For each major railroad in 2003, the average R/VC of captive and competitive traffic was as follows:
RR …Captive…Competitive
BNSF… 215.6% 109.1%
CSX… 186.8% 93.1%
NS …209.2% 102.3%
UP …210.4% 106.8%
These are averages. A study I participated in two years ago showed that on BNSF for a given commodity there was a variation of 20% either way system wide in competitive areas and ranged from between 240% and 380% in captive areas. In other studies, BNSF’s average has been as low as 101% R/VC for competitive traffic.
It is inarguable that there is cross-subsidization occuring by even the narrowest definition of cross-subsidy.
So let me ask you the question that I don’t seem to be getting a plethora of responses to:
What is the definition of cross-subsidy?
I thought that the term referred to the subsidy of a money-losing LOB, e.g. the sale of a product at a price less than the cost to manufacture it - something that a manufacturer might do in order to gain name recognition/boost market share.
Certainly the term can’t just refer to one LOB being more profitable than another LOB within a company. So what determines when cross-subsidization is taking place? I’m not trying to define the term, I’m trying to understand it.
It is a regulatory concept from decades ago. When railroad’s prices and services were regulated, government agencies would tell they had to run a money losing service and could make up the difference on the profits from another service. As an example coach passenger rates were held low but the railroads were told they could make it up on their share of Pullman travel. It was almost like a income transfer from wealthy passengers to poorer passengers. The STB no longer has the authority to make such a ruling.
The generic definition of cross-subsidy that I have seen repeatedly is this:
"Cross subsidization occurs when one purchaser pays a relatively high price and thus enables another purchaser to pay a relatively low price.’
True enough. If the game is to earn, achieve or exceed WACC [weight adjusted cost of capital], the market no doubt determines the feasible price mix to achieve the goal. The problem is when a customer is captive. There is no market based price. And this presents the dilemma when a company – and many do – utilizes market share pricing rather than marginal cost pricing.
The company can buy more market share – at lower prices than it would or could ordinarily charge – if other customers can be forced by their captivity and inelastic demand needs to contribute the additional profit margin necessary to meet the earnings goals of the company. If all the players in a given market have captive shippers, that enables those companies to compete on price for market share at rates lower than any of those companies would or even could otherwise charge those customers. Sometimes they don’t have captive customers and still go for market share pricing for a variety of reasons. Auto companies do it all the time, and pay the economic consequences from time to time.
In essence, the customers paying the lower share are subsidized by the very existence of the captive shipper – receiving services at prices that could not be rationally offered to them based solely on market principles, without the compensating income gained by the company at the other end of the price spectrum.
It is both an artificial subsidy, for some, and an artificial price penalty, for others. It could not exist in either case if genuine market conditions prevailed.
It is not differential pricing, rather it is inverse pricing – the captive customer often pays more for less service, rather than for a differentiated service.
In the world I’m familiar with, there’s some pricing theory, and what works. Not all this mincing about cross-subsidy and such. I think lean manufacturing and lean accounting are starting to clean out some of this stuff.
You bring up an interesting example here. Let me posit a counter-example:
This hypothetical railroad named BNSF decides to set rates to 160% R/VC across the board. In doing so, it loses 25% of its business to competitors (and I don’t care whether those competitors are other railroads and any other means of transportation or if it just means that the shipper goes out of business). Straight up, there is a 25% reduction in variable costs, so VC is now $5.55 billion. But fixed costs only decrease 5% (just for sake of the example) so FC is now $2.47 billion. Total costs are now $8.02 billion. At 160% R/VC, revenue now comes in at $8.88 billion for a profit of $0.86 billion or a 10.7% rate of return on revenue.
If BNSF decides that it wants to return to the same level of profitability (15.63%) that it had before losing all that business, it will need to boost its rates to 171% R/VC across the board. Of course, doing so will drive away even more business…
Ordinarily, textbook production reflects that the variable costs of production fluctuate with changes in product numbers. No less so than with trains. A 20 car train has much the same overall variable costs of operation as a 110 car train, and some differences as well (less fuel, lower capital charges). However, for the smaller train, the cost per revenue unit will be higher than with the big train. But, as a rule, less business higher VC, more business lower VC per unit. Marginal cost pricing theory grants to the railroads the entire variable cost, no matter what it is. That is, 100% of VC are recovered.
The historical FC on BNSF is about 35%-37% of the VC. FC and VC are linked more closely than people might imagine, at least statistically, and at least on railroads. There is certainly an interesting hypothetical out there to imagine “what if”, but the 180% R/VC threshold is a very liberal standard, for which we now have 25 years of actual experience and data to draw from.
