Thursday, December 6, 2012

Amtrak routes by 2012 cost recovery

With the release of the September 2012 Monthly Performance Report, we now have information on costs and revenues for Amtrak's routes throughout Fiscal Year 2012. With that information I've put together a spreadsheet of all of Amtrak's routes ranked according to their 2012 cost recovery level. In order to remove the affects of state support subsidies, I've considered only ticket revenue (found on Page A-3.5) against total expenses (page C-1) for the state supported routes. This has the side-effect of removing food and beverage revenues as well while keeping their costs and so state supported routes may actually be a few percentage points better for farebox recovery. The spreadsheet follows after the cut and is color coded for convenience in assessing between Northeast Corridor, state supported, and long distance trains.




Revenue 2012
Expenses 2012
Farebox recovery 2012
Revenue 2011
Expenses 2011
Farebox recovery 2011
Change in farebox recovery
Change in revenue
Change in expenses
Acela
$521,100,000
$314,500,000
165.69%
$510,300,000
$331,600,000
153.89%
11.80%
2%
-5%
Washington-Lynchburg
$11,411,821
$8,100,000
140.89%
$9,826,802
$6,900,000
142.42%
-1.53%
16%
17%
Northeast Regional
$552,800,000
$480,600,000
115.02%
$505,300,000
$477,300,000
105.87%
9.16%
9%
1%
Washington-Newport News
$34,286,847
$31,900,000
107.48%
$29,682,574
$31,300,000
94.83%
12.65%
16%
2%
Carolinian
$18,652,552
$20,700,000
90.11%
$17,720,525
$21,900,000
80.92%
9.19%
5%
-5%
Albany-Niagra Falls-Toronto
$24,600,726
$28,700,000
85.72%
$23,406,596
$30,700,000
76.24%
9.47%
5%
-7%
Keystone
$32,970,951
$47,800,000
68.98%
$29,366,992
$47,000,000
62.48%
6.49%
12%
2%
Auto Train
$74,100,000
$109,000,000
67.98%
$69,900,000
$101,500,000
68.87%
-0.89%
6%
7%
Empire
$43,877,344
$66,400,000
66.08%
$40,077,158
$71,900,000
55.74%
10.34%
9%
-8%
Palmetto
$18,400,000
$30,000,000
61.33%
$17,400,000
$34,000,000
51.18%
10.16%
6%
-12%
Pennsylvanian
$9,281,813
$15,700,000
59.12%
$8,856,539
$16,800,000
52.72%
6.40%
5%
-7%
Hiawatha
$15,963,261
$27,100,000
58.91%
$14,953,873
$25,900,000
57.74%
1.17%
7%
5%
Empire Builder
$72,200,000
$131,200,000
55.03%
$57,700,000
$112,300,000
51.38%
3.65%
25%
17%
Ethan Allen
$2,829,307
$5,200,000
54.41%
$2,504,308
$6,600,000
37.94%
16.47%
13%
-21%
Pere Marquette
$3,276,210
$6,100,000
53.71%
$3,197,106
$6,800,000
47.02%
6.69%
2%
-10%
Silver Meteor
$42,600,000
$81,300,000
52.40%
$41,600,000
$85,600,000
48.60%
3.80%
