Amtrak accounting

[Tadman]

I have had several jobs where we were supposed to allocate our time into small intervals, often 6 minutes. What happened universally was that everyone just made up the numbers to add up to 8 hours a day. No one ever questioned the numbers except if you made a mistake and the numbers didn't add up to 8 hours. Then, in my college days I worked for a while at the main post office in NY. You got a ticket for each tray you sorted, and it wasn't hard to meet the quota. Some people had trouble meeting the quota. Usually they just swiped tickets and handed them in as their own. The most inept guys just took trays of unsorted mail and labeled them as sorted to some random destination. I suspect that went on enough so eventually management scrapped the system.
I guess the point of this is that overly onerous allocation systems will be evaded. A company needs an allocation system that will actually reflect reality. I would also suspect too many significant figures in any published allocation numbers.

Totally agreed.

I know a lot of lawyers (I am one but do not practice) and they do not allocate their time by 6 or 15 minutes, more like by the hour. Ergo, if they call their wife or hit the bathroom, it's on someone's tab. When I worked at a factory in college, it was much like the postal example above. I need to allocate 8 hours of my day... well I walked past projects 1, 3, 7, 24, and 14, so that goes on my time sheet. We all knew it was bunkos and so did management, but it was a "pick your battles" situation and as long as one project wasn't totally out of whack, nobody got in trouble. Problem was when one 40 hour project turned into 200 hours, the big boss was on the floor asking who screwed up, and let him see the rework. "uh, there is none, we just kept shovelling hours to project 24 cos we're idiots"...

Which leads back to Amtrak: It's good that they're keeping track of operating profits. But it would be really nice to know the real cost of running a 79mph slot with hourly stops, not just the "we do it because we have to" rate that BN gives Amtrak. It would also be nice to know the real cost of running an Acela or NER. It could take a long time and there are probably three different ways to do it.

For example, because the NEC would still run with or without Acela, should any NEC costs below-rail go to Acela? They'd sure still run regionals and NJT, and did for 30 years before Acela showed up. I'm not advocating this line of logic, just pointing out that it is there and could be argued, leading to a long process that likely not everybody is happy with.

[eolesen]

That's a good point. How do you properly allocate IT costs? Per ticket sold? Train mile traveled? Passenger-mile?

Ticket sold strikes me as a good idea, because most of the IT costs are in processing the sale. But we don't know that's how they do it. They might do it by train mile, which would make no sense but would help management support their claim that the LD is not profitable.

Allocation model for airline IT depends on the system.

Core selling systems like the website, ticketing, inventory control, availability to offline systems are usually measured on a per transaction cost.

Systems required to support each departure (onboard wifi, boarding control) are costed out based on passengers boarded because you could have a one to many relationship depending on how many trains a single traveler was taking.

Core operations management systems like flight planning, fuel, and weight/balance are per departure because again it's tied to the movement and not the number of passengers onboard.

[east point]

Had much thought about reservations costs. On line costs would seem to be same for any single leg trip. Where it gets complicated is how long a potential passenger takes for him to complete the transaction. Then you have the persons htat need or have to contact an agent to complete a reservation or modify it. A smart system could average the time spent for each route and use that number for allocating IT for that train.
Station costs are of course different.

[Arlington]

People--staff & maintenance costs are the only *cost* large enough to swing routes in/out of profitability via scale per passenger.

Right of way costs for LDs are significant (in the $200m/yr -$400m/yr if I remember correctly), but all indications are that this is a fair price or even a bargain. The problem is that if you double the number of LDS the right-of-way costs would double getting you no economies of scale.

Equipment costs also double if you were to double the LDs.

The problem for the LDS is that lack of demand keeps them from being scaled profitably.

Doubling frequencies would double every major cost (station costs along the route are minor) but revenues would almost certainly less than double--increasing losses.

[Tadman]

The other problem with LD's is the erratic demand. Acela sees heavy use 5-6 days a week if not 7. Same with corridors. Another train or more cars would probably be filled regularly and see an operating profit.

The LD's see heavy bookings on weekends, holidays, summer, but not winter or mid-week. If you double an empty train on a January Wednesday, what does that get you? Two empty trains, both sucking up track and equipment costs. No creative accounting could hide that.

