With trains running at 50% of capacity, how to improve?

I see from the Amtrak August 2014 monthly report that trains fill only about 50% of capacity with passengers.

Airlines have done a huge push to eliminate excess capacity, by cutting back routes, using smaller planes, aggressive pricing, etc., and their capacity is now pretty much totally used up at peak periods, at mostly used up at off-peak periods.

Is Amtrak doing the same thing? I think not, given how frequencies of trains and train sizes have either stayed the same or increased in the past few years. If not, what should Amtrak be doing to fill more capacity and cut its losses?

An apples to oranges comparison, as I'm sure you're aware. Airliners fly point to point with (unless something goes really wrong) no intermediate stops. Trains make intermediate stops, so they'd better have some empty seats available. The proper comparison for Amtrak would be commuter trains or other (i.e., foreign) intercity trains. I don't know how Amtrak compares on scale, but it would be more appropriate than comparing it with the airlines. That said, Amtrak should always be trying to sell more tickets.

Simple solution.

Remove 1/2 of all cars and you suddenly have 100% occupancy.

And no room for surges. More customer dissatisfaction as they NEVER get an empty seat next to them.

Greg Moore wrote:Simple solution.

Remove 1/2 of all cars and you suddenly have 100% occupancy.

And no room for surges. More customer dissatisfaction as they NEVER get an empty seat next to them.

If a train starts at one end 20% full and ends at the other end 80% full, you only have 50% occupancy. On the reverse trip, you can expect to start 80% full and end 20% full. Of course some days of the week (Fridays) occupancy will be higher and other days it will be lower.

Now if you could add a car every 50 miles or so inbound and drop off a car every 50 miles or so outbound, you could keep the trains mostly full and keep the graffiti artists busy. No, I realize its not a practical solution. The rolling resistance of an empty rail car is low enough that its much easier to keep the consist together than it is to scatter loose cars all around the country.

If you read the august 2014 report it specifically mentions more aggressive pricing to fill seats (in reference to the regionals). will amtrak ever match airline load factors? probably not. can they do a better job filling existing capacity? absolutely.

Is this just for coach on Amtrak? One can hardly get a sleeper reservation on MANY overnight long distance trains....typically I had no problem booking a bedroom a few weeks in advance, but on a recent trip - all sold out and I took a roomette. I wanted to be upstairs and had to take room 9 on a Superliner. This room is right over the wheels and you have the door opening and closing all night long it seems.

Greg Moore wrote:Simple solution.
Remove 1/2 of all cars and you suddenly have 100% occupancy.
And no room for surges. More customer dissatisfaction as they NEVER get an empty seat next to them.

As noted, cutting 1/2 of all cars isn't quite the solution, but I'll also say that 100% of capacity is not a problem. Nobody ever went broke running seats that were too full. (you go broke when seats are too empty, and 50% is too empty...and many seats go only ~45% filled)

If Amtrak could run at 100% LFs, all the time that would be a very good thing. Surges would be managed by pricing. Yes, other modes would pick up the slack, but the switchers are almost by definition not in Amtrak's sweet spot.

If the train is 100% full, it will be full with people who are satisfied with full trains. And people who want an empty seat next to them will be dissastified, but we're all net better off replacing 1 passenger in 2 seats with 2 passengers in 2 seats.

And 100% full means that Amtrak is providing the mobility to Americans that it was founded to provide.

Meanwhile, I can attest that in BOS-PVD, Amtrak has a Saturday $15 fare that is very competitive with the MBTA's $13.50 cent fare (and, although 2 kids <12 ride free on an MBTA fare, Amtrak had a 2 kids @ $15). BOS-PVD is exactly the sort of market that Amtrak should strive to max out--but "max out" means that all the "new" BOS-PVD locals alight at PVD and leave a warm-but-empty seat for all the PVD-South trippers.