2%
-5%
Adirondack
$6,748,333
$13,000,000
51.91%
$6,301,649
$13,300,000
47.38%
4.53%
7%
-2%
Lake Shore Limited
$35,000,000
$68,100,000
51.40%
$32,900,000
$70,400,000
46.73%
4.66%
6%
-3%
Downeaster
$7,741,844
$15,100,000
51.27%
$7,149,257
$13,500,000
52.96%
-1.69%
8%
12%
City of New Orleans
$22,000,000
$43,500,000
50.57%
$18,800,000
$41,600,000
45.19%
5.38%
17%
5%
New Haven-Springfield
$11,723,569
$23,400,000
50.10%
$11,204,575
$24,400,000
45.92%
4.18%
5%
-4%
Pacific Surfliner
$58,595,820
$118,400,000
49.49%
$55,317,127
$115,400,000
47.94%
1.55%
6%
3%
Capitol Limited
$22,600,000
$47,200,000
47.88%
$22,400,000
$47,000,000
47.66%
0.22%
1%
0%
Silver Star
$38,700,000
$84,600,000
45.74%
$36,300,000
$86,900,000
41.77%
3.97%
7%
-3%
Wolverine
$17,704,897
$38,800,000
45.63%
$18,769,770
$37,200,000
50.46%
-4.83%
-6%
4%
Texas Eagle
$28,500,000
$63,100,000
45.17%
$26,600,000
$56,700,000
46.91%
-1.75%
7%
11%
Coast Starlight
$45,300,000
$101,300,000
44.72%
$44,300,000
$98,100,000
45.16%
-0.44%
2%
3%
Cascades
$30,886,455
$69,100,000
44.70%
$30,025,126
$66,100,000
45.42%
-0.73%
3%
5%
Crescent
$34,900,000
$78,200,000
44.63%
$32,300,000
$77,100,000
41.89%
2.74%
8%
1%
San Joaquin
$38,661,536
$87,200,000
44.34%
$35,704,109
$77,900,000
45.83%
-1.50%
8%
12%
Vermonter
$4,761,018
$10,900,000
43.68%
$3,961,115
$9,300,000
42.59%
1.09%
20%
17%
Chicago-Carbondale
$9,258,647
$21,600,000
42.86%
$8,802,288
$20,600,000
42.73%
0.13%
5%
5%
Piedmont
$3,077,031
$7,200,000
42.74%
$2,498,540
$7,100,000
35.19%
7.55%
23%
1%
California Zephyr
$53,200,000
$124,700,000
42.66%
$49,800,000
$112,500,000
44.27%
-1.60%
7%
11%
Southwest Chief
$48,200,000
$115,900,000
41.59%
$48,000,000
$114,500,000
41.92%
-0.33%
0%
1%
Blue Water
$6,094,659
$15,400,000
39.58%
$5,797,878
$14,000,000
41.41%
-1.84%
5%
10%
Capitol Corridor
$27,927,540
$75,800,000
36.84%
$25,720,252
$69,600,000
36.95%
-0.11%
9%
9%
Chicago-St. Louis
$13,353,833
$39,900,000
33.47%
$12,262,325
$32,400,000
37.85%
-4.38%
9%
23%
Cardinal
$8,400,000
$25,900,000
32.43%
$7,800,000
$26,400,000
29.55%
2.89%
8%
-2%
Kansas City-St. Louis
$5,139,069
$15,900,000
32.32%
$4,763,442
$14,100,000
33.78%
-1.46%
8%
13%
Chicago-Quincy
$5,687,467
$17,600,000
32.32%
$5,580,227
$16,800,000
33.22%
-0.90%
2%
5%
Sunset Limited
$13,000,000
$55,100,000
23.59%
$12,600,000
$51,700,000
24.37%
-0.78%
3%
7%
Heartland Flyer
$2,086,587
$9,200,000
22.68%
$1,911,994
$8,700,000
21.98%
0.70%
9%
6%
Hoosier State
$856,675
$4,700,000
18.23%
$836,057
$4,900,000
17.06%
1.16%
2%
-4%


