases, you're adding somewhere in the range of $20-30 for baggage fees per person. Only Southwest doesn't charge for bag fees at this point (though their fares seem a bit higher overall), so I'd universally add baggage fees to the calculation. Southwest also wins an award for creative connections between New York and Florida (connections available at Atlanta, Nashville, MIdway, and Houston).

(2) PPR for sleepers on the Star is $241.32. For the Meteor it is $269.88. Both figures FY13. PPR in coach is substantially lower in both cases, but I'm going to consider those numbers to be utterly useless given that those include a lot of short-hop tickets (the Star being particularly bad in this respect, what with the south-of-Orlando turnover providing a big slug of traffic for that train). For the record, the Meteor's PPR was $106.01 for FY13 overall; I would nudge that up a bit for through traffic on the basis of actual ticket prices and the aforementioned traffic situation.

(3) That level of penetration assumes sleeper penetration of somewhere under 2% for the region (2.5% on core airports, 0.5% on the surrounding areas). I'm still assuming that the majority of people travel coach on those numbers. (Edit: I also assume that at least some share of that comes from driving...and north of DC, driving to Florida is not a one-day trip).

(4) An implicit assumption is quite frankly that Amtrak would be able to scale a number of things on multiple trains. Amtrak did 3 round-trips to Miami with 10 sets of equipment versus 2 trips with 8 sets now. I do figure that you'd have similar scaling in effect (probably around 20-22 sets with 6 trains, depending on scheduling) as well as at least some efficiency in crew shuffling. Some other things would scale, additional station costs would be near $0, and as has been discussed at length elsewhere the Florida trains are already at break-even on direct costs. In such a situation, I'm guessing that fares would be about what they are now or less.

(5) Finally, another implicit assumption traces back to the travel situation: The vast majority of these trips are, if I am not mistaken, leisure and not business (so an exact arrival time into Florida is not a huge deal), so taking the train does in fact save a good portion of a day. Let me run down the list using January 15th 2015 in all cases for the SB leg, using January 21 2015 for a return:

-Southwest's earliest available arrival from EWR is at 1350, from LGA at 1305. ISP has a lone direct flight

that gets in before noon...which is also already $425 almost two months out.

-United has two flights that get you there before mid-afternoon. One departs EWR at about 6 AM and gets you to MCO at 1030 (price $368 before any fees). The other is 0735 to 1034 (and is, in fact, direct). Price: $323 before fees (though I cannot even begin to guess what the fare pricing).

-Delta is in the same boat (pricing starts at $330 for the most part, though there's a stray $295 thrown in). You've got multiple morning arrivals, though a lot of those are mix-and-match connection combos involving departures from different NYC-area airports connecting to the same flight in Atlanta.

More in a moment; I'm worried my browser will die on me.

Edit:

Ok, Nathanael, I need to make a correction: In that post when I said that domestic FC was on par with Amtrak's long-distance coach, I was incorrect. Domestic FC is on par with Regional coach. That's my bad.

As to California, there is a city pair in the sleeper sweet spot there: Los Angeles-San Francisco. Northbound travel time is 10:01; a 100-series local takes 93 minutes from there to 4th and King while a 300-series Baby Bullet takes 59 minutes. Assuming a 10-minute stop at SJC and a Baby Bullet schedule you'd be looking at 11:10 end-to-end from LAUS to 4th and King, which would enable such options as a 2000 departure and an arrival at 0710 (probably timetabled at 0730 or so for padding purposes, with the train running discharge-only from SJC onwards). Los Angeles-Phoenix and Los Angeles-Las Vegas are both pretty close to it as well (both falling a bit short but close enough to make work with a bit of a schedule massage...and Las Vegas might well be the single most flexible city in terms of scheduling arrivals and departures in the country, if I had to guess).

Edit2: There's always something else, isn't there?

http://finance.yahoo.com/news/9-airports-where-flying-lot-141839938.html

The article mentions a stack of airports facing various issues with crowding; MCO, EWR, BWI, and BOS are all on the list and relevant to my Florida discussion.

 

[neroden]

As to California, there is a city pair in the sleeper sweet spot there: Los Angeles-San Francisco.

Major problem there: the speed is not driving-competitive. By a substantial margin, not just an hour or two.
Second problem: that's just two cities. In the East we can get whole strings of cities, which gives better ridership.

Los Angeles-Phoenix and Los Angeles-Las Vegas are both pretty close to it as well (both falling a bit short

That's exactly the thing -- they're actually too close for sleeper service, and should have faster-than-driving daytime service instead.
If you have driving-competitive runtimes, *San Francisco to Las Vegas* is about right for sleeping car service. However, once CAHSR gets built, even that will be too short for sleeper service. And there's no "next city" beyond Las Vegas.

Again, with driving-competitive runtimes, Reno to Salt Lake City should be about right for sleeping car service. (The westbound scheduling of the California Zephyr is OK, the eastbound scheduling is no good.) But it just isn't a big enough market. And 8-12 hours out of Emeryville lands you in Winnemucca.

With driving-competitive runtimes (presumably along the Wyoming route) Salt Lake to Denver would again be a bit *short* for sleeper service, though it should work.

