In my past life I worked in LTL trucking, as a manager of pricing, which included filing of tariffs.
I agree that “tariff rates” published were almost always mileage or distance driven. Exception ratings or commodity rates were developed to meet competitive pressures. When these rates were published it was generally worded to be for a specific shipper, or based on point to point movements. The fact that these rates were published did not mean that all other filed rates were lowered. Hence the general level of rates were preserved.
Which regulated tariffs? If you’re going to cite a tariff, then name the tariff.
I agree with MP173, when they first tried to “regulate” rates the government attempted to enforce a rate structure that reflected three things:
Terminal Costs (which were the same regardless of distance moved)
Line haul distance
Value of commodity
The LTL rate structure of his motor carrier reflected this because, 1) the truckers basically copied the railroad rate structure when they became regulated in the mid 30’s and, 2) the LTL motor carriers had a lot of protection from compitition. The government restricted who could haul the stuff (barrier to entry) and enforced the rates. You couldn’t kill a truckline under regulation.
The rail rate structure quickly broke down because it made no allowance for things such as supply, demand, compitition, etc. Hence, the point to point commodity rates that basically resulted from negotiations between shipper and carrier. They moved the vast majority of rail freight.
The inability of the railroads to get out from under the government structured rate plan on LTL played a big hand in moving that traffic to the highway. They did better with their carload business.
The existance of a tariff rate, in itself, means nothing. If the freight wasn’t moving on that rate, it’s just a piece of paper. You’d have to do your correlation on rates that actually moved business, and 25
I thought I was watching the ad for “Chicken Little” again…
But, no, it’s just FM with his usual, “the sky is falling” approach to railroads…
Of course, in the pre-Staggers world many of the granger roads were nearly or really bankrupt. Look at the finances of the CRIP, MILW or CNW and you’ll find that all were having a tough go of it.
In this instance, the data was developed in March of 1979 in preparation for ICC Hearings and most of the data for that was provided by Glenn Reynolds, Milwaukee Road’s Director of Pricing.
I guess I’d call that meaningless. When you go into a hearing you generally select evidence to support your position. So, again, what tariffs are you citing?
I know we took mileage into consideration, but it was part of a mix and it didn’t govern our decisions. And we charged less to move freight north than we did south because the demand for northbound freight was less than the demand for southbound freight. Even if the miles were the same, we charged different rates.
“A” tariff could tell you just about zilch. . A credible linear regression requires as many data points as possible. Which tariff? As many as possible. In this instance, a data set of Milwaukee revenue carloadings including, from actual waybills, distance (from origin and destination) and product.
Statistically, that is the relevant data.
For a linear regression to assess rate/mileage correlation, rate and mileage is extracted from the data set. A citation to a specific tariff is irrelevant at that point because the linear regression doesn’t care: it wants actual, real data.
We requested waybill data and that’s what we got. Since the purpose of the study was not to determine if Milwaukee waybills matched published tariffs, nor any reason whatsoever to suspect they didn’t, there was absolutely no reason whatsoever to be looking for individual tariff numbers in a compilation of thousands of waybills no doubt including hundreds of tariffs.
At that point in time, it was a safe guess that all such rates were regulated and so there was no question that every single combination of price and distance was governed by a regulated rate.
For current pricing on wheat carriage, BNSF Rate Book 4022K was used. Within that rate book are subsets of unregulated and regulated rates, corresponding neatly with competitive and captive shippers.
On the way to work, I was wondering what on earth this meant. Why look at the price differential on direction, when there are hundreds of price differentials on the same mileage in the same direction, because a railroad hauls different commodities/products at different prices, even over the same distance. Direction has nothing to do with it. There are simply another set of tariffs over the same mileage.
Chicago to New Orelans has/had hundreds of tariffs at different rates for the same mileage. Does that change the statistical relationship between each tariff and the mileage? Not one bit.
