Sleeping Car Revenues are Pure Gravy for Amtrak

Sleeping Car Service

§         Sleeping car service on Amtrak’s long distance trains generates an incremental operating profit of at least $40 million (based on Fiscal 2004 data, see Table 1).

§         Using the same detailed cost data employed by the USDOT Inspector General, we calculated that the incremental revenue from sleeping car service exceeded the incremental cost on every route.  The gross margin on sleeping car service averaged 32% for the system; it was 48% on the Lake Shore Limited (Table 2).

§         Sleeping car service performs an important transportation function. Sleeping cars carried people over 630 million miles in each of the last two years.  For perspective, sleeping cars carry people more miles than Amtrak’s premium high speed Acela Express & Metroliner services in the Northeast – 22% more in FY 2004 and 51% more in FY 2005 (Table 3).

§         The federal cost of moving one person one mile in a sleeping car is less than it is in coach.  In fact, the cost of carrying a person one mile on a full service train that provides passengers with a range of attractive choices including sleeper, lounge, dining car and checked baggage service is actually lower than what the IG projects it would be on a bare bones train offering only coaches without any amenities (Table 4).

§         The DOT IG’s analysis further demonstrates that the federal cost “per passenger” (rather than passenger-mile) reflects only distance traveled, not class of service.  That is, the federal cost of moving one passenger in coach and another passenger in sleeper on trips of equal length is roughly equal—slightly lower in sleeper!

§         The federal cost of operating a national network of trains with just coaches and no food or other amenities would be far greater than the cost of continuing the current level of service.  Under any scenario, the goal of break-even food service is unrealistic and ignores the fact that, worldwide on all transportation, when food is provided it is to enhance overall revenues, not to serve as its own profit center.

§         In order to realize the projected “savings,” the IG makes the unrealistic assumption that coach passengers would continue to patronize a coach only service. This assumption is not credible because of the large number of coach passengers who use coach to make very long trips.  Indeed coach passengers outnumber sleeper passengers in every distance category.  

§         Passengers making long distance trips in coach are “high value” customers. In FY 2004 (the year the IG analyzed), coach passengers  traveling 400 miles or more accounted for 81% of total coach revenue; coach passengers traveling 800 miles or more accounted for 54% of revenue. It is highly likely that most if not all of these passengers would not use a bare bones coach only service. The $170 million of coach revenue that a bare bones coach only service would put “at risk” would more than offset even the high end “savings” the IG estimates its recommendations would achieve. (Table 5)

§         Reducing long distance trains to coach only service would dramatically reduce public utility, volume of use, and revenue while simultaneously increasing both cost per passenger mile and per passenger, making the service inefficient, irrelevant to the traveling public and unjustifiable.

§         The proper goal is to drive efficiency. That means reducing federal cost per passenger mile not total federal cost. Volume drives efficiency by increasing revenue and lowering unit costs. USDOT IG has turned these important priorities upside down.

Table 1

Sleeping Car Incremental Revenue and Incremental Cost in FY 2004

Notes

1 – Sleeper transportation & accommodation revenue (this includes California Zephyr sleeper revenue, almost all of which RPS incorrectly identified as Custom & club accommodation revenue).

2 – Direct sleeper attendant wages plus a proportionate share of OBS support (sleeper at percent of sleeper, dining car and coach labor).

3 – 50% of total linen and laundry cost (balance to food service, see also note 12).

4 – Total route fuel cost divided by total car miles times number of sleeping car miles. Cost per car mile averaged $.18 for the system but varied wildly from a low of $.08 for the Zephyr to a high of $.36 for the City of New Orleans.

5 – Annual turnaround cost divided by annual trips (corrected for Builder Portland section and Lake Shore Boston-Albany segment), divided by number of cars per trip (car miles divided by train miles). Cost per car averaged $268 for system but ranged wildly from a low of $151 on the Eagle to a high of $711 on the Sunset.

6 -- Assumes sleeper elimination would allow all trains to operate with only one locomotive.  This assumption may understate the incremental profitability of sleepers, since Amtrak says most of the trains not already running with a single locomotive could not safely operate with only one locomotive.   [Locomotives per train are calculated by dividing locomotive miles by train miles. Cost per locomotive determined by dividing annual turnaround cost by number of unit trips (corrected frequency multiplied by number of units). Savings calculated by multiplying cost per unit by units saved by corrected annual frequency.]

7 – Percent of sleeping car miles times total car program maintenance.

8 – Percent of sleeping car miles times total car heavy maintenance.

9 – Percent of sleeping car miles times total car wreck and accident.

10 – Percent of sleeping car passengers multiplied by total redcap and ticket (both route & shared station) cost.

11 – Percent of sleeping car revenue multiplied by total cost of commissions and inconvenience.

12 – Sleeping car transfer as percent of total F&B revenue multiplied by total cost of food, liquor & tobacco and ½ of linen (other half allocated entirely to sleeping cars).

Table 2

Net Sleeping Contribution by Route

Table 3

Passenger Miles in Sleepers and Premium NEC Trains

Table 4

Federal Cost per Passenger mile Based on USDOT IG Analysis

Table 5

Coach Revenue by Distance Traveled

FY 2004

GLC 12/18/06