In telecommunications, regulators turn increasingly to the nebulous term known as "cost-based" to set pricing policy. An example is the new Telecommunications Act of 1996 (Act), whose pricing standards for interconnection and network element charges stipulate that the just and reasonable rate for the interconnection of facilities and equipment should be "based on the cost ... of providing the interconnection or network element,"1 and "may include a reasonable profit."2 The notion of "cost-based rates" also surfaces in many state regulatory proceedings.3
To economists, however, cost bears no universal or formulaic relationship to price. Though it is not a term of art, "cost-based pricing" for the practicing economist usually denotes a price that relies on incremental cost as a price floor, and that exceeds this floor in efficient ways based on market information and other variables.4 But regulators often use the term "cost-based rates" to describe the setting of prices via some sort of formula, using cost data as the primary determinant. That notion bears scrutiny. While cost-based pricing itself is not objectionable, formula-based pricing should give cause for complaint when it is thinly disguised as "cost-based pricing."
In telecommunications, at least for local exchange carriers (LECs), regulatory agencies cannot reduce the rate-setting function to a concise formula (em such as by adding an appropriate ("just and reasonable") level of contribution to overhead to the "true" incremental cost to arrive at a price to charge the end user or intermediate supplier.5,6 Simplistic, formulaic methods will not likely arrive at prices that allow LECs to address the challenges posed by emerging competitors, changing consumer demand, or developments in technology. The post-Act world demands much more of pricing policy.
Though it sounds reasonable, formula-based pricing is the economic equivalent of one of Al Capp's shmoos, the affable, smiling, bewhiskered creatures that taste better than anything, and cheerfully cook themselves in a jiffy. It's too good to be true.
While it is certainly possible to use cost computations to arrive at prices that, at least "on paper," are subsidy-free and seem reasonable, such prices will quite likely prove absolutely meaningless in a competitive market.
Competition Means a Small Markup
The proponents of formula-based pricing would simulate a competitive price simply by appending a "competitive" markup to incremental cost. They presume that such markups are low. But in truth, neither the markup (a percentage of costs) nor the gross contribution margin (a percentage of price)7 follows any predictable pattern in competitive markets. Contribution margins for firms in fiercely competitive markets are surprisingly large, as illustrated by the following examples.
Compact Disks. If ever there was a competitive market, here it is. But record stores will purchase CDs for between $10.72 and $11.20, and sell them for $16.98, yielding markups over the wholesale price of between 52 and 58 percent.8 Such a markup equals
a gross contribution margin of between 34 and 37 percent.
Running Shoes. While Nike's midpriced running shoes cost only $15 to $20 to produce, they retail for four times the factory cost in the United States.9,10 This gap adds up to some substantial gross margins.
Computers. In 1991, Compaq Computer offered personal computers (PCs) in the highly competitive computer market at prices that contained a 37-percent gross contribution margin.11 Even today, Compaq's gross margin target reaches 23 percent.12 IBM's numbers are similar: Gross margins run 48 percent on large servers; 40 percent on disk and tape drives; 33 percent on memory chips and PowerPC microprocessors; and 65 percent on software.13
Jeans. The contribution margin on a pair of jeans can be surprisingly large. Levi Strauss & Co. reports a 40-percent gross margin on sales in this industry, even though its market share is only 19 percent.14,15
Cereal, Cigarettes, and Razor Blades. Even after deducting the average value of coupons, cereal boasts a gross margin of about 20 percent.16 Philip Morris cigarettes run a margin of 32 percent, while Gillette blades and razors claim a whopping 36 percent.17
Souvenir Spoons. Even souvenir spoons, something nobody really needs, have a markup of over 400 percent!18
These randomly chosen examples of common household items demonstrate that regulatory markups of 5 or 10 percent over cost do not necessarily simulate the pricing behavior one would observe in a competitive market. The reason: Large markups are required to cover the joint and common costs and opportunity costs of competitive multiproduct firms. Looking at just the contribution margin for one product of a multiproduct firm does not tell the entire story. Similarly, applying competitively "small" contribution margins to regulated telecommunications services does not address the entire pricing problem.
Myth #2: Price Equals Long-run Marginal Cost
Basic economics tells us that, in the long run, a market reaches equilibrium when the market price equals both short- and long-run marginal cost as well as both short- and long-run average cost. At such an equilibrium, all costs are covered and profits are zero because prices equal average costs. This idea is fine as a theoretical maxim, but a literal and myopic reading can lead the policymaker astray.