A copy of the this as an excel file is available. It was originally made with Apple's Numbers, however, so I apologize if the conversion resulted in any screwiness.

I do not think it merely coincidental that, with the exception of the Auto Train, the top 11 performing routes have either a major stop or their terminus in New York City. Were clearances raised such that the Auto Train (or a section) terminated in the vicinity of New York City, I would expect a somewhat substantial increase in its cost recovery as well, train lengths and equipment permitting.

14 comments:

  1. Thanks for putting this together. Do the "expenses" include anything for depreciation on equipment or other capital costs? Also, is anything included for infrastructure on (or off) Amtrak-owned track? (I'd try to look this up in the report, but I can't get Adobe to work this evening.)

    On one hand, it's encouraging see how well some of these services are doing in terms of cost recovery. On the other hand, it's discouraging just how heavily subsidized even the major corridors outside the NE appear to be in comparison.

    ReplyDelete
    Replies
    1. To the best of my knowledge, none of those are attributed in the train operating costs; they're all in a separate budget.

      Delete
  2. Thanks for you hard work.

    I'd be interested in seeing how these routes break out by segment.

    Like NYC to DC is not just Acela and Northeast regional, but also Cardinal, Crescent, Carolinian, and others, at least for part of their routes. So maybe NYC to DC travel is under reported.

    Also for routes like the Empire Builder; how much travel is Chicago to Minneapolis and Seattle/Portland to Spokane versus Havre to Minot?

    ReplyDelete
    Replies
    1. Well, if Amtrak ever gets around to posting their performance improvement plan for the Empire Builder, we'd be able to see, but unfortunately, they've not done that with the last LD routes that were being studied this year.

      Delete
    2. Go to NARP, click Resources, click Facts Sheet by Train.

      The Empire Builder had (2008 figures latest shown) 8.1% of the trips end to end, Chicago-Seattle/Portland. And 60% of the trips were for less than 600 miles.

      So there's some info there. But yeah, until Amtrak posts the PRIIA study on the Empire Builder, I don't know where to get the full story.

      Delete
    3. Anonymous

      NYC-DC actually IS the Acela and Regionals. The long distance trains only stop to receive new passengers. So you board in Philly with your ticket to Raleigh or Charlotte, but Amtrak won't sell you a ticket just to Baltimore or even Washington.

      (You could get off in Alexandria and take the Metro into D.C. Don't know how you'd get home; northbound Alexandria is a discharge-only stop. LOL)

      Delete
    4. Anonymous and Paul,

      I groped for an answer to mid-route performance on the Empire Builder by using the Great American Stations site, where Amtrak posts figures for its various stations. (Very nicely done, I must say.)

      The figures for Chicago and Seattle are useless because they both have so many trains stopping there.

      But look at stations near Chicago, and see Wisconsin Dells with 13,500 users, and LaCrosse 28,900. Maybe most of those are coming from/going to Chicago or Milwaukee, but some are traveling to/from the west.

      That gets us to St Paul, what seems a biggie with 120,000 users. Doesn't seem like so big to me, tho: the Twin Cities metro has a population of 3.3 million, while LaCrosse area has 134,000.

      Of course, St Paul could quadruple its count with (more and probably faster) corridor trains to ChicagoLand.

      When we look at North Dakota station counts, Fargo and Grand Forks boast about 20,300 each. Minot comes in at 37,200. Williston can brag of 54,300 oilfield roughnecks and other riders. (Most of these loyal passengers get 'served' with stops in the dark of night.) North Dakota way, WAY outperforms the Twin Cities.

      In the NW, Spokane gets only 62,800, half the riders of St Paul. And Pasco, an intermediate stop Spokane-Portland, has 25,500.

      These numbers do support the claim of heavy traffic out in the middle of nowhere. Indeed, Amtrak does well in the flyover heartland.

      Delete
  3. It makes me wonder what is the range of cost per passenger and why it varies so much

    ReplyDelete
    Replies
    1. The two worst performing long distance trains both run only three times a week, while 'normal' trains run daily. Until Amtrak can get more locomotives and passenger cars, those disproportionate losses will continue.

      The PRIIA study of the Cardinal estimated that taking it daily would more than double its ridership (to nearly 250,000). That, and more efficient use of staff (who are put up overnight in NYC, and receive 'away pay' while they wait to work the next train) would bring its loss per passenger close the rest of the pack.

      According to the PRIIA study, a daily Sunset Ltd would also gain about 125,000 new riders. (It gets about 100,000 now.) And save on hotels and 'away pay' for on-board staff.

      The Sunset's loss per passenger now exceeds $400, a shocking figure even for an Amtrak fan like me. It's the number the Amtrak haters like to use to attack the whole system.

      The Union Pacific blocked the PRIIA proposal to go daily with the Sunset, demanding over half a billion in track upgrades. Boardman bargained for a much faster schedule (started May, 2012) while delaying a formal request for daily by two or three years.

      Meanwhile Amtrak may come up with the added equipment needed to go daily, and the UP is double tracking the line thru New Mexico and Arizona on its own. So if we can just hold on. Daily service would not end the losses, but the huge increase in riders would slash the loss per passenger by 25%, maybe 33%.

