Passenger Trains > Reuters: Why Amtrak losses mount..hop on the Empire Builder

To see why Amtrak's losses mount, hop on the Empire Builder train
Reuters
ABOARD THE EMPIRE BUILDER
May 28, 2015
by Ernest Scheyder

> Its passengers are mostly silver-haired retirees, oil-field workers
> and a few young families gazing out the windows of Amtrak's
> least-profitable and third-longest line, rumbling from Chicago
> through eight states and on to the American West Coast.
>
> As the Empire Builder snakes along 2,230 miles (3,590 km) on a
> 46-hour journey, it offers insights into the financial problems
> plaguing America's passenger railroad at a time of intensifying
> scrutiny over its aging infrastructure and safety record.
>
> Following Amtrak's May 12 fatal derailment in Philadelphia, the
> 86-year-old line sums up its problems — and its promise. Running
> most of its length near no major highways, the Empire Builder reaches
> into some of America's most rugged, remote and rural regions. A
> coach ticket between Spokane, Washington and Williston, capital of
> North Dakota's oil boom, costs $111, much cheaper than flying.
>
> Its supporters include both Republican and Democrat politicians,
> illustrating how the often-vicious partisan rancor over Amtrak in
> Washington evaporates in sparsely populated areas of the country
> that need it most.
>
> And like many of Amtrak's long-distance routes, it's growing...

So far, if you ignore the (probably intentional) inflammatory headline, not so bad.

But farther down is this tidbit, quoting so-called statistics and
ignoring....as Amtrak's propaganda ministry does when characterizing
the long distance trains as "money-losers" and the Northeast Corridor
as "making money"...the enormous infrastructure cost of the Corridor. 
For the longhauls, the maintenance and capital cost of the rails, substructure,
signal system, bridges...indeed, the entire railroad the trains run on...
are lumped by Amtrak into "operating costs" since they are charged
to Amtrak by the freight railroads.  Amtrak is responsible for its long
haul station and switching facilities and for support functions such as
reservations and ticketing operations, on an equal basis for Corridor and
longhauls.

> In total, long-distance routes carry about 14.6 percent of Amtrak’s
> passengers but soak up about 41 percent of the costs.

Typical "propaganda ministry" wording, too...the long distance trains
"carry" passengers but "soak up" money.


Full story:

To see why Amtrak's losses mount, hop on the Empire Builder train



Edited 1 time(s). Last edit at 05/28/15 10:00 by GenePoon.

"In total, long-distance routes carry about 14.6 percent of Amtrak’s passengers but soak up about 41 percent of the costs." That's comparing doughnuts to dollars. How about comparing dollars to dollars, as in "long distance routes produce X percent of Amtrak's revenues, but soak up Y percent of the costs"? 



Edited 1 time(s). Last edit at 05/28/15 11:07 by CPR_4000.

How about dealing  with Passenger MILES ? ? ?
Why do we equate someone going from Bridgeport to Trenton with someone going from Wisonsin Dells to Williston ?

Difficult to reconcile with the capital-cost black hole that's the Northeast Corridor.

I have done studies on all these LD routes using Amtrak's own numbers and none lose money like Amtrak reports.  Most cover their operating costs.  The main costs for these trains are labor, fuel, track rent and maintenance.  The rest of the costs are small and have no impact on operations.  The big loses come from Amtrak's allocation of it's overhead, most of which would continue even if all these LD trains were discontinued.  And, of course, we all know the NEC doesn't make any money as it's costs ignore the huge infrastructure costs whcih are allocated separately by Congress.

By definition, the accounting should be consistent year after year, using the same principles and breaking out costs the same way. But there's no requirement that the actual method (train miles, revenue miles, revenue per seat miles, avoidable costs, etc) be described. And, many companies combine methods in a consistent way

Amtrak's own overhead costs have always been a difficult item to break out.  How to divide the costs of the substantial law department?  Or Mr Boardman's staff? Or the Chicago diesel facility with its work on corridor trains, long distance trains, etc. The police department.  And, is this consistent year after year, or do the numbers move around as the political winds blow?