The maxim applies most readily to a firm for which all costs may be directly attributed to a sole product. However, telecommunications companies are multiproduct firms that exhibit scale and scope economies as well as substantial joint and common costs.19 These costs are not included in any long-run incremental-cost calculation for any single product.
The maxim of P=MC implies that the sum of the separate long-run incremental costs for all of the products of a multiproduct firm add up to all of its costs, including the joint and common costs. This assumption simply isn't valid. Substantial joint and common costs are residual to the incremental-cost calculations, which is why we observe substantial margins in excess of long-run incremental-cost calculations for separate products in the real world.
Prices in competitive markets must cover not only incremental costs but also 1) the joint and common costs incremental to all services, and 2) the opportunity cost of capital. However, in applying "cost-based" pricing that is actually formula-based pricing, regulatory agencies may overlook these other costs (which can loom large for network-based industries such as telecommunications) and assume that markups on individual services must be "small" to be "reasonable."
Margin, Markup, and Return are Interchangeable
In attempting to add a "reasonable" contribution margin to a service's incremental cost, one common pitfall is to confuse three concepts:
s A contribution margin expressed as a percentage of price
s A percentage markup over cost to arrive at a price
s A reasonable rate of return on investment.
These three measurements yield strictly noncomparable numbers, yet they are casually used as interchangeable in telecommunications. However, a service with a high contribution margin or percentage markup over cost can have a low rate of return. Confusing rate of return for contribution margin or markup over cost tends to lead to prices that are too low. In telecommunications, regulators often recommend a 5 to 10 percent markup above incremental cost to arrive at a price, but such markups actually seem generated with rates of return in mind.
Consider a simple example. Suppose a product has a production cost of $5 per unit, and a price of $25. Its contribution margin would be 80 percent when expressed as a percentage of price (em in other words, a markup of 400 percent over cost. Now assume that the total level of sales the first year is 1,000 units, and that it cost $200,000 to buy the manufacturing equipment required to produce the product. In this example, the annualized rate of return on investment is 10 percent. Marking up the $5 cost by just 10 percent instead of 400 percent would yield a price of only $5.50, even though the market price is $25.
Myth #4: Prices Differ Only if Costs Do Too
Regulators expect service prices to vary as costs vary, and this notion seems to be a basic condition of formula-based pricing. Thus, if a service always costs the same amount to produce, but is offered at two sets of prices (e.g., two differing multipart tariffs), one could conclude erroneously that one or both sets of prices are not cost-based. However, given any significant differences in consumers' demand characteristics, it is economic naiveté to conclude that service prices should maintain the same ratios as their respective costs. Competitive markets just don't operate that way.
Efficient pricing in telecommunications usually requires price discrimination, in the economic sense. In fact, the practice is as pervasive in telecommunications as in other industries. Clear proof lies no further than the optional calling plans offered by interexchange carriers, or the simultaneous LEC pricing of basic local service at both flat and measured rates.
Microeconomics teaches that optional multipart tariffs are usually more efficient than pricing a service at a simple per-unit price.20 An insistence on strict formula-based pricing may preclude efficient price discrimination, such as volume and term discounts or discounts on packages of services.
The Cold, Hard Facts
Telecommunications pricing exhibits a disturbing trend of late: the belief that one can set "cost-based" prices by applying a multiplier to an LEC's incremental cost data. The multiplier is designed to build a "competitive" level of contribution into prices, and one can observe multipliers as low as 1.05 in the setting of "cost-based" rates. Unfortunately, this practice lacks all economic merit. Considering that the industry has largely broken itself of the habit of setting fully-distributed cost (FDC) prices for services, it is unsettling to confront the popularity of a pricing method that shares many of the economic infirmities of FDC pricing, and adds a few of its own.