      Both routes serve a good string of cities between their big city end points. Charlottesville, Charleston WVa, Cincinnati, Indianapolis for the Cardinal. Houston, San Antonio, El Paso, Tucson, kinda-sorta Phoenix. (Operators of airport ground transportation told Amtrak they'd gladly serve Maricopa if it got daily stops, but not with 3 a week.)

      But both routes travel thru deserts. The Sunset thru the Chihuahuan and Sonoran, and the Cardinal thru Indiana.

      The Cincinnati-Indianapolis-Chicago corridor should have frequencies and ridership like St Louis-Chicago, but it definitely does not. When/if that line gets upgraded to a corridor, the Cardinal will benefit the way the Texas Eagle hitches a ride on the St Louis corridor, and the Coast Starlight hitches a ride on the Cascades.

      Delete
  4. 50% is a perfectly respectable farebox recovery ratio, looking at Wikipedia. For transit systems, the iffy ones have recovery ratios of less than 30%.

    What this tells me is that
    (1) people like intercity trains
    (2) The Hoosier State, Heartland Flyer, and Sunset Limited are unusually poor services
    (3) People hate disruptions which mess up the schedule. (This goes with most of the drops in farebox recovery.)
    (4) Albany-NYC trains should be extended beyond Albany
    (5) There are basically two networks: a NYC-centered network and a Chicago-centered network. Then there are a few trains on the West Coast.

    ReplyDelete
    Replies
    1. Nathanael, You're admirably succinct while I'm uncontrollably verbose. Sorry.

      Do agree with your points, especially (1).

      (2) The Hoosier State is probably using equipment being held in the yard for the next run of the Cardinal, the worst stuff they've got. And it keeps the terrible schedule of the Cardinal, leaving Indianapolis uncomfortably early but getting into Chicago with most of the morning pissed away. One sunshiny day the tracks will be upgraded, the trip time slashed, new equipment secured, and ridership will rise.

      The Heartland Flyer has the next most raggedy trains. Stimulus funding was to bring new signaling and cut the trip time by half an hour. Maybe that project is parked on a siding somewhere.

      The tri-weakly Sunset and the Cardinal are the biggest wasted opportunities in the national grid. Well, if we can just hold on a few more years.

      (3) People must have hated it when the NS suddenly declared go-slows on the segment of the Wolverine line Dearborn-Ann Arbor-Kalamazoo that it still controlled, just weeks before Amtrak was to go to 110 mph Kalamazoo-Porter en route to Chicago. Call me paranoid, I'm surely cynical, but I'm thinking that the surprise attack of go-slows was a deliberate undermining of Amtrak's PR efforts. They crowded out the good news with the awful service disruptions.

      (4) The Empire Corridor and beyond (what's that to be called, the Mohawk Corridor?) will be benefiting from a nice little set of improvements: Modern signaling on a stretch south of Albany, a new passing track in the Albany station, Amtrak taking over traffic control south of Schenectady, best of all double tracking the bottleneck section Albany-Schenectady, and a passing track for the Rochester station. Maybe 10 minutes out of the trip times to Buffalo?

      Extending NYC-Albany trains to Buffalo looks juicy, with that extra 20%. One train should go on to Chicago, with daylight stops in Cleveland. Sure, as soon as Amtrak gets more locomotives and passenger cars.

      (5) There's enuff population in the Red Sea to build a network centered on Dallas, Houston, Memphis, New Orleans, or Atlanta. For now, Amtrak is where people and politicians want it to be.

      Delete
  5. Paul, A very nice table. Love seeing the routes ranked like this.

    I'm optimistic, because the rankings are about to get all shook up, for the better, all across the board.

    Good thing Obama won a second term, because all those shovel-ready projects his stimulus funded are still works in progress -- but they are starting to come to fulfillment.

    The two new Talgo trainsets have at last left the factory and are undergoing testing at Pueblo, Colorado. Then more testing on the Cascades route before they can go into revenue service. Hope they can get to work before summer (and the fiscal year) is over.

    Two more departures will mean many more riders. Adding two frequencies Portland-Seattle makes the wait between departures shorter, effectively reducing trip times for many customers. The convenience could allow a modest price increase on a route soon to have over a million riders. Even two bucks per ride dropping straight to the bottom line will improve the ranking of the Cascades.