I've advocated in the past for a few leased Canadian cars in winter to come to the Silvers, where capacity is needed. You could also attach a few superliner sleepers at Washington. Or run a third train. But that's another ball of wax and not under the accounting thread anyway.

[JoeG]

It looks as if, in the end, trying to find big allocation mistakes may just be rearranging the deck chairs on the Titanic. Amtrak just does not have enough money. Its upper management often seems unable to master the job of running a railroad. For some reason many people want Amtrak to lose less money, or to make a profit. It can't. Other modes of passenger transport lose money all over the world. In this country no one expects highways or commuter lines to make money. Last I saw, airlines, over their history lost money despite large subsidies for airport, air traffic control, various tax breaks, etc.

Sure, individual trains or maybe some lines could break even, but not the system as a whole. Amtrak tries to cut its way to solvency. All that does is make their service and utility less and less. They fail to make reasonable deals with their host railroads. As a result, their trains, especially the LD routes, are so late they can't be counted on, especially if some kind of connection is planned. Part of Amtrak's expenses should be to pay their hosts enough so Amtrak trains are treated with some respect. The rest of the developed world manages this stuff better than we do. At least one reason is they don't expect their passenger transportation to be profitable. It's a public service like police and firemen and roads.For that matter, the rest of the developed world provides its citizens with tax-revenue-funded healthcare. We of course provide neither free healthcare nor good transportation. I'm glad I'm old enough to get Medicare. Oh, and should I become unable to drive, very little help is available. My county has announced a 30% cut in bus service and says that in 2 years it will have no money for any bus service unless new funding is provided. Of course the legislature is proud of its low tax rates and may not pony up. I guess when I can't drive anymore I could limp to the curb and stick out my thumb. If I lived in Germany or even England I would have none of these problems. If I needed to go anywhere I could go using public transportation, including trains.

[prokowave]

People--staff & maintenance costs are the only *cost* large enough to swing routes in/out of profitability via scale per passenger.

Right of way costs for LDs are significant (in the $200m/yr -$400m/yr if I remember correctly), but all indications are that this is a fair price or even a bargain. The problem is that if you double the number of LDS the right-of-way costs would double getting you no economies of scale.

Equipment costs also double if you were to double the LDs.

The problem for the LDS is that lack of demand keeps them from being scaled profitably.

Doubling frequencies would double every major cost (station costs along the route are minor) but revenues would almost certainly less than double--increasing losses.

I have to disagree with most of your points here. Although I do like your quote "Trying to solve congestion by making roadways wider is like trying to solve obesity by buying bigger pants" as it demonstrates induced demand. More service equates to a greater than commensurate increase in ridership and airlines know this well - just look at Delta's operation in Atlanta. I strongly believe that making most long distance routes double daily or increasing the 3/weekly service to daily would lead to a greater than doubling of ridership and revenue.

To illustrate, look at the long distance routes that pass through Ohio in the middle of the night. If you added a service that was offset by 12 hours, you'd have a mid-day service that would see much higher ridership. Same thing with the Crescent passing through CLT at 2am both directions. To give another example from my area of New Orleans, we have 3 long distance routes, but none of them connect to each other without an overnight stay. If you double services, you'd have far more connecting opportunities - BHM-HOU; SAT-ATL, MEM-LFT, etc.

As for costs, the staffing costs per mile would decrease as you grow, because you'd be hiring younger employees with less seniority and the average would shift lower. You'd also have more flexibility to more efficiently schedule crew - for example instead of working 4 hours outbound and then staying overnight at a hotel to work the next inbound, you could have that crew work an inbound that would be scheduled later the same day.

The LD's see heavy bookings on weekends, holidays, summer, but not winter or mid-week. If you double an empty train on a January Wednesday, what does that get you? Two empty trains, both sucking up track and equipment costs. No creative accounting could hide that.

Low frequency and poor timings as I mention above are okay for leisure travelers, but if you increase the frequency, you'll attract more business travel which would fill those weekday and winter, etc. seats. And at higher last minute fares too.

[electricron]

I have to disagree with most of your points here. Although I do like your quote "Trying to solve congestion by making roadways wider is like trying to solve obesity by buying bigger pants" as it demonstrates induced demand. More service equates to a greater than commensurate increase in ridership and airlines know this well - just look at Delta's operation in Atlanta. I strongly believe that making most long distance routes double daily or increasing the 3/weekly service to daily would lead to a greater than doubling of ridership and revenue.