The Telecommunications Act of 1996 will make the pricing of telecommunications services more critical, but formula pricing will not answer. The more competitive the market, the less valid a formulaic approach to pricing (em particularly where the formulas are based on specious assumptions to begin with. No formula can substitute for market-based pricing, especially in an industry poised to change dramatically (em as telecommunications will in the wake of the Act. t
Alexander Larson is senior economist in the regulatory and external affairs department at Southwestern Bell Telephone Co. in St. Louis, MO. Mr. Larson has an MS in economics from the University of Illinois, and is widely published in the area of telecommunications regulation. The author wishes to thank Curt Hopfinger, Dale Lundy, Tom Makarewicz, Doug Mudd, Steve Parsons, Terry Schroepfer, Margarete Starkey, and Jack Van Pelt for their valuable comments. Chris Graves assisted with research.
1. Telecommunications Act of 1996, Pub. L. No. 104-104, 110 Stat. 56, (sc 252(d)(1)(A)(i) (1996).
2. Telecommunications Act of 1996, Pub. L. No. 104-104, 110 Stat. 56, (sc 252(d)(1)(B) (1996).
3. For example, the Texas Public Utility Commission set the prices of integrated services digital network (ISDN) at a markup of 5 percent over incremental cost. TEX ADMIN. CODE tit. 16 (sc 23,69(f)(2)(D) ("To further the commission's policy that ISDN be made available at a reasonable price and that ISDN be as accessible as possible to those customers who want ISDN, [basic rate interface] should be priced ... at not less than 100% of [long run incremental cost] and at not more than 150% of [long run incremental cost].")
4. For example, in the paradigm of perfect competition, efficient cost-based prices are equal to marginal costs, and efficiency requires that prices not exceed marginal cost Ramsey pricing is a cost-based approach that uses demand elasticities to determine the most efficient ways for prices to exceed marginal cost.
5. This corresponds directly to the Act's concept of a "reasonable profit."
6. "Incremental cost" is the additional cost caused by engaging in a given economic action or decision, such as producing and marketing a service. For a service, incremental cost includes volume-sensitive costs and any service-specific volume-insensitive costs associated with the provision of that service, or change in the quantity supplied of that service. It excludes costs directly attributable to the production of other services, as well as unattributable costs that 1) are incurred in common for all the services supplied by the firm, and 2) do not vary with the level of output. The incremental costs of service are economic costs that represent the-forward-looking value of the resources caused to be used or expended by the provisions of that service.
7. The term "profit margin" is often used interchangeably with the term "contribution margin" in this article. This is potentially confusing because the term "profit" usually refers to the residual revenues after deducting total costs, including joint and common costs, and as used popularly, the term "profit margin" really refers to an individual product's or service's contribution toward overheads (which in turn must be deducted from total revenues in computing a firm's total profits).
8. Neil Strauss, "Pennies That Add Up to $16.98: Why CD's Cost SO Much," Wall Street Journal (July 5, 1995), at B1 (midwest ed.).
9. This depends on where they are made. The cost is $15 in China or Indonesia, $20 in Korea. Mark Clifford, "Spring in Their Step," 5 Far Eastern Economic Review 56, 59 (November 1992).
11. "Infotech: Fast Times at Compaq," Fortune 121, 122 (Apr. 1, 1996).
13. Brent Schlender, "Big Blue is Betting on Big Iron Again," Fortune, 106 (April 29, 1996).
14. Robert Lenzer & Stephen S. Johnson, "A Few Yards of Denim and Five Copper Rivets," Forbes 82, 3 (February 26, 1996).
15. Id. at 86.
16. "Cereal Wars: A Tale of Bran, Oats, and Air," Fortune 30 (May 13, 1996).
18. William M. Bulkeley, "Today's Mystery: Who Buys All Those Souvenir Spoons? Fort Inc. Sells 6 Million a Year, to Mr. Fort's Surprise," Wall Street Journal (June 6, 1991), at A1.
19. The discussion in John C. Panzar, "Technological Determinate of Firm and Industry Structure," in 1 Handbook of Industrial Organization 3, 13 (Richard Schmalensee & Robert D. Willig eds. 1989) makes this quite clear. Note, however, that incremental costs include product-specific fixed costs.
20. This holds true under a wide variety of conditions found in telecommunications, See, Robert D. Willig, "Pareto-Superior Nonlinear Outlay Schedules," 9 Bell Journal of Economics 56 (1978). See also, Daniel F. Spulber, Regulation and Markets, 552-53 (1989); Louis Philips, The Economics of Price Discrimination, 181-82 (1983).
Articles found on this page are available to Internet subscribers only. For more information about obtaining a username and password, please call our Customer Service Department at 1-800-368-5001.