    Track upgrades between Charlotte and Greensboro will cut 20 or 30 minutes out of the ride for the Piedmonts and the Carolinian. Leaving NYC a bit later (now 7:05 am) with time to wake up, or more likely, arriving in Charlotte (now 8:12 pm) in time to kiss the kids goodnight, well, small things make a big difference. And the Carolinian so sooo close to breaking even.

    Work is supposed to take half an hour out of the trips Detroit-Chicago. Arriving in Detroit "before" midnight instead of "after" midnight, that right there is gonna add riders. Then another half hour coming out next year; be in Ann Arbor in time for lunch. Ridership will boom on the Wolverines and the Blue Water branch.

    Likewise St Louis-Chicago, quicker trips coming this year and next as work is finished on various sections. The Lincoln service will conquer.

    The resultant better times in St Louis could help boost riders on the River Runner to Kansas City.

    The Ethan Allen shows revenue up 13%, expenses down 21%, due to investments in the r-o-w. With more to come, when those Empire corridor investments come on line. Because the Ethan Allen and the Adirondack split off at Schenectady, they'll enjoy the times shaved around Albany.

    The Vermonter showed 20% revenue increase, not sure why the 17% cost increase. But I know revenue will again improve nicely; on March 1 the Vermonter introduces a new schedule. It will leave St Albans at the far north end 28 minutes later, and return 28 minutes earlier. Little things can count for a lot. That morning rush -- get the kids to school, grab a bite, have I got everything? -- will be not so hectic with a 9 am departure. The ride home on a dark night will be a lot less stressful at 9 pm than at 9:30. Next year, another half an hour will be cut from the times in Massachusetts. Still later, more speed-up to come in the Springfield-New Haven segment. These improvements will all build ridership. Waiting expectantly for the extension to Montreal in three years or so.

    The San Joaquins will soon get more cars (rehabbed Comets from New Jersey) to add capacity, and perhaps another frequency.

    Other lines will get tweaked, benefit from CREATE projects around Chicago, and so forth.

    Paul, you'll have to do this again next year. There'll be plenty of changes.

    Thanks again.

    ReplyDelete
  6. Echo the earlier comments on the importance of every-day frequency.
    Not 3 days per week, but not even 5 days per week; every day.

    The other thing- why make fares predictable? Airfares aren't. Megabus and greyhound aren't. Gas prices aren't. If Amtrak takes advantage of those short weeks where all other forms of travel get bogged down and bid up (summer holidays, thanksgiving), you could see a lot of these routes get out of the red on an annual basis.

    The other thing- food and non-ticket revenue has to count bigger than it does. Spirit Airlines is among the most profitable- How? Because they nickel and dime for EVERYTHING, but also there's more options than a normal flight. Normal flight- free drinks, a snack. Alcohol starts at $7 and high-class food for even more. Spirit- nothings free, bottled water and crackers are like $2-4, and there's more expensive fare available. There's no free carry-on bag.

    They capture more of the willingness to pay in the plane, and scrimp on the ticket price, where people make their rational (at-home) decisions. Then people pay $50 for a return-flight carry-on bag of souvenirs and $9 worth of bottled water.

    Meanwhile, Allegiant air is making profits hand-over fist with poor on-time performance (30% late!)? How? They're a home-town airline, they face competition on just 17 of their 203 routes. It was mentioned before, amtrak is big in fly-over states. The big advantage to trains isn't just endpoints, it's the in-betweens.

    Amtrak has acted all too often as if it's a luxury service, with dining cars and sleeper cabins and upsales the model.

    An allegiant model, with some spirit, some megabus, DAILY SERVICE EVERYWHERE, WIFI ON EVERY TRAIN, and INTERMODAL CONNECTIONS is the vision for a profitable amtrak.

    Also, make thruway not require a train connection. Amtrak could make a lot from me alone just by connecting me to the jax airport from Gainesville. The busses aren't full- add a stop worth taking, keep the connection guarantee, and make a lot of money doing it. If it's a regulatory hurdle, privatize it and auction the co-booking, co-branding, and connection guarantee.

    Amtrak's far and away closer to making money than the operating costs of airlines (who don't pay ATC), highways, or intercity bus routes. It's the margin that needs a chance, by and large.

    ReplyDelete