The “standard error” is a different statistical measure than correlation (r-squared). The strength of statisitical analysis is to be able to measure a relationship that may in fact have offsets up or down for a variety of reasons, but for which a specific factor is still the controlling factor in setting a rate.
Greyhounds is discussing a situation which would affect the standard error (in large samples, the standard deviation), not the correlation. The standard error has nothing, or at least little, to do with determining whether or not the rates are set by reference to mileage, only whether or not such rate differentials exist – which is a completely different measure than correlation.
Over very large data sets, a correlation can be very strong, and still have a relatively large standard error, that is, a variety of different rates over the same distance, and a multiple of such distances, all of which ultimately are set by reference to mileage, will yield a statistical analysis t
The underlying question remains, how does the BNSF justify charging Montana farmer more $$ to move grain a shorter distance than North Dakota and Minnesota farmers? The crops from the east wil use more fuel, more crews and plug up more terminal facilities than those from Montana. What gives here? The folks in Montana want to know.
It is an interesting note about shipping rates between Chicago and New Orleans and the reverse route. The return charges for the mty grain cars are figured into the loaded rate but what about cotainers from the midwest to the Pacific northwest? The railroad has no containers of their own any more so it is not their problem to balance loads and mtys to keep the equipment where it is going to be needed like they must do with grain cars. The balance component for containers is with the customers. Loads weight more but mtys take the same amount of space on the train. So it would seem there is no benefit to the railroad to charge much less for the mty move and besides, depending upon the time of the year, sometimes it is loads west and mtys east and a few months later it is loads east and mtys west. The railroad does not care about that any more, a box is a box is a box to them. Some are heavier than others but the customer needs to stage them, no longer a railroad pricing problem. They no longer have an incentive to price the move differently in order to give the market an incentive to route the loads and mtys for the benefit of the railroad.
On the way to work, I was wondering what on earth this meant. Why look at the price differential on direction, when there are hundreds of price differentials on the same mileage in the same direction, because a railroad hauls different commodities/products at different prices, even over the same distance. Direction has nothing to do with it. There are simply another set of tariffs over the same mileage.
Chicago to New Orelans has/had hundreds of tariffs at different rates for the same mileage. Does that change the statistical relationship between each tariff and the mileage? Not one bit.
The “standard error” is a different statistical measure than correlation (r-squared). The strength of statisitical analysis is to be able to measure a relationship that may in fact have offsets up or down for a variety of reasons, but for which a specific factor is still the controlling factor in setting a rate.
Greyhounds is discussing a situation which would affect the standard error (in large samples, the standard deviation), not the correlation. The standard error has nothing, or at least little, to do with determining whether or not the rates are set by reference to mileage, only whether or not such rate differentials exist – which is a completely different measure than correlation.
Over very large data sets, a correlation can be very strong, and still have a relatively large standard error, that is, a variety of different rates over the same distance, and a multiple of such distances, all of which ultimately are
Oddly, the results of that study are very similar to the regulated rates on the BNSF today. Probably just a coincidence that regulated rates show similar characteristics, at different points in time, on different railroads. Of course, if you are contending that therefore BNSF is also a basket case …
On the way to work, I was wondering what on earth this meant. Why look at the price differential on direction, when there are hundreds of price differentials on the same mileage in the same direction, because a railroad hauls different commodities/products at different prices, even over the same distance. Direction has nothing to do with it. There are simply another set of tariffs over the same mileage.
Chicago to New Orelans has/had hundreds of tariffs at different rates for the same mileage. Does that change the statistical relationship between each tariff and the mileage? Not one bit.
The “standard error” is a different statistical measure than correlation (r-squared). The strength of statisitical analysis is to be able to measure a relationship that may in fact have offsets up or down for a variety of reasons, but for which a specific factor is still the controlling factor in setting a rate.
Greyhounds is discussing a situation which would affect the standard error (in large samples, the standard deviation), not the correlation. The standard error has nothing, or at least little, to do with determining whether or not the rates are set by reference to mileage, only whether or not such rate differentials exist – which is a completely different measure than correlation.