To illustrate, look at the long distance routes that pass through Ohio in the middle of the night. If you added a service that was offset by 12 hours, you'd have a mid-day service that would see much higher ridership. Same thing with the Crescent passing through CLT at 2am both directions. To give another example from my area of New Orleans, we have 3 long distance routes, but none of them connect to each other without an overnight stay. If you double services, you'd have far more connecting opportunities - BHM-HOU; SAT-ATL, MEM-LFT, etc.

As for costs, the staffing costs per mile would decrease as you grow, because you'd be hiring younger employees with less seniority and the average would shift lower. You'd also have more flexibility to more efficiently schedule crew - for example instead of working 4 hours outbound and then staying overnight at a hotel to work the next inbound, you could have that crew work an inbound that would be scheduled later the same day.

Low frequency and poor timings as I mention above are okay for leisure travelers, but if you increase the frequency, you'll attract more business travel which would fill those weekday and winter, etc. seats. And at higher last minute fares too.

Adding more trains is the same as adding lanes or using larger pants. To have induced demand there must be a demand first.
People don’t buy larger pants until they can’t fit into their existing pants, and traffic planners do not plan highway expansions until there traffic jams already on the highway.

LD train crews don’t run a train for 4 hours at a time, which would allow them to return on the same day. They run the trains for 8 hours at least, and can’t run a return train later that day, they are required off time by federal regulations. That is why they wait for the next day.. Of course, Amtrak could double their number of crew change stations, doubling management costs and crews. But that will not lower costs at all.

Have you ever taken the time to consider what the Lake Shore Limited or Capital Limit schedules would be displaced by 12 hours?
Let’s look at both trains, limiting the stations to the two terminating stations and one intermediate station (Cleveland).
LSL WB
New York City 3:40 am
Cleveland 3:33 pm
Chicago 9:50 pm
LSL EB
Chicago 9:30 am
Cleveland 5:38 pm
New York City 6:35 am

CL WB
D.C. 4:05 am
Cleveland 2:53 pm
Chicago 8:45 pm
CL EB
Chicago 6:40 am
Cleveland 1:45 pm
D.C.1:05 am

I would suggest that this is a worse schedule than the existing schedule.
Maybe both trains schedules could be changed into a very long day train with coaches only; with the removal of sleeper cars from it. Move the CL WB to a 7:05 am departure from New York, 5:53 pm in Cleveland, and a 11:45 pm in Chicago. Keep the EB to where it was moved earlier.
I don’t think a coach only day trip can be done with the LSL in either direction. It just too long for a 6 am departure with a before midnight arrival at the terminating stations.

[R36 Combine Coach]

Over at another site at which I now only read, their regulars enjoy saying how Amtrak accounting is non-GAAP compliant and is "fraudulent" (LD'are break even to profitable).

"Creative accounting" or "Enron accounting" (or "Madoff accounting" *) is the term, making one division look profitable on paper by transferring money or cash flow only on the books to give the impression.

(*Refer to the Madoff Trustee's reports on the investigation and other files. The "phantom profits" from the investor advisory brokerage was used to boost up the other wing of the firm, the "legitimate" commercial trading floor.)

[Arlington]

The S curve is what airlines call it when increased frequency *captures* disproportionate market share--where frequent service functions like faster service. It isn't so much inducing demand as better capturing demand that's already proven.

This works with business trips where trip duration and workday convenience matter.

When Anderson proposes corridor service on the busy subsegments (MSP-CHI but not the whole Empire Builder, BUF-NYP/CHI but not the end-to-end LSL) he is doing frequency the correct way.

[prokowave]

Adding more trains is the same as adding lanes or using larger pants. To have induced demand there must be a demand first.
People don’t buy larger pants until they can’t fit into their existing pants, and traffic planners do not plan highway expansions until there traffic jams already on the highway.

There is enormous demand for intercity travel, it's just that most of it uses planes, buses, and cars since Amtrak does not offer enough schedule flexibility and speed.