Here in the east, whatever fast daytime service you may run, you can just push the route another few hours out, and find yourself at another decent-sized city where the timing is right for sleeping cars. On the west end of the "east", you can push this process as far as Minneapolis, Omaha, Kansas City, OKC, Dallas, and San Antonio, probably, and possibly stretch to Denver. On the east end, you can go as far as the Atlantic Ocean. On the north, as far as Boston, and on the south, all the way to the Gulf.

This just doesn't work out west; you run into these empty deserts and mountains and grazing land. And wine country, in the case of northern California. And the Mexican border as well. It's much harder to hit the sweet spot for sleepers because there's so much empty space.

Of course, discussion of driving-competitive runtimes brings up another issue: the highways are fast out west, *particularly* in the empty areas. The state speed limit in New York is 65 mph, and that only applies to rural expressways; everything else is 55. Even in states with higher speed limits, traffic means that *speeds* aren't terribly high on I-95 parallel to the NEC, or anywhere in the Chicago metro area. Slower cars means it is easier for trains to match or beat cars on speed.

Looking at the speed limits by state...

http://www.motorists.org/speed-limits/state-chart

...I find that, off the NEC (speed limits are relatively slow in nearly all NEC states) New Hampshire, Oregon, Washington, Arkansas, Missouri, Kentucky, and Wisconsin look "likely" for trains to have speeds competitive with cars. (Of those, Arkansas and Kentucky were unexpected. Sadly, the direct line from Little Rock to Memphis seems to have been severed, but improvements to the Texas Eagle route through Arkansas might be highly successful. Kentucky suffers from very twisty railroad routes, though the Lexington-Cincinnati route looks OK.)

There's always something else, isn't there?

http://finance.yahoo.com/news/9-airports-where-flying-lot-141839938.html

The article mentions a stack of airports facing various issues with crowding; MCO, EWR, BWI, and BOS are all on the list and relevant to my Florida discussion.

I note Chicago Midway being right up on the top of that list. Expect Midway flights to become more expensive as Midway hits its people capacity limit.

 

[jis]

Let me start by saying that I like the analysis. However I have a few observations to share, so please take them as such.

(5) Finally, another implicit assumption traces back to the travel situation: The vast majority of these trips are, if I am not mistaken, leisure and not business (so an exact arrival time into Florida is not a huge deal), so taking the train does in fact save a good portion of a day. Let me run down the list using January 15th 2015 in all cases for the SB leg, using January 21 2015 for a return:

-Southwest's earliest available arrival from EWR is at 1350, from LGA at 1305. ISP has a lone direct flight

that gets in before noon...which is also already $425 almost two months out.

-United has two flights that get you there before mid-afternoon. One departs EWR at about 6 AM and gets you to MCO at 1030 (price $368 before any fees). The other is 0735 to 1034 (and is, in fact, direct). Price: $323 before fees (though I cannot even begin to guess what the fare pricing).

-Delta is in the same boat (pricing starts at $330 for the most part, though there's a stray $295 thrown in). You've got multiple morning arrivals, though a lot of those are mix-and-match connection combos involving departures from different NYC-area airports connecting to the same flight in Atlanta.

The United fares are round trip, not one way. The one way fare on the 7:35am is something like $217. Actually the air fares vary so wildly that it is hard to nail anything down precisely. For example, I am going on a weekend trip to EWR from MCO first weekend in December (Friday evening out Sunday late afternoon back) and my round trip fare is $178 on United.
As for baggage fees I reckon somewhere between half and two thirds of the travelers land up paying any fee at all on United. The rest have $0 fee for various reasons.

The cruise ship crowd usually uses evening flights to get into Florida well in time for mid morning departure on the cruise the next morning. I have noticed that there are very few leisure travelers on early morning flights, simply because they are the most expensive flights, as are early evening flights. If arrival time truly does not matter I'd expect people to choose the cheapest flights, not the most expensive ones.

I happen to travel quite frequently on this route as many times as twice a month round trips at various times of day and week,

I have some anecdotal evidence that you may be over-estimating typical prices paid and also baggage fees, though I have no idea how one could get a good guesstimate. I have noticed large families going on vacation to Disney who land up paying basically no baggage fees on mountains of baggage, because one or more members of the group either have an appropriate affinity credit card or have status enough to make it free. And then of course there are others who pay through their nose. But it is never a sudden surprise, since while buying the ticket and checking in on line they will be happy to tell you exactly what the baggage fees will be. I don't see too many people eating much on those flights and all soft drinks are free, served at least twice in course of the slightly to well under 3 hours flight.

As for airport congestion, on the rail side one needs to take into consideration the congestion situation on the Atlantic Coast route, which is not in a very pretty shape even right now. And like everything else, things will only get worse, not better. Just like selective mitigation on the rail side, selective mitigation on the air side is starting to happen with upsizing of equipment, and rebanking to reduce the mad dash to have every flight leave at the top of the hour. Also, in general there is much more committed resources for new terminals and related infrastructure at the air side than on the rail side. So I'd take that aspect out of the equation for now since I think it is a wash between air and rail as far as that goes.

