I have a map of the Chicago Rock Island and Pacific by my desk in my office. Sometimes it is really interesting to sit and look at all of its routes and where they ended up--it really is amazing how many separate Railroads now operate the former Rock. Anyway, here is my question: You have the Iowa Interstate and the KYLE Railroad. I assume the portion between Omaha and Belleville Kansas was abandoned. Would the two railroads (IAIS & KYLE) be more viable today had they/someone just tried to run it as one railroad from Chicago to Colorado Springs instead of two independent sections? Given that Rail America CEO recently indicated that he considered the KYLE on of their better and more promissing properties and the success of the Iowa Interstate, I am left to wonder, would they function more profitably together?
It is a question I often wonder when I see two--or more--short lines abutting one another: wouldn't they work more efficiently and effectively as one Railroad?
I understand there is going to be some obvious answers to illuminate the flaws of my logic--but, if I knew those answers . . .
Gabe
Gabe -- this is sort of hard to explain, so bear with me. There's two questions you are asking, the first about operating efficiencies and the second about managerial efficiencies.
First, operational efficiencies. Short lines want to be gathering/distributing entities with the least line-haul possible, not line-haul roads, because at low carload volumes, railroad costs are inelastic with volume. Though the handling charge or interline settlement the short line might get for a longer haul will go up, it won't go up as fast as its costs go up. That's because when the cars are on the short line, they must bear 100% of the cost of the short line's infrastructure. That's a very high burden per car because certain individual features of the short line's railway plant, such as bridges, grade-crossing signals, drainage, brush clearing, tunnels, locomotive shop, tools, hi-rail truck, each cost a virtually identical amount to maintain and renew whether the line carries 100 cars a year or 1,000,000 cars a year. Once the short line cars are on the Class 1, then their costs become shared with thousands of other cars and their individual portion is very low, thus the per-car price for each car can be reduced and the business can be won.
It makes sense for a short line to try to become a line-haul road only if it has a very high volume, so high that it in effect it can apportion the costs of transportation onto each of its cars as cheaply as the Class 1 these cars would travel on otherwise. Otherwise the short line is literally throwing money away because it can, in effect, "rent" the Class 1's plant at a lower cost than paying for its own.
But even if the short line gains the sort of volume where it can rent line-of-road from itself as cheaply as it can from the Class 1, it has to be careful. Suppose the short line buys and fixes up some leftover line-haul road that parallels the Class 1 it connects to, in order to take its cars to some big city terminal that the Class 1 was delivering to. When it does so, it is taking away those cars from the Class 1, which now has to spread its fixed costs onto fewer cars. If the short line can take away 100% of the cars it previously interchanged to the Class 1, and put them all onto its "found line-haul," no problem for the short line. But if some of its cars remain on the Class 1 (they're going only part way toward the terminal), suddenly the cost burden on those cars shoots up, and when carried into their freight rate might price them out of the market. The short line can find itself in a less-than-zero-sum game, with total costs post-acquisition of its line-haul plant much higher. (This is why duplicate trackage is such a bad idea; the idea is to put all the cars onto one line and spread the inelastic costs onto as many cars as possible.)
There's a frequent exception to the above, which consists of -- how can I say this nicely -- "using up the plant." In other words, the short line acquires a lengthy piece of track and pushes its interchange with the Class 1 much farther away than the nearest possible interchange point. This can persist so long as the short line isn't actually renewing the physical plant. But once the ties and rail are used up in 20 or so years, the game is over. There's a couple of other exceptions which are "take or pays" that are in effect carved-out sections of a line-haul railroad that are de facto contract operators for a line-haul. Those are unique and unlikely to be repeated.
The example you gave, the Rock Island from Chicago to Denver, supposes the cars even would want to move that way today. But do they? What business is there that originates anywhere en route that wants to go either to Chicago or to Denver? There wasn't much historically; it was mostly end-point to end-point and overhead. The Rock Island had a nice business to/from Denver back in the regulated era when carload freight moved at the same rate between rate-territory boundaries regardless of which road they were on. The instant freight was deregulated, the high rates enforced by the ICC to protect the weak roads collapsed, and freight that once could move profitably via the Denver gateway on a D&RGW-Rock Island or D&RGW-BN routing now could not, and shifted to an SP-UP or Santa Fe routing to gain the cost-sharing of dense traffic.
It would have been possible for the entire Rock Island between Chicago and Denver to land post-1980 under one owner. But, that owner, even though it owned the entire route, would have -- if he knew what he was doing -- done everything possible to get rid of its carloads to the first Class 1 interchange possible, even if the destination for those cars was actually at the other end of its own railroad. Trying to haul those cars from one end to another of its railroad would have just have used up the physical plant and returned a revenue insufficient to pay the costs of maintaining the physical plant, because there were not then or now nearly enough cars moving end-to-end to pay the bills for all that track. A number of sprawling short lines and regionals have tried to do this anyway (that is, presume to be a line haul but without the traffic base to support it). Whether they chose that path through ego or lack of understanding of railroad economics I do not know, but those managements are not with us now.