Do you use train miles and give the Saluki etc a break and stick it to the Builder / Southwest, etc?  Or revenue miles, or just divide it by the number of trains?

The law department's costs wouldn't go away if the Sunset vanished tomorrow.  Nor would Amtrak's robust congressional relations department.

 

CPR_4000 Wrote:
-------------------------------------------------------
> "In total, long-distance routes carry about 14.6
> percent of Amtrak’s passengers but soak up about
> 41 percent of the costs." That's comparing
> doughnuts to dollars. How about comparing dollars
> to dollars, as in "long distance routes produce X
> percent of Amtrak's revenues, but soak up Y
> percent of the costs"? 

The missing numbers are 25% of revenue (excluding State subsidies).

reading the article myself all i can say is they should look at what the actual costs iss compared to them saying most of the money is spent from the "Profitable" NEC! If the NEC was priofitable, the infrastructure would be much better off! I for one also still don't like how the Republican Congress demand Amtrak to be financially responsible, I mean tney're doing what they can, but I know those jerks won't be satisfied since it seems they still demand Amtrak to make a profit!

joemvcnj Wrote:
-------------------------------------------------------
> How about dealing  with Passenger MILES ? ? ?
> Why do we equate someone going from Bridgeport to
> Trenton with someone going from Wisonsin Dells to
> Williston ?

Sure, but revenue and cost have to be part of the discussion.  A major problem with the economics of LD trains is that coach fares are kept low by the presence of cheap airfares between endpoints.  And unlike airlines who can command a premium for travel between secondary airports, the linear nature of an LD route means that Amtrak can't charge higher fares for secondary stations.  Case in point - for three random dates this summer I compared the cheapest airfares Chicago - Seattle with Amtrak's.  Amtrak's cheapest fare ranged from 55 to 85% of the lowest airfare.  On the same dates, Amtrak's lowest fare Fargo - Spokane was only 32 to 45% of the cheapest airfare on that route.

Regarding cost, consider the amount of equipment needed to generate revenue on LD trains vs. on the NEC.  As currently operated, the Empire Builder requires 10 locomotives and 55 cars, not including spares.  That's a lot of equipment to maintain, let alone depreciate.  On the corridor, 1.5 Acela sets generate essentially the same revenue as all that EB equipment (using January 2013 revenue data).   Add to that the labor costs to staff five EB sets vs. 1.5 Acelas.

I'm a supporter of LD trains and consider them worth the subsidy, but let's not kid ourselves about the economics of operating them.

Dave

joemvcnj Wrote:
-------------------------------------------------------
> How about dealing  with Passenger MILES ? ? ?
> Why do we equate someone going from Bridgeport to
> Trenton with someone going from Wisonsin Dells to
> Williston ?

Because taxpayers who do not use Amtrak which, let's face it, are most taxpayers, do not care if someone is riding 100 miles or 1,000 miles on Amtrak.

That said, I do not have any objection to federal investments in long-distance trains and wish more funding was allocated for that purpose.
 

A more complete article would mention the multi hundred dollar bedroom tariffs in addition to the low coach fares.
And might explore the concept that the Government sometimes does the right thing even if it doesn't "make money".

DavidP Wrote:
-------------------------------------------------------
> joemvcnj Wrote:
> --------------------------------------------------
> -----
> > How about dealing  with Passenger MILES ? ? ?
> > Why do we equate someone going from Bridgeport
> to
> > Trenton with someone going from Wisonsin Dells
> to
> > Williston ?
>
> Sure, but revenue and cost have to be part of the
> discussion.  A major problem with the economics
> of LD trains is that coach fares are kept low by
> the presence of cheap airfares between
> endpoints.  And unlike airlines who can command a

But that is the free market at work.  The train has to find a way to make it or it will be the next stage coach.