But, other than those caveats, i do like the direction your analysis is taking. Frankly irrespective of the fare differences that you and Paulus are arguing about, the percentage is in the ballpark IMHO within +/- 0.5% point

 

[Anderson]

Well, a couple of thoughts pop into my mind:
On speed limits: Those are maximum speed limits, and are often not found anywhere close to "everywhere". In VA, for example, I-64 is only 70 MPH from Williamsburg up to about I-295, and then again west of Richmond to a bit short of Charlottesville. Most of the rest of the way it varies between 55 and 65. Moreover, in a lot of areas the speed limit is academic...you could set a speed limit to 90, but if traffic is jammed up it's not going to matter. This is often the case in many parts of Florida and Virginia (I think I can make it across Orlando about 25% of the time without getting caught in a massive slowdown). The same can increasingly be said for other areas as well (California comes to mind).

On airports: There are open question marks in a lot of cases. In the case of LaGuardia, Logan, National, and Midway the airport is exceedingly hemmed in (this is at least part of the reason that LaGuardia and National have their perimeter rules in place). If MWAA /really/ wanted to push traffic to Dulles, they'd just charge something to the effect of an extra $5/passenger at National and use that fee to cross-subsidize Dulles (there's a decent amount of traffic that is either going to come in at National or at Union Station; Dulles is sufficiently inconvenient to downtown DC to dissuade shorter-hop traffic from landing there).

Beyond those airports that are hemmed in, we've seen a mixed trend: Some airports are seeing overflow traffic, such as those mentioned in the article. Others are seeing either major drops in traffic or substantial reductions in the number of destinations being served. Basically, you're seeing a "winners and losers" situation slowly play out, especially as what seems like the last few rounds of airline consolidation play out (we're literally running out of airlines to see merge, and some of the ownership structures of international airlines are getting complicated due to various limitations on mergers). Some airports will invest in additional capacity, others won't. Moreover, some areas (such as Southern California) are running into issues with airspace capacity that is limiting the use of close-in secondary airports.

I will say that terminal congestion, airport congestion, and airspace congestion are different animals to some extent. You can de-congest a terminal with more space inside (adding seats, extending the terminal outwards some, etc.) allowing higher-capacity planes to be used. You can, in some cases, work on the airport with additional runways. Airspace congestion requires a lot of ATC work, and even it has limits.

On fares and fees: One thing I wouldn't rule out is some level of distortion. Those average fares do, in fact, include Business/First upgrades, and anyone buying a full-fare coach ticket is likely to increase the average fare paid on their flight by a dollar or two. As to baggage fees...I think the estimate of 1/3 of folks dodging those fees is fair. You've got a complicated mess of exemptions (some airlines seem to have universal exemptions for ID-carrying military for multiple checked bags, for example); though those won't be surprises, they still matter insofar as what one is "actually" paying. I'd also argue that some of the "discount" carriers are distorting average fares in the other direction by a bit given how much they seem to make off of various fees (if they knock $20 off their average fare but make it back through various fees then their average fare is going to look lower while they make the same amount).

====================================

As to those services, I guess the question is this: If a train can't cover a given distance in less than 6 hours or so, and driving is a similar story, then does it make sense to look at a sleeper service with early boarding or something similar?

For example, let's take Hampton Roads-New York City. Right now, that train trip takes around 8:00-8:30. Under the most aggressive (realistic) plans I can find, you'd be looking at maybe 7:00 for the trip (you're not going to knock much more off north of Washington; there's maybe 30 minutes to be saved RVR-WAS and another 15-20 you could save at WAS by using dual-mode...and possibly a bit more NFK/NPN-RVR, but getting the total savings over an hour is going to take a lot). Driving to New York is no bueno (I've seen parking fees at hotels in the $40-50 range) while going the other way, plenty of people simply don't own cars. Both ways, traffic tends towards being atrocious to the point that Williamsburg has complained about losing repeat visitors to jams in DC.

Yet a day train, leaving at 0500, isn't going to get to the other end of its run before noon under the most rosy plans. Leaving at 1700 it won't get to its destination before midnight. That is a very real problem that limits the train's utility due to a day lost in transit. There are other markets (Charlotte-Washington/New York comes to mind) where even if you /can/ get in under the sweet spot, it might make sense to run a train a bit slow to land in it given the amount of time lost on a day train.

 

[Paulus]

As to those services, I guess the question is this: If a train can't cover a given distance in less than 6 hours or so, and driving is a similar story, then does it make sense to look at a sleeper service with early boarding or something similar?

I don't think it necessarily does; trying to go overnight is trying to poorly compete with flying while those who've chosen to drive are doing so in the daylight hours (generally speaking), so there's no waste of time by running in the daylight. I think arrival time might be tripping you up rather than viewing it from the stance of most preferred departure time.

 

[northnorthwest]

Was out of town and am glad to see this topic still going strong a few days later!

I have a feeling that if a half dozen or dozen of the posters on this board could seize control of Amtrak it would be in much better shape....

I think the core of my question was really getting at the idea of whether the population could be taught (Please forgive my propagandistic language, but that's really what it would be.) to take trains again. My sense is that with advertising people can be led to buy just about anything, so there may be hope after all. In 1950 trains were part of the culture, and it was not shocking to imagine traveling by train. But today I just get the sense from the vast majority of people I personally meet (all across the country, many income and education levels, mostly in cities), that the train is just not on the radar. Before I relocated to my current home in Georgia I frequently took the Crescent to and from PHL, and I did so for reasons of scheduling, cost, and to a certain extent comfort/enjoyment/pleasure. I sometimes had the sleeper, always with AGR points, and I would have paid for it if I could have. (Current prices are prohibitive for me.)