There are some examples where a regional railroad to outward appearance seems to be in the line-haul business, accepting cars in interchange at one end and hauling them to interchange at the other. But if you look closely the volumes it is accepting in interchange are very low and are usually "tonnage fill" on a train that had to be run anyway, so the labor costs are spread thinner. If it tries to take more, it will be losing money instead of making money because its competing with high costs per car against a Class 1 line-haul with low costs per car, and the price at which the traffic will move on the regional is less than its costs -- a case of the more business you get, the more money you lose.
The second question you ask is managerial. The trend in the rail industry has always been toward consolidation because it reduces overhead. But for a short line it may not accomplish very much to weld two or more lines together, because so much of what the short line management and support staff does is irreducible. In other words, the short line business has already cut management and overhead to an extremely low level that consolidating two end-to-end short lines might only trim 1 or 2 jobs, if that.
RWM
This could be one of the most informative posts I have ever read.
Thanks,
Concur. Thank you.
To grossly oversimplify this and reduce it to a "popular" level, it seems like the railroad version of young people when they move in to live together as an unmarried couple - "But Dad, two can live as cheaply as one !' Well, maybe so, and maybe not quite.
The key would be the relative proportions and lengths of each of the traffic routes, and the inelastic (fixed) costs, and the elastic (variable) costs of each. Specifically, how much of the Class I's costs are variable with the diversion of some percentage of its traffic to the short line. If instead a good proportion of the Class I's costs are inelastic and the amount of traffic diverted to the short line is not a huge proportion of the Class I's traffic base, then the remaining traffic on the Class I would not necessarily suffer a huge increase in its share of the inelastic costs. For example, if 10 % of the Class I's traffic goes to the shortline, the increase in the inelastic cost burden on the remaining traffic on it would be about 11 % (100% / 0.90) - not 20 % or 50 %, etc.
And as postulated, the short line's costs are inelastic in absolute amount - the tunnel or bridge doesn't cost any more each year because of more traffic over it, hence their incremental cost to the short line for the additional traffic is near zero. So if the short lines remaining traffic on the Class I route can tolerate an 11 % increase, the short line may well make out. If it can't, then the short line has shot itself in the foot just as RWM warns about.
Finally, viewing both lines together as a total system: Once the short line is in existence, its inelastic cost meter is running full time anyway, as is the Class I's. The sum total of those inelastic costs in absolute $ amount each year is a constant figure - that's the definition of inelastic cost. So those costs won't change for either line regardless of how much traffic is routed on each (within reasonable limits of capacity and others, of course). John Kneiling used to refer to this as the "station-painting" fallacy of the ICC's costing system - just because twice as many trains and cars are run, does not result in the annual cost of painting the station doubling, too.
So the optimum traffic balance between the Class I and the short line - for the purpose of minmizing the total costs of these 2 lines for their aggregate traffic - will be found by piling cars onto the route with the lowest elastic (variable) costs(presumably the Class I) until they rise to the point where those variable costs on equal the variable costs of the other line (presumably the short line). That's a truism as the definition of the least cost of the total system (see Calcullus 101 or Economics 201). Any other solution or traffic allocation is by definition sub-optimum = total costs higher than they have to be.
That's why RWM's special examples of the short line carrying the traffic work. For the "tonnage fill", the incremental cost of a few more cars on a train and lcomotive set that has to run anyway is near zero, so there's a clear advantage over the Class I with a fully-allocated cost structure. But that's only good for that 1 train at a time or that 1 train on the schedule each day or so, not as a long-term or sustainable strategy (like having additional passengers with you on a taxi ride). And the "use up the infrastructure" approach (also known as "cannibalization" or "throwing the furniture on the fire to keep warm" or "eating your seed corn to survive", etc.) is likewise merely a short-term approach to artificially lowering the apparent short-term cost by ignoring the replacement cost component of the inelastic costs. (Like running a car until the tires are bald - then walking away from it.)
Interesting intellectual exercise. Again, thanks to both of you.
- Paul North.
EDIT at 2:35 PM: One other important aspect of this discussion that is not mentioned and that I overlooked above is that it is implicity based on the price / rate/ value of service being the same for either the Class I or the shortline between the service points. To the extent that is not the case and the rate or perceived value by the shipper may be different - such as by or from more frequent service (such as daily on the Class I), or more personal and customized service (matching the plant's schedule by the short line), etc. - then the profit or margin on each traffic route would be more relevant than merely the difference in costs (which is of course a negative component of the profit/ margin calculation). The above analysis would then have to be revised accordingly, which would make it more complex, and difficult to explain this simply - these things are always extremely sensitive to the specific fact contexts.
- PDN.
Our community is FREE to join. To participate you must either login or register for an account.