Too, that liberal threshold permits a wide range of rate making and the average is only that – a statistical artifact of thousands of independent prices tailored to each competitive market or service need, or “not”, as the case may be, and varying as noted above, from a net loss on variable costs to a very healthy profit margin from captive shippers who are plainly subsidizing competitive shippers. Averages, in that sense, are difficult however to budge. Very difficult.
Elasticity plays a large role, and that contributes to the stability of the statistical average. Coal and agriculture are about 60% of the whole RR ball of wax, and the demand for railroad services for those industries are relatively inelastic. For the 31% of RR revenue that is extracted from captive shippers, the elasticity approaches zero.
Or improve its productivity which is, in general, the long term approach to maintaining profitability. That, of course, is why the economists who designed the deregulation process felt it necessary to impose a marginal cost pricing guideline to restrict cross-subsidization – the easy alternative to productivity increases is simply to raise prices – and that tends to be for captive shippers.
That design, then, is important. Productivity improvements are the driving engine of capitalism, not raising prices because it easier. The whole idea of “the market” is to drive price efficiency – actively compel it. There is no incentive to efficiency in a captive market, for the obvious reason that there is no actual “market.” That’s why the protections are given to the captive shippers in the Staggers Act and should be enforced.
They do not protect the captive shippers, they protect the economy as a whole.
Could you provide me with some examples of what is considered FC vs VC? I mean, some things are obvious, like fuel costs are VC while company management is FC. What is track repair? The engineer’s (prorated) salary? Or an agent that deals with three shippers in a terminal area? If a track services only one shipper (such as branch line to a coal company), is that track maintenance considered VC or FC? How about a manager (and staff) that provides service to a single intermodal firm? RR ownership of boxcars? Are there guidelines for this somewhere? Or is it up to each individual RR to decide if a particular cost can be allocated to a particular shipper?
Improvement of productivity is another area that seems a little counter-intuitive to me. No, I don’t mean that RRs shouldn’t improve their productivity, but let me use another example—
VC = 100, R = 180, R/VC = 180% – shipper is happy
Now if improvements are made resulting in a VC reduction of 20 –
VC = 80, R = 160, R/VC = 200% – shipper is unhappy even though the rate lowered?
In other words, if the improvement in productivity occurs on the VC side of operations, R/VC will rise even if the full savings of the improvement in productivity are passed on to the shipper.
As the designer of deregulation in America, Alfred Kahn was no dummy. That’s how all the world works – businesses routinely pass on more than 100% of their productivity increases – and still become more profitable. Passing on less than 100% creates a level of profit which absolutely at some point invites competitors – and then the rate drops back down, with the new competitors in hot pursuit. That invites the “race to the bottom.” A sophisticated manager does not want to be there. Producers always end up yielding more than 100% to the customer, even if they “think” they can do otherwise.
Yet still earn more profit.
That is the genius of the market system, it is genuine win-win, and a careful producer passing through more than 100% will carefully price to minimize competition at that price.
You will note that the R/VC does in fact rise – but if the equation is completed through to the penultimate factor – profit – the higher R/VC % very coincidentally begins to show net profits which are unattainable in a genuine market. The R/VC formula is, in fact, a good predictor of that if you finish the calculation. In that instance, you will see that the seller that is driven by authentic market forces would be compelled to lower the price charged – pass through more of the productivity – and the R/VC ratio declines as a result, and both complies with the law and its purpose to simulate market forces in non-market circumstances which is … to drive productivity increases and price efficiency.
Ouch. I am not a railroad accountant and never played one on TV. My first lessons on the topic were from Ernest Poole, Costs – A Tool for Railroad Management, 1962. I still have the book. To the extent that there are always some discretionary decisions – arguments – battles royale – between what is fixed and what is variable in any industry, the STB prescribes these currently under the UCRS and to the extent that they are statutorily defined for reporting purposes, there is a complete list of such costs which STB uses in its Railroad Cost Program for the specific R/VC function.
Why is it that whoever, by whatever method, attempts to dictate a paradise of “equality” and plenty ends up producing misery and want? There is a certain irony in the collapse of the I. C. C. and the Soviet Union coming within such a short time of each other.
Okay, having read more than anyone would ever want to know about how STB adjudicates shipping rate complaints, I have come to the conclusion that the STB includes absolutely everything that could possibly be attributed to the cost of supporting one particular shipper to the variable costs side of ledger. I’ve even seen line items such as the number of steel welding guys that are needed to repair cars.
I’m left with wondering exactly what is on the fixed costs side as those costs are not mentioned. No real need to mention those costs as the pricing ratio is against only the variable costs.