But for my fiancée the train was and is never an option. Regardless of costs, the idea of a 15 hour train trip (Admittedly in my original scenario that time would have been cut down with various improvements, but the point is the same.) for her was just never ever an option. She has done LD train travel with me and loved it, but she said she would never do it alone. So I do get the sense that for most people it's either fly or drive, and I expect it is just some kind of perception that the train is not an option. It's just not part of their thinking.

But as I said, if the options were there and with proper investment in effective advertising (Think a beer ad, but on the train, so some hot 20-somethings board separately in coach, hit it off in the diner, and end up UPGRADING to a sleeper en route...TALK ABOUT AN UPGRADE....), maybe sleepers would have a chance....

Or perhaps I've just watched the film from which my name is taken too many times.

 

[Anderson]

As to those services, I guess the question is this: If a train can't cover a given distance in less than 6 hours or so, and driving is a similar story, then does it make sense to look at a sleeper service with early boarding or something similar?

I don't think it necessarily does; trying to go overnight is trying to poorly compete with flying while those who've chosen to drive are doing so in the daylight hours (generally speaking), so there's no waste of time by running in the daylight. I think arrival time might be tripping you up rather than viewing it from the stance of most preferred departure time.

Well, the issue is looking at both. For many purposes right now, and for the foreseeable future, you've got a choice of a bad departure time, a bad arrival time, or both: It's a fair question as to whether an 0400 departure is better or worse than one somewhere in the 2300-0000 range, but with the 0400 you're still losing half of your day (assuming you could get the arrival back to around noon) whereas with the late night departure you at least keep the day intact.*

Another point is that you really don't need a huge passenger base to support these trains. Assuming 22.5 passengers per Viewliner (I'm working with the eastern seaboard; I could run numbers for stuff on the west coast as well), a four-sleeper train would max out at around 90/day or 65,700/yr in the sleepers. Assuming that such a service doesn't need more than a cafe (possibly slightly enhanced somehow), operating costs are going to be generally in line with a regular train (plus sleeper crew costs). Don't worry, I agree, four sleepers would be a lot...I'd look towards two sleepers, 4-5 coaches, and a cafe. But let's play with some numbers.

The operating costs for the longest single-frequency daytime routes I can find are $28.2m (Palmetto), $20.3m (Carolinian), $16.5m (Pennsylvanian), $13.5m (Blue Water), and $12.4m (Adirondack). Kicking out the Palmetto (it's almost too long for our purposes), $15-20m seems like a reasonable cost estimate for a train running this length...and based on various numbers I've seen that seems like a good guess for such a service if you don't have a diner (the costs for adding additional sleepers being relatively small). Can that be managed? Good question, and one that I'd need a bit more data than I have now:

-I'd need to be able to make assumptions on coach traffic and seat turnover.

-More importantly, I'd need to also be able to make assumptions on daily traffic generated on the route...and the estimates I have on that front, from various EIS work, is hit-or-miss. In the case of Hampton Roads, 9x daily trains are projected to generate somewhere around a million riders per year out of Hampton Roads. I specify because this does not take into account, for example, traffic to/from Richmond not originating in Hampton Roads that those trains would carry (and the closest thing to an estimate I've heard on the Richmond-Washington route is at least another million added, possibly two million).

So...there's a lot of ridership to be had. The share of ridership you'd need to divert to an overnight service out of that to make a success is, if not negligible, at least not massive.

*The late night departure could also facilitate better equipment cycling.

 

[James Patten]

I'm no where near retirement and pretty much need to the take the train now if I'm going anywhere long distance and I don't want to drive - I discovered this year that flying is no longer much of an option for me (vertigo) unless it's extreme need. As my wife travels with me, we use sleeper. Not cheap, and I would like it to be much cheaper!

One of my coworkers would love to be able to take the train long distance, but there are 2 kids and a wife so sleepers are out of his price range. If the price came down significantly - or the trains sped up significantly - he'd be willing/able to take a long distance trip.

 

[Anderson]

Was out of town and am glad to see this topic still going strong a few days later!

I have a feeling that if a half dozen or dozen of the posters on this board could seize control of Amtrak it would be in much better shape....

I think the core of my question was really getting at the idea of whether the population could be taught (Please forgive my propagandistic language, but that's really what it would be.) to take trains again. My sense is that with advertising people can be led to buy just about anything, so there may be hope after all. In 1950 trains were part of the culture, and it was not shocking to imagine traveling by train. But today I just get the sense from the vast majority of people I personally meet (all across the country, many income and education levels, mostly in cities), that the train is just not on the radar. Before I relocated to my current home in Georgia I frequently took the Crescent to and from PHL, and I did so for reasons of scheduling, cost, and to a certain extent comfort/enjoyment/pleasure. I sometimes had the sleeper, always with AGR points, and I would have paid for it if I could have. (Current prices are prohibitive for me.)

But for my fiancée the train was and is never an option. Regardless of costs, the idea of a 15 hour train trip (Admittedly in my original scenario that time would have been cut down with various improvements, but the point is the same.) for her was just never ever an option. She has done LD train travel with me and loved it, but she said she would never do it alone. So I do get the sense that for most people it's either fly or drive, and I expect it is just some kind of perception that the train is not an option. It's just not part of their thinking.