> premium for travel between secondary airports, the

Airlines actually are generally abandoning secondary airports and even retrenching from major airports in second tier metros because of the revenue-cost issues reflected in the fares.  US Air travel is ever so gradually being reduced to only going to major US cities (metro areas of 1 million or more, but more preferrably those of 3 million or more).  This is really a huge opportunity for Amtrak against the airlines if they could just figure out the service needed.  The train has a major advantage of being able to serve hundreds of origin-destination combos on a single trip without a major increase in operating costs for that particular departure.  It has to sell this vs. the airliners cruising speed.  Going 600 mph doesn't help so much when the trip requires two layover connections or a 1000 mile detour in the air distance to make the connection.

Look at the types of markets available to Amtrak that the airlines are not serving in a serious manner - Harrisburg/Lancaster/York, South Bend, Dayton, Grand Rapids, Fresno, Wichita, Canton, Peoria, Springfield (all of them - IL, MO, MA, etc.), Mobile, Syracuse, Albany, Charleston, etc.

> linear nature of an LD route means that Amtrak
> can't charge higher fares for secondary
> stations.  Case in point - for three random dates
> this summer I compared the cheapest airfares
> Chicago - Seattle with Amtrak's.  Amtrak's
> cheapest fare ranged from 55 to 85% of the lowest
> airfare.  On the same dates, Amtrak's lowest fare
> Fargo - Spokane was only 32 to 45% of the cheapest
> airfare on that route.
>
> Regarding cost, consider the amount of equipment
> needed to generate revenue on LD trains vs. on the
> NEC.  As currently operated, the Empire Builder
> requires 10 locomotives and 55 cars, not including
> spares.  That's a lot of equipment to maintain,
> let alone depreciate.  On the corridor, 1.5 Acela
> sets generate essentially the same revenue as all
> that EB equipment (using January 2013 revenue
> data).   Add to that the labor costs to staff
> five EB sets vs. 1.5 Acelas.

That is instructive, and the comparison to the NE Regional network is not particularly more favorable.  That operation takes around 36 locomotives and 210 cars.

> I'm a supporter of LD trains and consider them
> worth the subsidy, but let's not kid ourselves
> about the economics of operating them.

Their economics are determined in the main by a (1) skeletal network, (2) slower than 1970 operating times, (3) one-train per day frequency.  A denser, faster, multi-frequency network would generate more revenue at much lower unit costs due to spreading of overhead, reduction in equipment needs and the increase of travel options.  Some examples - the Chicago-Pac NW could have 2 trains each day via the NP and GN (as in days gone by), using one route Chicago-Fargo and consolidated and covering more routes Spokane-PAC NW.  It is likely the costs of this operation would be around $300 million (3 times today) due just to spreading of shared overhead (stations, route management, terminal costs, corporate overhead), but revenue could be closer to $250 million if not more depending on the growth that can be accomodated.  This equation could be further changed by tighter equipment turns by taking trains arriving overnight in Chicago in the AM and dispatching them out in the mid-afternoon (meaning probably 9 sets of equipment to cover 2 runs on a route instead of 5 sets for 1).  A great deal of the costs for every train are the maintenance of equipment and terminal costs - less equipment lowers these costs.

The static thinking of the skeletal 1971 route system is what drives the losses.

abyler Wrote:
-------------------------------------------------------

> Their economics are determined in the main by a
> (1) skeletal network, (2) slower than 1970
> operating times, (3) one-train per day
> frequency.  A denser, faster, multi-frequency
> network would generate more revenue at much lower
> unit costs due to spreading of overhead, reduction
> in equipment needs and the increase of travel
> options.  Some examples - the Chicago-Pac NW
> could have 2 trains each day via the NP and GN (as
> in days gone by), using one route Chicago-Fargo
> and consolidated and covering more routes
> Spokane-PAC NW.  It is likely the costs of this
> operation would be around $300 million (3 times
> today) due just to spreading of shared overhead
> (stations, route management, terminal costs,
> corporate overhead), but revenue could be closer
> to $250 million if not more depending on the
> growth that can be accomodated.  This equation
> could be further changed by tighter equipment
> turns by taking trains arriving overnight in
> Chicago in the AM and dispatching them out in the
> mid-afternoon (meaning probably 9 sets of
> equipment to cover 2 runs on a route instead of 5
> sets for 1).  A great deal of the costs for every
> train are the maintenance of equipment and
> terminal costs - less equipment lowers these
> costs.
>
> The static thinking of the skeletal 1971 route
> system is what drives the losses.