LD train crews don’t run a train for 4 hours at a time, which would allow them to return on the same day. They run the trains for 8 hours at least, and can’t run a return train later that day, they are required off time by federal regulations. That is why they wait for the next day.. Of course, Amtrak could double their number of crew change stations, doubling management costs and crews. But that will not lower costs at all.

I don't the specifics of crew scheduling and rules, but I'm sure increase frequences could allow for better scheduling efficiency. In some cases it might allow for more crew bases and less need for overnight stays.

Have you ever taken the time to consider what the Lake Shore Limited or Capital Limit schedules would be displaced by 12 hours?
Let’s look at both trains, limiting the stations to the two terminating stations and one intermediate station (Cleveland).
LSL WB
New York City 3:40 am
Cleveland 3:33 pm
Chicago 9:50 pm
LSL EB
Chicago 9:30 am
Cleveland 5:38 pm
New York City 6:35 am

CL WB
D.C. 4:05 am
Cleveland 2:53 pm
Chicago 8:45 pm
CL EB
Chicago 6:40 am
Cleveland 1:45 pm
D.C.1:05 am

I would suggest that this is a worse schedule than the existing schedule.
Maybe both trains schedules could be changed into a very long day train with coaches only; with the removal of sleeper cars from it. Move the CL WB to a 7:05 am departure from New York, 5:53 pm in Cleveland, and a 11:45 pm in Chicago. Keep the EB to where it was moved earlier.
I don’t think a coach only day trip can be done with the LSL in either direction. It just too long for a 6 am departure with a before midnight arrival at the terminating stations.

I would suggest that to the CLE, PIT, CIN, etc travelers the schedule is 10 times better than what is run currently. Obviously you'd want to adjust things slightly to avoid overlap and maximize connections, but this would much better serve the cities in the middle of the route, which is critical to the success of a long-distance service.

I'm not advocating for a complete change of what already exists, since no long distance schedule is going to be perfect for everyone. But offering two roundtrips would greatly increase the usefulness of a given route. Or maybe a second train that does a morning departure from Chicago, splits at Cleveland and goes overnight to NYC and late night arrival in WAS. In an ideal world we could also get some investment in more capacity and higher speeds so that more long day trips would be possible.

I could spend all day discussing hypothetical schedules, but my main point is that there is not enough discussion on how to attract greater ridership to the long distance routes and to grow those services.

[electricron]

I could spend all day discussing hypothetical schedules, but my main point is that there is not enough discussion on how to attract greater ridership to the long distance routes and to grow those services.

Whatever needs to be done to satisfy you will not work until all LD trains actually ran on time. Not on time at the last station, where Amtrak pads the schedule by a half hour to full hour on every route. But on time at every station along the route. The best place to grow ridership the fastest is at the intermediate stations, and as long as the trains average more than an hour late, growth is not going to happen.

[Suburban Station]

Over at another site at which I now only read, their regulars enjoy saying how Amtrak accounting is non-GAAP compliant and is "fraudulent" (LD'are break even to profitable).

"Creative accounting" or "Enron accounting" (or "Madoff accounting" *) is the term, making one division look profitable on paper by transferring money or cash flow only on the books to give the impression.

enron transferred debt to subsidiaries and kept it off the consolidated P&L, it didn't simply transfer money between divisions or even subsidiaries which normally would come out in the consolidated P&L.

[David Benton]

And then there are multinationals, transferring profits/costs between countries to make the most profit in the country with the lowest tax rate.

[WashingtonPark]

I could spend all day discussing hypothetical schedules, but my main point is that there is not enough discussion on how to attract greater ridership to the long distance routes and to grow those services.

Whatever needs to be done to satisfy you will not work until all LD trains actually ran on time. Not on time at the last station, where Amtrak pads the schedule by a half hour to full hour on every route. But on time at every station along the route. The best place to grow ridership the fastest is at the intermediate stations, and as long as the trains average more than an hour late, growth is not going to happen.

That's an excellent point. Riding the Vermonter last week at one point we were 45 minutes late. The conductor kept on assuring everybody there was plenty of padding in the schedule and they would arrive Washington on time. We arrived only 4 minutes late in Philadelphia, which was great for us but stinks for the people north of New York who are arriving and departing late at all these stations. As long as management convinces itself that all is fine as long as the train arrives within thirty minutes of it's scheduled time at the terminal station, Amtrak is not viable as a long distance alternative to get somewhere.