But as I said, if the options were there and with proper investment in effective advertising (Think a beer ad, but on the train, so some hot 20-somethings board separately in coach, hit it off in the diner, and end up UPGRADING to a sleeper en route...TALK ABOUT AN UPGRADE....), maybe sleepers would have a chance....

Or perhaps I've just watched the film from which my name is taken too many times.

I don't think it's necessarily a matter of people being "taught"...simple availability and flexibility goes a long way. Take my Florida scenario and combine it with the suggested 4x daily CHI-NYP LSL corridor. Ok, now you've got a whole swath of cities from Chicago to New York to Miami that can take a train with some flexibility and an assurance that if they miss one train (be it due to traffic or due to a missed connection), there's another in a few hours instead of the next day. Same if one train is sold out...you're not likely to be stuck having to move trips around by days due to space availability issue (affordability is another story entirely). Throw in a Broadway/Capitol corridor of some kind and suddenly you actually start getting regional connectivity...especially if some of those trains can be extended to Boston instead of just New York. You see where this goes...

The main problem, as I see it, is that one's travel options on the train are constrained. For example, take the mess on the LSL/CL corridor...were there 2-3 trains per day out of Chicago going west, I would have grabbed one. Were an option there to take an earlier train (even one with a nominally awful arrival in CHI) I could have booked that with the expectation of it running late. But instead, I effectively had one option and was up the creek if it failed...and that's the case in most of the system.

Edit: Something that should probably be said, as to differences between 1960 and soon-to-be 2015:

(1) The US has about 140 million more people. In 1960 the population was 180 million; we're creeping up on 320 million now.

(2) Public transit ridership has bottomed out and started recovering. One of the factors that brought trains down in the past was the disintegration of mass transit at all levels, something that has stopped (and in some cases reversed in a very real way).

(3) America's love affair with the car is increasingly an open relationship. Back around 1960, having a car was the thing everyone wanted. While I think it is fair to say that most people probably still want a car, not everyone wants to live in a multiple-car household and teenagers are getting their license right off the bat less and less now.

(4) Highway congestion isn't getting any better. In many cities, not only is traffic bad, but highways are basically built out to capacity. There was infill capacity back in the 60s when many roads were built; in many places, that capacity is now gone.

(5) And air travel isn't any better. Legroom has dropped from the middle 30s to the high 20s on many airlines (28-30 inch coach seating is increasingly pervasive), security is unpleasant, and a number of airports are slamming into capacity issues. Again, contrast with the 60s when many of those airports were new, security was non-existent, and (as noted) coach then was more like a lot of business class seats are now

Basically, most of the conditions that led to a decline in rail travel have at least changed, if they haven't reversed outright.

 

[Paulus]

(5) And air travel isn't any better. Legroom has dropped from the middle 30s to the high 20s on many airlines (28-30 inch coach seating is increasingly pervasive), security is unpleasant, and a number of airports are slamming into capacity issues. Again, contrast with the 60s when many of those airports were new, security was non-existent, and (as noted) coach then was more like a lot of business class seats are now.

On the other hand, air travel was a whole heck of a lot more expensive and dangerous back then. Coach was a lot more like business class seats are now precisely because coach cost as much as business class seats do now. So if you want 1960s treatment, you just have to pay for it.

 

[TML]

I despise flying since past flights have made me airsick. Furthermore, I tried taking overnight Greyhound buses and found that such trips invariable leave me fatigued at the end. Thus, I will always try to book a sleeper for Amtrak if there is overnight travel involved.

Furthermore, the size of this country means that demand for sleepers will always exist. This is in contrast to smaller countries like Japan, for example, where sleepers have almost disappeared completely due to its small size and availability of high-speed rail and air travel there.

 

[Anderson]

I despise flying since past flights have made me airsick. Furthermore, I tried taking overnight Greyhound buses and found that such trips invariable leave me fatigued at the end. Thus, I will always try to book a sleeper for Amtrak if there is overnight travel involved.

Furthermore, the size of this country means that demand for sleepers will always exist. This is in contrast to smaller countries like Japan, for example, where sleepers have almost disappeared completely due to its small size and availability of high-speed rail and air travel there.

Japan is an unusual case for a whole host of reasons. Of course, one thing rarely mentioned is that the original plans for the Shinkansens (which I probably misspelled) was to run overnight services as well. That largely fell by the wayside due to maintenance issues more than anything (having the tracks closed for five hours per night is useful for doing work without disrupting service).

 

[jis]

(5) And air travel isn't any better. Legroom has dropped from the middle 30s to the high 20s on many airlines (28-30 inch coach seating is increasingly pervasive), security is unpleasant, and a number of airports are slamming into capacity issues. Again, contrast with the 60s when many of those airports were new, security was non-existent, and (as noted) coach then was more like a lot of business class seats are now.

On the other hand, air travel was a whole heck of a lot more expensive and dangerous back then. Coach was a lot more like business class seats are now precisely because coach cost as much as business class seats do now. So if you want 1960s treatment, you just have to pay for it.