Over very large data sets, a correlation can be very strong, and still have a relatively large standard error, that is, a variety of different rates over the same distan
Mileage may be the determining factor for a “rate”, but the determining the final “cost” will almost always be based on demand and supply.
You can file all the rates you want, they are starting points for negotiating the final cost. Sometimes, as in Montana or southbound Chicago -NO freight, there is no negotiating. That is traffic imbalance. Happens all the time. Look at airline pricing. It even happens in my field. Got a memo yesterday to the effect that “backlog is very low”, take a look at any business out there.
Had I gone and picked up an MBA like some of you I would know if that is elastic or inelastic demand. If you run a business or are in sales you dont need to know the proper name, you just realize the necessity to adjust the throttle.
Seriously, from the LTL days, rates began based on mileage (rate basis which was a form of railroad mileage), but then found the water level mark based on all other factors. Some rate programs I developed did not even factor mileage, but were based on the number of aggregate pickups and deliveries and total tonnage tendered.
Pricing and yield management are very interesting aspects of any transportation company’s marketing and are often very much protected within that organization. To say there is no reasoning on a company’s pricing is often based on looking at it on a micro economic scale, rather than the big picture.
I have always maintained that if I were negotiating the Montana wheat rates (for the farmers), I would use a method of tying that freight into a much larger pool of traffic. Large agricultural concerns such as Cargil, ADM, etc possibly have, or should have much favorable rates, based on mileage.
Ed-We were always very concerned about the impact of a local chemical shipper turning over their logistics to someone like UPS. Many large rail shippers have contracted their physical distribution management out to UPS and other firms. They can come to the table with a very large portfolio of business to offer; most of which is competitve with other railroads or motor carries.
Also, in the early days (1980s) of deregulation we would bid on a customers entire package of traffic within North America. The only thing the customer wanted to hear was how much we would charge him to move all of his traffic for three to five years. Other distinctions such as the specific commodity or distance became secondary to this primary issue. I can recall cases where the charge from the TX/LA Gulf Coast to California would be lower than the charge to Alabama.
Ed, I was very, very clear that unregulated or deregulated rates are poorly correlated with mileage. and I hope this is not an intentional misreading of my comments just to get a rise. Please reread my comments.
I’m not sure that you are seeing the distinction that the discussion related to regulated rates. Hence the specific date, 1979.
If what you say is true for that era, then the Staggers Act was of no particular use at all.
ICC Chairman Daniel K. O’Neal testified to the US Senate in January, 1978 that there were four railroad “basket cases,” the Rock Island, the Illinois Central Gulf, The Milwaukee Road and the North Western, but that the other railroads were not far behind.
When Milwaukee petitioned for bankruptcy, it had something like $10 million cash on hand and $38 million accoounts payable. When Stanley Hillman left ICG, it had 0 in the cash drawer and $100 million in accounts payable outstanding. Interestingly, as of December, 1977, the Milwaukee Road was not technically bankrupt, and ICG technically was.
Well, these economic platitudes are starting to wear a little thin.
Demand is the greatest where? Supply is the most constricted where?
And the rates are the lowest on those corridors, not the highest.
So this explains why rates are lowest where demand is the highest? Well, that is exactly what you are saying. And rates are highest where demand is the lowest because of supply and demand?
I have long been puzzled by these frequent simplistic references to “laws” of supply and demand, because they have made just about zero sense in this context.
The discussion has been about captive and non-captive shippers. The economic fact of life is that supply and demand has nothing to do with it; it is a matter of market share pricing, rather than cost pricing, which leads to plummeting margins for most traffic, especially where demand is the greatest and capacity supply the lowest which is exactly why the railroads are struggling with excessive demand and inadedquate supply – because they keep lowering the price, not raising it. Except to certain parts of the system where, ironically, there is little congestion and seemingly, lots of supply of train and track capacity.
It does not take an MBA to see that the platitude and the reality are two different things.