I agree with your points on network size, transit times and frequency improvements likely leading to more revenue.  But the cost to do any of those things is massive.  If railroads are even willing to consider any of the above, their demands for infrastructure investment will be massive - consider UP's demand of nearly a billion dollars just to increase the Sunset's operation from tri-weekly to daily.  Improved running time in particular is costly as it eats up track capacity quickly in an increasingly constrained operating environment.  Add on the cost of more equipment and the total amortized capital costs make financial performance improvement through expansion a questionable proposition. 

Dave

DavidP Wrote:
-------------------------------------------------------
> I agree with your points on network size, transit
> times and frequency improvements likely leading to
> more revenue.  But the cost to do any of those
> things is massive.  If railroads are even willing
> to consider any of the above, their demands for
> infrastructure investment will be massive -
> consider UP's demand of nearly a billion dollars
> just to increase the Sunset's operation from
> tri-weekly to daily.  Improved running time in
> particular is costly as it eats up track capacity
> quickly in an increasingly constrained
> operating environment.  Add on the cost of more
> equipment and the total amortized capital costs
> make financial performance improvement through
> expansion a questionable proposition. 

This is why the Republicans are correct when they assert that we need to change the entire incentive structure regarding Amtrak operation.

If the non-NEC network were operated by the freight railroads on the basis of a refundable tax credit for train miles operated, I think you would see those billion dollar costs for service improvements fall dramatically and also provide a mechanism to make capacity investments.  The tax credit would be set to gaurantee a profit and encourage the needed investment.  The profit would come from from collecting ticket revenue and the train-mile credit.  The incentive would be to operate as many trains for the tax credit as would produce and positive overall revenue stream.  The tax credit could be set up to require service be provided to all towns/cities of a certain size with interline operation coordinated by Amtrak transformed into a joint subsidiary of the railroads.  And since the payments are tax credits, there woudl be no annual fight over appropriations, since there wouldn't be any.  Officially, "spending" would go down and "private investment" go up.

abyler Wrote:
-------------------------------------------------------
> (SNIP)
>
> This is why the Republicans are correct when they
> assert that we need to change the entire incentive
> structure regarding Amtrak operation.
>
> If the non-NEC network were operated by the
> freight railroads on the basis of a refundable tax
> credit for train miles operated, I think you
> would see those billion dollar costs for service
> improvements fall dramatically and also provide a
> mechanism to make capacity investments.  The tax
> credit would be set to gaurantee a profit and
> encourage the needed investment.  The profit
> would come from from collecting ticket revenue and
> the train-mile credit.  The incentive would be to
> operate as many trains for the tax credit as would
> produce and positive overall revenue stream.  The
> tax credit could be set up to require service be
> provided to all towns/cities of a certain size
> with interline operation coordinated by Amtrak
> transformed into a joint subsidiary of the
> railroads.  And since the payments are tax
> credits, there woudl be no annual fight over
> appropriations, since there wouldn't be any. 
> Officially, "spending" would go down and "private
> investment" go up.

That''s a very elegant solution, and one which solves several issues at once.  In some cases, it could even create the oddity of class 1 rails arguing over which wold serve particular end points.  Chicago to LA via Denver and Salt Lake on UP/BNSF, or via ABQ on BNSF (mostly).