Although it should be remembered that while Coach seat pitch has gone to hell in a hand basket, ever since the dawn of the jet age, Coach seats have never been much wider than 17" - 18", specially in narrow bodies. That pretty much got set in stone with the cabin width of the 707 and DC-8. Airbus improved things only slightly in the A320 family.

 

[Anderson]

(5) And air travel isn't any better. Legroom has dropped from the middle 30s to the high 20s on many airlines (28-30 inch coach seating is increasingly pervasive), security is unpleasant, and a number of airports are slamming into capacity issues. Again, contrast with the 60s when many of those airports were new, security was non-existent, and (as noted) coach then was more like a lot of business class seats are now.

On the other hand, air travel was a whole heck of a lot more expensive and dangerous back then. Coach was a lot more like business class seats are now precisely because coach cost as much as business class seats do now. So if you want 1960s treatment, you just have to pay for it.

That depends. In some cases, upgrading to some sort of "plus" category is pretty straightforward and not entirely unaffordable, though this depends on the airline and the aircraft. In others, the only "real" upgrade is to "First Class".

On the other end of things, though, I feel compelled to point out the fact that at least on the coach side, Amtrak LD fares now are quite a bit less than they were before A-Day once you adjust for inflation. I poke about at the Silver Service fares every-so-often since that's the most easily documented (and since those trains were doing well all along), and the general rule is that unless you're looking into a last-minute ticket at a peak day, coach fares tend to be something like half of the pre-1971 fare.

 

[jis]

I despise flying since past flights have made me airsick. Furthermore, I tried taking overnight Greyhound buses and found that such trips invariable leave me fatigued at the end. Thus, I will always try to book a sleeper for Amtrak if there is overnight travel involved.

Furthermore, the size of this country means that demand for sleepers will always exist. This is in contrast to smaller countries like Japan, for example, where sleepers have almost disappeared completely due to its small size and availability of high-speed rail and air travel there.

Japan is an unusual case for a whole host of reasons. Of course, one thing rarely mentioned is that the original plans for the Shinkansens (which I probably misspelled) was to run overnight services as well. That largely fell by the wayside due to maintenance issues more than anything (having the tracks closed for five hours per night is useful for doing work without disrupting service).

Also, most Shinkansen aficionados do not realize that there are 48 daily flights between Tokyo and Osaka (various airports), the two anchor points of the Tokaido Shinkansen (translates roughly to Tokaido New Line, the original) most of them using wide bodies (767, 777, 787) between JAL and ANA. The nature and size of that market is way beyond anything that we can conjure up apparently.

Can you imagine a flight every 20 mins from New York area airports to Washington area airports with wide bodies and in addition a Shinkansen level intensity of service on the NEC?

 

[Anderson]

I despise flying since past flights have made me airsick. Furthermore, I tried taking overnight Greyhound buses and found that such trips invariable leave me fatigued at the end. Thus, I will always try to book a sleeper for Amtrak if there is overnight travel involved.

Furthermore, the size of this country means that demand for sleepers will always exist. This is in contrast to smaller countries like Japan, for example, where sleepers have almost disappeared completely due to its small size and availability of high-speed rail and air travel there.

Japan is an unusual case for a whole host of reasons. Of course, one thing rarely mentioned is that the original plans for the Shinkansens (which I probably misspelled) was to run overnight services as well. That largely fell by the wayside due to maintenance issues more than anything (having the tracks closed for five hours per night is useful for doing work without disrupting service).

Also, most Shinkansen aficionados do not realize that there are 48 daily flights between Tokyo and Osaka (various airports), the two anchor points of the Tokaido Shinkansen (translates roughly to Tokaido New Line, the original) most of them using wide bodies (767, 777, 787) between JAL and ANA. The nature and size of that market is way beyond anything that we can conjure up apparently.

Can you imagine a flight every 20 mins from New York area airports to Washington area airports with wide bodies and in addition a Shinkansen level intensity of service on the NEC?

I can imagine it, but not with wide-bodies...though that is more a side-effect of mediocre transit from the airports that can host them into the cities (EWR and IAD are a good distance out from NYC and Washington, respectively). My understanding is that in Japan, the local transit connections are quite good by comparison (and if I'm not mistaken, there are a couple of private transit systems that are listed on the stock exchange there[!]). Of course, that brings to mind a serious question: Is airport security in Japan less of a mess than it is in the US?

 

[jis]

Narita Airport is something like 70km (~45 miles) from central Tokyo, and Osaka Kansai (KIX) is 38km (24 miles) from Osaka downtown. Local transport connection is very good. but 70km is 70km, and unless you want to shell out the big bucks to take the relatively fast though not Shinkansen speed NeX, it does take its time to get to the airport. Haneda OTOH is very close to Tokyo, a short monorail ride (17 mins on Express) from Hamamatsucho station on the Yamanote (Circle) Line station.

For frequent travelers airport security is no more of a mess in Japan than it is in the US. Japan afterall has their own internal crazies who occasionally try to insert poison gas into the air circulation system of Tokyo subways and such. OTOH, the messiness of Security for frequent travelers in the US at present is a bit overstated in AU anyway.

