Friday, August 3, 2012

Amtrak food and beverage losses break down

This is taken from the Inspector General's testimony to the House yesterday.

As the second image states, the three quarters of the state supported routes include state support for food and beverage services, and so they aren't a terribly good figure for comparison to the rest of the services.

The poor performance of the long distance trains isn't terribly surprising. They require more staff with the dining cars but couple it to significantly lower passenger loads.

In addition to the recommendations that the Inspector General makes for improving Amtrak, I'd like to see them do some more experimentation in preparation for the states' being entirely on the hook for their routes. As I understand it, Amtrak currently has a surplus of cross country cafe diner/lounges; playing around with these as a bar car or rapid meal service on current corridor trains is a fairly safe way of seeing if they can achieve a higher rate of return on their current services.


  1. Looking this over, there are two issues which I'm not 100% clear on:
    1) How are sleeper (and crew) meals accounted for here? It seems possible that there's some under-accounting of "revenue" from those meals.
    2) How much of the LD trains' costs come down to using two staffed FSCs where one might suffice? For example, you might need some seating room in a second car, but is there a way to achieve the same quality of food service with a "lunch counter" car (many railroads ran these back in the 50s) or a diner-club operation (such as Amtrak has mooted in their PIPs)? Likewise, what could you achieve with a bit more food preparation area in a cafe and a total of 2-3 staff (i.e. a cook, a bartender/register person, and a third staff person to assist the others during peak hours and/or allow them to take breaks)?

    1. Sleepers have some of their revenue assigned to food and beverage, though I'm not aware of how much or what the formula is. Staff food I don't believe is charged to F&B at all, though it is possible.

      I'm not certain how much of the LD train cost comes from two FSCs instead of one, but I don't think it's a terribly high portion. Fundamentally the problem is running a full diner service, even with reduced crew compared to years past, for very few people. LD trains are only averaging about five hundred or so passengers per train across several days travel if memory serves. By comparison, I asked a family member who works at a fairly popular restaurant chain, and they estimate about 1,500-2,000 customers per day at their location. So an LD train is fundamentally unprofitable in my opinion, though steps can be taken to increase cost recovery.

      Corridor trains, on the other hand, really should do remarkably better and using vending machines and the like on trains with relatively low ridership instead of an expensive cafe car should boost cost recovery.

    2. Paul,
      I'm not going to disagree with you there to some extent, though you do seem to highlight the age-old issue of Amtrak being stuck with trains "running short" compared to the past. Heck, even look at the Silvers now vs. in the 70s/80s.

      With that said, there's something rather telling: The NEC managed to improve CR by 23 points over the course of about four years. In that same time, the ridership needle only moved by about 10%. So...something happened there to /dramatically/ improve the picture and, I would submit, put the NEC food service in danger of making money if current trends continue. I know there was a partial overhaul of service there

      In that same time, the needle has basically not moved on the LD side of things (the corridor services are basically meaningless to analyze in this context, since some of the improvement there is likely a result of the subsidized services growing at the expense of those remaining legacy "system" corridor services), something that is rather striking as LD ridership has jumped even more than NEC ridership in that time. I'm not expecting them to have jumped by double digits, but even getting up to about 50% would seem to be a reasonable goal.

      With all of this said, I'd love to see what the 2012 numbers end up looking like, since the transfers seem to have jumped there (they're up about $5.8m vs their budget and about $10m year-over-year). It is entirely possible that we'll see a spike on the LD side of things there...but it's either that or the NEC service is just going to show up as being on a tear.

      Still, your point brings me back to the question of what Amtrak could/should do as far as tinkering with meal service on LD trains. I don't see abolishing it as a winning proposition (especially out west, with the multi-day trains).

      By the way, I do agree with LD trains being, by and large, unprofitable. I do think that Amtrak could, with some equipment additions and the like, get CR up into the 70s on those trains in general...but beyond that seems highly unlikely to impossible.

    3. The real problem is that on-train full diner service is needed when trains run "too long".

      People will be OK with abolishing it the moment the trains are fast enough that they don't run over meal time!

      This means that the solution is really just to run the trains faster. That would take a lot of money, but is certainly viable for the Lake Shore Limited and Capitol Limited; 220 mph service from Chicago to Toledo, and 110 mph service from there to Albany and Pittsburgh, might be sufficient to cut the schedule below the "diner threshold".

      In the shorter term, I expect you'll see an improvement in "transfers" from sleeper revenue to diner revenue, basically because the diners are starting to run full on many meals on many trains. This raises the point that Amtrak is limited by lack of table cars. One diner can handle a diner car plus a table car, if run carefully, but Amtrak hasn't *got* any table cars except on the Auto Train. And we're back to lack of rolling stock. Again.