 

Lackawanna484 Wrote:
-------------------------------------------------------
> abyler Wrote:
> --------------------------------------------------
> -----
> > (SNIP)
> >
> > This is why the Republicans are correct when
> they
> > assert that we need to change the entire
> incentive
> > structure regarding Amtrak operation.
> >
> > If the non-NEC network were operated by the
> > freight railroads on the basis of a refundable
> tax
> > credit for train miles operated, I think you
> > would see those billion dollar costs for
> service
> > improvements fall dramatically and also provide
> a
> > mechanism to make capacity investments.  The
> tax
> > credit would be set to gaurantee a profit and
> > encourage the needed investment.  The profit
> > would come from from collecting ticket revenue
> and
> > the train-mile credit.  The incentive would be
> to
> > operate as many trains for the tax credit as
> would
> > produce and positive overall revenue stream. 
> The
> > tax credit could be set up to require service
> be
> > provided to all towns/cities of a certain size
> > with interline operation coordinated by Amtrak
> > transformed into a joint subsidiary of the
> > railroads.  And since the payments are tax
> > credits, there woudl be no annual fight over
> > appropriations, since there wouldn't be any. 
> > Officially, "spending" would go down and
> "private
> > investment" go up.
>
> That''s a very elegant solution, and one which
> solves several issues at once.  In some cases, it
> could even create the oddity of class 1 rails
> arguing over which wold serve particular end
> points.  Chicago to LA via Denver and Salt Lake
> on UP/BNSF, or via ABQ on BNSF (mostly).

They could all do it.  The tax credit is "limitless" and not based on endpoints or franchises but on serving all online towns/cities over a certain size and connecting them to major cities.  The theory is that the private money would make as many trains to be operated as could be renumerative.  Say the tax credit is $50 per train mile, the revenue is $40 per train mile, and the costs are $80 per train mile - the operation makes a $10 per train mile profit.  A cross-country train would then make $20,000 net from LA to Chicago.

jfrank39 Wrote:
-------------------------------------------------------
> I have done studies on all these LD routes using
> Amtrak's own numbers and none lose money like
> Amtrak reports.  Most cover their operating
> costs.  The main costs for these trains are
> labor, fuel, track rent and maintenance.  The
> rest of the costs are small and have no impact on
> operations.  The big loses come from Amtrak's
> allocation of it's overhead, most of which would
> continue even if all these LD trains were
> discontinued.

In the 2010 PRIAA reports, Amtrak gave a route cost breakdown.  On the Zephyr, as an example, there was $45M revenue on $100M costs.

Cost breakdown was:

Direct - 51%
T&E/OBS Labor - 31%
Fuel - 12%
Host RR - 7%
Commissions - 1%

Shared - 31%
MOE - 16%
Stations, Commissary, Yards, Ops Management - 15%

Overhead - 18%
G&A, Insurance, Police, Safety - 13%
Sales - 5%

So revenue didn't even cover alleged direct costs of operation.

Labor and MOE are time-dependant because increased running times drive equipment needs and labor hours.  40 hours Chicago to the coast means you get an 8 hour turn on each end and can run a train with 4 sets of equipment instead of 5 or 6.  Using the Zephyr numbers above, that is a variance of $10 million in operating costs.  This is why the Rio Grande route makes no operating cost sense when you have the UP available for the Zephyr.  Not enough people pay Amtrak sufficiently for the scenery in Colorado vs. the costs of operating in Colorado.  There is undoubtedly a market for the Colorado scenery, but it is not an intercity market between Chicago and Oakland.

This is also why its a mistake to forgo the 100 mph speed the Santa Fe used to have on its mainline.  Losing that costs hours of schedule which adds millions in annual costs because its the difference between an 8 hour turn in Chicago and a 26 hour turn.

abyler Wrote:
-------------------------------------------------------
>(snip)
> This is also why its a mistake to forgo the 100
> mph speed the Santa Fe used to have on its
> mainline.  Losing that costs hours of schedule
> which adds millions in annual costs because its
> the difference between an 8 hour turn in Chicago
> and a 26 hour turn.

That also gets back to the return on investment question.  How much will it cost in new investment and annual maintenance to shave 10 hours, 12 hours, etc off the transcon etc trip? And will that be sufficiently rewarded in crew cost improvements and passenger revenue?