 

[neroden]

From what little I know of Japan, I'd expect a far more sensible, orderly, and above all *predictable* airport security system there. What finally caused me to say "to hell with this" to US airport security was the arbitrariness, the capriciousness, the way the so-called "rules" would change from day to day or depending on who was doing the scanning. Another way to describe this problem is "corruption".

Without knowing too much about Japanese airports, I'd expect that their security would be, at least, consistent, predictable, and documented in advance.

 

[neroden]

Well, a couple of thoughts pop into my mind:

On speed limits: Those are maximum speed limits, and are often not found anywhere close to "everywhere".

Certainly. However, the states with *low* maximums are worth looking at; the situation in New Hampshire, for example, seems to scream out for train service.
Furthermore, the more I look at it, the more promising Kentucky looks. There are promising train routes. Cincinnati-Lexington, in particular, is *already owned by a government* (much like the North Carolina situation); if the cities of Cincinnati and Lexington wanted to get passenger service going, I'm pretty sure they could get cooperation from their *lessee*, NS. Indianapolis-Louisville would require massive rehabilitation, but is geometrically fine and owned by a shortline who would be cooperative. Louisville-Lexington is geometrically somewhat worse, but again owned by cooperative agencies, and there was a previous push for service. Kentucky's worst problem in terms of rail development is its neighbors, Ohio and Indiana. :-(

As to those services, I guess the question is this: If a train can't cover a given distance in less than 6 hours or so, and driving is a similar story, then does it make sense to look at a sleeper service with early boarding or something similar?

Sommmmetimes.

For example, let's take Hampton Roads-New York City.

Yeah, here it makes sense to look at it. It's worth nothing one key feature of this which makes it worth looking at.
You can extend service north of New York to provide sane Hampton Roads - Boston service. (& Hampton Roads-New Haven, Springfield, etc.) Trains don't do so well if they're point-to-point, so the ability to push further matters. Even if you don't extend the actual train, the connections are valuable.

It would be better if you could also extend or connect to further daytime service south of Hampton Roads, but you can't, unfortunately. I guess ideally, then, you'd run "daytime" from Hampton Roads to Richmond, sleep from Richmond to NYC, and run in the day from NYC onward to Boston.

Right now, that train trip takes around 8:00-8:30.

Actually, doing a little research...

https://jawbone.com/blog/circadian-rhythm/

I had to figure out typical bedtimes and how long people sleep in particular markets.

Hmmm. The northbound should operate on a schedule something like this:

Hampton Roads 9 PM

Richmond 11 PM

NYC 7 AM

New Haven 9 AM

The southbound should operate on a schedule something like this:

New Haven 9 PM

NYC 11 PM

Richmond 7 AM

Hampton Roads 9 AM

This would likely be preferable to the current schedule of #66/67, which is designed for Boston-Washington service and doesn't really serve New York. (I'm going to bet that NY+points north to Richmond+Hampton Roads is a stronger market than Boston to Washington+points south. Maybe I'm wrong.) Of course, the commuter railroads would be unhappy about the service arriving in peak commuting hours in New York, but that's what it *should* be doing.

(Now, if I look at California, I can't come up with anything as attractive as this; the cities simply aren't spaced out right.)

I don't think it necessarily does; trying to go overnight is trying to poorly compete with flying while those who've chosen to drive are doing so in the daylight hours (generally speaking), so there's no waste of time by running in the daylight.

This is incorrect; going overnight is competing with driving. If I'm comparing driving to taking the train, if I can sleep on the train, I end up with an effective shorter trip time from "last hour of operations at home" to "first hour of operations at destination". Going overnight -- over the *entire* night -- is very much competing with driving. Or more accurately, it's competing with driving + hotels. It's a non-linear relationship; travelling over *half* the night fails.
To use the most western and longest-run example I feel comfortable with, consider Denver-Chicago. There's no competition with air, so consider only competition with driving. Driving is a solid 14-16 hours -- a long day, longer if you stop for meals. The *entire* day is gone, along with the nights before and after it. At the moment the train takes, annoyingly, almost 19 hours. But suppose that Iowa's 2013 "Chicago to Omaha Service Development Plan" has been implemented, with trip times of 7.48 or better from Council Bluffs to Chicago. (7.48 is for 79 mph service, 5.40 for 110 mph service.) Then you run overnight for about 9 hours (from Denver to Omaha, where there are no intermediate cities to speak of) and in the day for less than 8 hours (from Omaha to Chicago). You have an afternoon and evening in Chicago; you've saved half a day over driving.

Obviously, if you're going from Denver to Omaha, you save a lot more time; either you can drive for 8 hours, or you can go to sleep in Denver and wake up in Omaha. Unfortunately Denver-Omaha is a pretty small market, but it presents an example of the time savings of sleeping car service over driving. The current eastbound California Zephyr schedule, however, is *not* ideal; it leaves Denver too early & arrives in Omaha too early. (The westbound is about right.)

In most of the Northeast and Heartland, you don't have such a convenient layout of population with an almost-exactly-8-hour-long gap in population as you do on the route to Denver; instead, you have cities all along the way. So by running overnight, you lose most of the ridership from some intermediate cities which would be good ridership sources (Cleveland). It is therefore important that you already have day trains covering those cities before you add an overnight train. And this is where the 1960s train-offs left us with a really ass-backward situation, leaving us with the overnight "name trains" while losing the underlying day trains on the route. In general, a good sleeper service is running along a route which already has daytime service.