The "unsolvable" trains are the one-ended ones (there aren't enough people in Carbondale or Rantoul to fill the seats between the two on the way to/from Chicago)

Also tough are the "empty middle" LD markets, where a "whole length" trip is too long to fill the middle and the endpoints are too far apart so that the "tapering off" from one end happens faster and ends sooner than the "ramp up" to the other end starts.

Dropping/adding coaches is effectively out of the question, but starting more trains "midway" (like the LYH train overlaid on the Crescent) *is* a way of adding capacity "mid-trip" ...just a little bit chunkier, that's all. The question there is who should pay (and that the LDs are designed-to-fail given that they must run endpoint-to-endpoint).

BTW, for anyone not recognizing my style, my proposal was NOT a realistic one, more a sarcastic one.

More to the point as others have pointed out, trains are not like airliners in many ways. They don't run just point to point. Secondly, an airline has more difficulty changing seating numbers. (someone like 737 can only change them in increments of "another 737"). Adding a 2nd car to a 1 car train doesn't suddenly double costs (I'd take a SWAG and say 20% more cost/). But it does double seating. So you could go from 100% load factor to 75% and actually be doing better.

Ultimately, 50% may be an issue but by itself it's just a number. One has to really look at where it's 50% of capacity and why and try to improve each area separately.

I would also point out that Airlines achieve their high occupancy rates by "Hubbing". Try booking a flight from one small city to the other, more than likely it will be via Atlanta or Houston for e.g.

Arlington, the MBTA fare to Providence is only $10.50, not $13.50. Much less competitive with Amtrak.

Honestly, I disagree about BOS-PVD being a good market for Amtrak. At less than an hour trip time, it's much better suited to the MBTA's commuter service, which is cheaper and easier to use.

David Benton wrote:I would also point out that Airlines achieve their high occupancy rates by "Hubbing".

Not really...or certainly don't let trains off the hook because they don't have hubs.

Airlines have had fortress hubs since the early 1990s, at a time when load factors were in the upper 60%s. It wasn't hubbing, but rather capacity reduction and smarter pricing that drove average load factors from 71.2 percent to 79.7 percent from 1999 to 2006. It reached a then-record-high of 80.5 percent in 2007, and most US airlines strive to operate consistently in the 80% range today.

If trains actually went where people needed them, and cut capacity where it isn't used, they'd run fuller. This requires either nipping routes back (the Downeaster added a whole lot of empty miles--but few new passengers--when it went beyond Portland ME) or not extending them to empty places in the first place

Fixing it should mean things like:
1) Reduce dangling ends. Illinois would do better if it had a big city (even an out-of-state one) on the other end of its routes.
- Carbondale trips should continue on to Memphis.
- Quincy routes should either be trimmed back to Peoria or extended on to STL or KCY or MEM

2) Not serving big "mid-route" cities at bad times (e.g. Charlotte on the Crescent, Cinci on the Cardinal, PIT on the Capitol Ltd.)

deathtopumpkins wrote:Honestly, I disagree about BOS-PVD being a good market for Amtrak. At less than an hour trip time, it's much better suited to the MBTA's commuter service, which is cheaper and easier to use.

You're thinking like an engineer, and not an economist, as required here. Empty seats are killing Amtrak. Markets like WAS-BWI and BOS-PVD should be pursued on the basis of price for their marginal dollars. If people value their time at just $10 per hour, saving a half hour (:40 on Amtrak vs 1:10 on MBTA) is easily worth the $5 more to go Amtrak. So $15 on Amtrak vs $10 on MBTA is competitive. Frankly, since MBTA isn't free to discount on Saturdays, Amtrak's profit-maximizing strategy is going to be somewhere in the $10 to $15 range.

Arlington wrote:

David Benton wrote:I would also point out that Airlines achieve their high occupancy rates by "Hubbing".

Not really...or certainly don't let trains off the hook because they don't have hubs.

Airlines have had fortress hubs since the early 1990s, at a time when load factors were in the upper 60%s. It wasn't hubbing, but rather capacity reduction and smarter pricing that drove average load factors from 71.2 percent to 79.7 percent from 1999 to 2006. It reached a then-record-high of 80.5 percent in 2007, and most US airlines strive to operate consistently in the 80% range today.