Amtrak is likely seeing the problem as many Acela travelers realize that saving a half hour of their time is not always worth a $200 premium over coach.

abyler Wrote:
-------------------------------------------------------
> jfrank39 Wrote:
> --------------------------------------------------
> -----
> > I have done studies on all these LD routes
> using
> > Amtrak's own numbers and none lose money like
> > Amtrak reports.  Most cover their operating
> > costs.  The main costs for these trains are
> > labor, fuel, track rent and maintenance.  The
> > rest of the costs are small and have no impact
> on
> > operations.  The big loses come from Amtrak's
> > allocation of it's overhead, most of which
> would
> > continue even if all these LD trains were
> > discontinued.
>
> In the 2010 PRIAA reports, Amtrak gave a route
> cost breakdown.  On the Zephyr, as an example,
> there was $45M revenue on $100M costs.
>
> Cost breakdown was:
>

> This is also why its a mistake to forgo the 100
> mph speed the Santa Fe used to have on its
> mainline.  Losing that costs hours of schedule
> which adds millions in annual costs because its
> the difference between an 8 hour turn in Chicago
> and a 26 hour turn.

The 2014 numbers were 55m in revenue and 86m in operating costs for an operating loss of 31m.  The overall loss per Amtrak for this train was 62.5m.  It is Amtrak's biggest looser.  The next biggest are the EB and the SWC.  Time is money.  The longer these trains take to negotiate their route the more it costs.  The biggest costs for all of them is labor.  Followed by maintenance, fuel and rent.  The revenue side problem is the coach fares.  These are often less than Greyhound Bus fares.  Buses and airlines do not have a rolling diner and sight seer lounge car.  Rail coach fares should be at a premuim compared to bus and air. Sleeping car fares are high, too high.  Amtrak needs to only offer one meal with the sleeper ticket price as riding on a train all day one doesn't need three full meals a day.  For breakfast, Amtrak should just offer a standard breakfast for free to all passengers as most motels and hotels do now, both here and in Europe. Perhaps they could set it up as some kind of buffet to reduce labor requirements.  As for labor costs, you are correct there, as the faster a train completes it's route, the lower the labor costs.  Amtrak also needs to look at fewer on board personnel.  European train have very skeletal crews.  If you are going to use these trains as the 'full employment' act, then don't gripe about the subsidies.  There are probably many other improvements that could be made, but Amtrak isn't interested in promoting nor improving and growing the business.  All they care about is the NEC.    

If Amtrak had as many Superliner sleepers as possible - including deluxe sleepers that are currently only on the Auto Train - it would be interesting to see how this would change the economics. Maybe the Canadians have something here. Two coaches and what 18+ sleepers for the premium rate? Then add in VIA's Prestigae class revenue, which they are selling space for!

Booking sleepers on LD Western trains in July (when I first looked into this in late May) was putting me in Bedroom B or A already! Some trains just one room with letter A available. I'd rather take a roomette. But then comes the interesting time in winter - how to keep the sleepers full. Well, maintenance on some, others are put on Florida trains perhaps.

It is no seret that Amtrak needs sleepers in peak periods - Empire Builder, Southwest Chief, CA Zephyr, Coast Starlight. And EVEN some of these trains are running with THREE sleepers. But a premium rate can be had for bedrooms and roomettes that coach doesn't offer. The business class seating on the Coast Starlight is an interesting concept to boost "coach" revenue. Not sure if it will work.

But basic economics would call for Amtrak to have more sleepers (bedrooms - deluxe sleepers) in the peak season to get more revenue per train. Amtrak can't even move sleepers off the Auto Train because in summer time (when one would think it would be off peak for that train) - this train actually carries a lot of families to Disney World. 

European countries were trying to get cars that could be converted from coaches to sleepers and vice versa to meet demand. But Amtrak is not making investment in its Superliner fleet (i.e. replacements). It can't even figure out how to do "upsells" for existing sleeper or coach patronage.