Here's the way to think of overnight sleeper service: it's a way of getting a "virtual" runtime advantage over driving. You can have a longer runtime, but to the customer, it is effectively shorter, because the time sleeping doesn't count.

This is also the correct way to think of food service on the trains. For people who would have *stopped at a restaurant to eat* when driving, the food eaten while moving provides a "virtual" runtime advantage.

I think if you use the "virtual runtime advantage" analysis consistently, it vastly clarifies when sleeping car and dining car service are beneficial, and exactly how they are beneficial. The subtle point here is that a real runtime advantage benefits all city pair markets along the route, whereas these "virtual" runtime advantages only benefit *some* of the city pairs -- and in the case of overnight running, hurt other city pairs. So it requires careful planning.

 

[Tokkyu40]

.... The airlines that ran passenger rail out back in the 50s and 60s had baseline seating in the mid-30s in terms of seat pitch and included a lot more amenities in many respects...

Some of the amenities included government provided airports, paid for at government expense and exempt from property tax and government paid and provided traffic control.

It's not easy for private tax paying companies to go head-to-head with government mandated and sponsored competition.

 

[Anderson]

.... The airlines that ran passenger rail out back in the 50s and 60s had baseline seating in the mid-30s in terms of seat pitch and included a lot more amenities in many respects...

Some of the amenities included government provided airports, paid for at government expense and exempt from property tax and government paid and provided traffic control.

It's not easy for private tax paying companies to go head-to-head with government mandated and sponsored competition.

That's not quite what I was thinking. I'm more thinking of the reductions in food service, etc. that we've seen over the years.

 

[neroden]

Hmm. Reading another article, I stumbled across a link to this:

http://www.huffingtonpost.com/2014/11/20/airports-with-most-delays_n_6140298.html

This may have some relevance to determining the markets where passenger trains might have unusually high attractiveness vs. airplanes. (From this, you'd think Chicago, New York, Florida, and possibly Denver.)

New York comes up over and over when you do any sort of "where does the road & air competition stink" analysis, with Chicago coming up nearly as often. It's interesting to see Florida and Denver show up, and I'm not entirely surprised.

 

[domefoamer]

potential to reI thouht we started out talking about sleepers and their potential. But all but one response on this page ignore sleepers, delving instead into the familiar big-picture questions of rap vs air vs highway. Maybe that's what everyone wants to talk about, but there are other places for that discussion.

To put it as simply as I can, I'd love to see several dozen new sleepers added and operated at reasonable rates. Price is important, and the common prices of $600-1000 per night are out of reason and reach for most of us. Twenty years ago I slept on one of the last Slumbercoaches for $55 extra; you ought to be able to offer a bed for five or six times that price now.

All of this takes investment and subsidies that are unlikely right now, I know. But sometimes you have to spend money to make money. I believe it would be cheaper and far quicker to make trains more comfortable, at present speeds, than to rebuild the nation's overloaded rail routes to increase their speeds.

 

[Anderson]

NEC-Florida, NEC-Chicago, and Chicago-Denver are the "obvious" markets. Some secondary ones that jump to mind would be NEC-Atlanta and LA/SD-SF.

For what it's worth...yes, when CAHSR is complete LA-SF becomes a bit short. However, even on the most optimistic build-out we're looking at around 15 years from now...and frankly, I think that schedule has about as much a chance of being kept as most megaprojects out there...and really, until the build-out and electrification is complete you can't do a full end-to-end run and a sleeper service still has a good market (in fact, an SD-LA-SF run over the Central Valley alignment should be doable in 8-9 hours LAX-SFO while the HSR option, with transfers at San Jose and Burbank, is probably not going to be capable of much better than something in the six-hour range: 30 minutes LAX-BUR (Amtrak Surfliner), 60 minutes SJC-4th and King (Caltrain Baby Bullet), and budget 20 minutes at one station for a quick transfer and 40 minutes on the other end...that's 2:30 on the ends, bare minimum for safe budgeting. Add in...I think we'd be looking at about three hours for BUR-SJC? So that's 5:30-ish, which is probably favorable. If the link were to slide back to either Merced (that connection doesn't happen until 2026) or Palmdale (as was formerly the case in old versions of the plan), those connection times get very ugly very quickly. LAX-Palmdale would run closer to 90 minutes on an express (the locals, IIRC, run about two hours over that route)...throw another hour in and suddenly the "high speed" trains aren't that much faster, in effect, than an overnight service would be...and they take a single-seat trip and turn it into a three-ride trip before you get to local transfers (i.e. BART, LA Metro).

 

[Paulus]

1) Where are you getting all these transfers from? By the time the Authority hits San Jose, Caltrain will be electrified, so it'll be a straight shot in to downtown San Francisco, just not as many frequencies doing so until full buildout.

2) IOS is Merced to San Fernando Valley (likely LAUS actually, lots of pushing for that and it's a cheap extension) in 2022 which gets you a five hour one or two seat ride LA-SF (depends on whether transfer at Merced or diesel locomotive hauls from there).

3) 2026 is one seat three hour ride LA-SF

4) LA-SF via the CV by diesel is 11 hours, not 8. It's five hours over Tehachapi and fat chance of getting an agreement to use that.