One statement that I ran across a couple of times that has me rather baffled. I don’t have the exact quote, but the STB makes the claim that even if R/VC is less than 100%, that is not necessarily an indication that cross-subsidization has taken place. The only thing that I can suggest is that the phrase “not necessarily” in this case means that there is a distinct probability (however small) that under the most awesome of circumstances it just might be possible for R/VC to be less than 100% and yet cross-subsidization has not taken place.
The STB does not care about cross subsidies. The only thing they care about in a rate case is if a rate is “reasonable”. If the rev/lrvc is over 180 and their is a lack of competition the railroad has a very heavy burden to prove their rate is reasonable.
That is accurate. Cross subsidies are part of the economic theory that underlies the Staggers Act and the 180% R/VC. Railroads were given free reign for so long as the rate was under 180%, including the right to cross-subsidize to their heart’s content for so long as they did not exceed the rate threshold. Administrative agencies are not good places to undertake broad discussions of policy and economics.
The 180% threshold is the tangible economic (and legal) expression of a large body of economic theory and discussion to achieve underlying economic goals for society. Railroads bear a burden of proof of “reasonableness” above that threshold because on a case by case basis there can, in fact, be legitimate reasons under the marginal cost theory as to why 180% R/VC does not offer a reasonable rate of return – a high fixed cost installation, for instance, to serve a given shipper.
As with any legislation, the trick is in the proper administration of the intent of the law, as applied on that case by case basis referred to above. And, as with any agency charged with that administration, the execution of Congressional intent is often less than stellar. Nothing new about that.
Part of the momentum leading to the Staggers Act was the increasing bureaucracy of the ICC (and FRA) in dealing with Congressional intent, coupled with (often conflicting) court decisions. Congress acted, nearly every 20 years, to try and fix the Interstate Commerce Act that the railroads themselves had originally clamored for. As the original post on this thread shows (I take a different lesson from the PRR drama), the railroads were often t
Well, it won’t surprise many of you to read that I strongly disagree with Mr. Sol, his numbers, his “study” and his conclusion.
When someone does a “study”, or analysis, like this, a final step is to ask yourself “do my results make sense?” If they don’t, there is a need to go back and check the methodology used.
Mr. Sol’s conclusion of a sure thing “cross-subsidization” just doesn’t make sense. No for profit corporation would willingly subsidize a customer. As Bob pointed out, in the past the government forced cross-subsidization through their economic regulation. But that’s pretty well gone now so Mr. Sol’s claim of voluntary cross-subsidization flies in the face of all reason.
Time to check things out. The first question is obvious. How did Mr. Sol classify customers as “Captive” or “Competitive”. This is a very subjective thing, but he seems to have readily been able to put rail customers in two very distinct classifications. So my first question to Mr. Sol is: "What criteria did you use to classifi
Ford, GM, and Chrysler do it all the time. They admit it. They are clearing inventories, they are attempting to retain market share, sometimes they are simply driving up the risks of entry for prospective market entrants. Obviously, this is something you simply know nothing about, as you admit.
Cross subsidization is all over the place. I am not going to try and justify here what you should have learned in school, or especially at International Harvester which did it all the time. I will suggest this.
Go to the bookstore that handles business books.
Go to the section of books that are written by Michael Porter. I am simply using his name as an example because people genuinely familiar with management will instantly recognize his name as one of the foremost authorities on business strategy.
Go to the book entitled Competitive Advantage: Creating and Sustaini
Greyhounds once strongly argued that wheat rates had gone down over the past 25 years, and offered every misreading of every study he could lay his hands on, refusing even to read the explicit texts of the studies he cited to which stated that prices had gone up.
He was vociferous, adamant, expending a great deal of time and energy, even though he has no experience at all with the wheat industry or its rate history-- at one point selecting out of seven or so commonly used inflation indexes the only one that had nothing whatsoever to do with either railroads or agriculture to try and “prove” his point. It was all heat and no light. So much for appropriate methodology.
It was interesting then to see the October 6, 2006 GAO study which showed … wheat rates had gone up, and in fact GAO singled them out as one of the few categories of rail traffic that had in fact gone up over that time.
So, it is interesting now to see that he has a methodology for “stepping back to check the methodology used.”
Judging by the wheat rate example, I gather he has always considered himself exempt from the analytical process he describes as important.
In nearly every study I have ever seen, the 180% R/VC threshold is used as the measure of captivity for study purposes. The underlying econometric studies found captivity to exist in typical market settings of around 160% R/VC, but in any case, the 180% R/VC used was a “probable” in statistical terms, even though railroads got the benefit of the doubt when the statute, a political co