If trains actually went where people needed them, and cut capacity where it isn't used, they'd run fuller. This requires either nipping routes back (the Downeaster added a whole lot of empty miles--but few new passengers--when it went beyond Portland ME) or not extending them to empty places in the first place

Fixing it should mean things like:
1) Reduce dangling ends. Illinois would do better if it had a big city (even an out-of-state one) on the other end of its routes.
- Carbondale trips should continue on to Memphis.
- Quincy routes should either be trimmed back to Peoria or extended on to STL or KCY or MEM

2) Not serving big "mid-route" cities at bad times (e.g. Charlotte on the Crescent, Cinci on the Cardinal, PIT on the Capitol Ltd.)

And SWA has no hubs (at least not technically, although they sort of do in practice). Yes, Carbondale trains scream to be extended to Memphis, but we know IL won't pay for it. Also, the Quincy trains don't go through Peoria, and there is no good route from Chicago to Peoria in any event. You could also throw in extending the Vermonter back to Montreal (mighty few people getting on and off in St. Albans).

Another area to look at in regard to load factors is operational requirements/exigencies. For example, nb LDs run discharge-only north of WAS based on their uncertain arrival times. Also, going back to the Vermonter, it runs to St. Albans because that's where it can be turned/serviced. Likewise, the Downeasters will (once it's built) be going to Brunswick to access the layover facility there. In those situations, Amtrak is faced with either deadheading the train (0% load factor) or selling what tickets they can to either St. Albans or Brunswick. As for not serving mid-route cities in the wee hours, if you're going to have LD's, someone's going get shafted on arrival/departure times.

Finally, in regard to pricing, Congress (ever wont to micromanage Amtrak) puts restrictions on the amount of discounting Amtrak can do on non-state-supported routes.

That would make sense, seeing as I am an engineer, not an economist!

But let me illustrate my thinking with an example. Occasionally me and some coworkers will need to travel to the Providence office for something. It never crosses anyone's mind to take Amtrak. In fact usually most people drive, and me and a couple others hop on the MBTA. Even if the schedule might be slightly better on Amtrak, none of us justify the added cost. Plus since I have a monthly from another zone, the fare was only $3.25.

So anyone from metro Boston who regularly rides the CR would pay even less than $10.50, further driving them to the MBTA. And even people who don't aren't going to think to take an intercity train to Providence. They're going to think to take the T.

I know that's all anecdotal, but I just don't see Amtrak being able to woo very many people on routes that it shares with commuter trains that are fairly short and direct and not ridiculously crowded.

Station Aficionado wrote:Yes, Carbondale trains scream to be extended to Memphis, but we know IL won't pay for it.

I'm thinking that it's Amtrak's job to show Illionis how to cut its losses and boost ridership. Illinois isn't the expert here, Amtrak is. And adding MEM (or STL)

Station Aficionado wrote: Also, the Quincy trains don't go through Peoria, and there is no good route from Chicago to Peoria in any event.

Galesburg, then. Make it an intermodal terminal and increase its frequency (make it worth a drive from Peoria or Moline).

The point is that it is expensive to send trains out to the far hinterlands only to run empty at the dangling end, vs cheap to make people drive those last 50 to 100 miles. What airlines do really well is sending out a prop feeder to those little towns, not a "whole" airplane. Lynchburg VA goes like gangbusters with a "real" NEC train because it can make folks drive (or bus) from Roanoke and Blacksburg (VaTech/Radford/etc. Going to Roanoke is more politics than good route-planning (it'd have be better to double down on the siding there at Lynchburg and make it "always worth driving to" Oh well.)

Station Aficionado wrote:You could also throw in extending the Vermonter back to Montreal (mighty few people getting on and off in St. Albans)

Yes. Many of these dangling ends would probably be more cheaply served with a free taxi ride than sending a whole train, and have the "taxi" go to a place that has more (but perhaps smaller) trains per day.