THE ROAD TO RETAIL COMPETITION IS A LONG ONE. HAVING realized that, utility management has quelled its initial panic and has begun to concentrate on longer-term objectives. For instance, how much market share am I likely to lose during competition's early stages? And what prices can I charge to various customer classes without incurring a loss in market share?
The answer could affect decisions about future load, asset divestiture and competitive strategies.
Quantitative measurement of future load in a competitive world relies on methods that capture customer sensitivity to changing prices. %n1%n Estimates of the price elasticity of demand, defined as the percentage change in demand for electric power caused by a given percentage change in the price of power, can be used to formulate generation market share. The problem might be characterized in this way: How much market share would the incumbent utility lose to competition if it should continue to charge regulated rates while competitors were free to enter and drive prices down toward the variable cost of producing electricity?
We know, for example, that even under regulated rates customers still have some ability to substitute for energy purchased from their electric provider. For instance, customers could always choose to cut their electricity consumption if prices rise. (This ability to curtail consumption is reflected in studies that estimate negative demand elasticities for electric power. %n2%n) Yet while customers have never been completely at the mercy of the incumbent provider, it is indisputable that their choices will expand dramatically under retail competition.
This widening of choices will transform historical relation- ships between electricity use and price. Thus, we analyze three possible scenarios: (1) no competitive alternatives (historic franchise regulation); (2) moderate customer choice, as exists today in a few dual-served territories; (3) full retail competition, which has been approximated using evidence from highly analogous industries that have already undergone a monopoly-to-competition transformation. By analyzing all three cases, the increase in customer responsiveness as a function of acceptable substitutes is explained.
No Competition: Small Losses
The "no competitive alternatives" case is based on actual information from Niagara Mohawk Power Corp. %n3%n NiMo's data for tariffs and customer load by rate class along with information about the economy and weather patterns were used to produce a load forecast. The demand elasticities were drawn from this exercise. The estimates were similar to elasticity estimates published elsewhere. %n4%n For example, the NiMo data produces estimated elasticities for residential customers of about -0.25. In other words, for every 10-percent change in electricity prices, the change in electricity used was -2.5 percent. Similarly, the estimated demand elasticity for commercial customers was also about -0.2, while the estimated demand elasticities for industrial customers were just greater than -0.3. The estimates for NiMo's customer price sensitivity were combined with results from previously published studies to suggest a "lower bound" for market share losses due to retail competition. These findings appear in Table 1.
The predicted changes in market share provide a picture of modest price sensitivity in a regulated environment where customers do not have an option to choose another (retail) supplier and only industrial customers see benefits from competitive wholesale transactions. However, with the advent of retail competition, customer price sensitivity will increase as the availability of acceptable alternatives grows.
Territories, Greater Risk
The moderate customer choice case is drawn from customer responsiveness to price changes in various parts of the country where two utilities compete for joint service territory. If these customers show a strong propensity to switch suppliers when confronted with a single alternative supplier, then the "lower bound" market share losses shown in Table 1 will understate potential declines.
Information exists about a group of nonstandard utility markets where two electric companies currently provide retail service to a single city or town. There are at least 46 such communities in the U.S. where limited retail competition exists, allowing new and/or current customers to choose their electricity supplier usually at no cost, with as little as three day's notice before switching, and with no time commitment to the chosen utility. Among the cities in this sample are Cleveland, Traverse City, Mich., Culpeper, Va., and Duncan, Okla.
Three observations drawn from these markets where retail competition exists offer a useful middle ground for understanding how a competitive retail environment might affect utility market share.
NONUNIFORM PRICES. First, communities served by two utilities usually do not face sustained uniform electricity prices. In Cleveland, two utilities compete at retail and the sustained price difference ranges from 7 to 10 percent.
INFREQUENT SWITCHING. Second, despite a persistent price difference between sellers, customer switching at any point in time is infrequent, ranging from as little as 0.3 percent per year on average in Columbus, Ohio, to 6 percent in Duncan, Okla.
CUSTOMER INERTIA. Third, additional evidence also suggests this customer inertia cannot be completely overcome in the short run by competitive strategies like new service offerings or reduced prices. In Ohio, local municipal provider, Newton Falls Utility Department, competes with Ohio Edison for customers within city limits. Despite NFUD's 30-percent lower electricity price, only a few customers have left Ohio Edison.
A Digression: The Drug
The case for estimated market share losses under full retail competition relies on a combination of evidence from existing generators, beginning with ÑiMo, and other industries that have undergone the transition from protected to competitive markets.
Arguably the most germane example comes from the patented drug industry where the product is protected - like a utility's service territory - from competition for the 17-year term of the patent. However, once the protection is lifted, the full force of competition is brought to bear on the determinants of supply and demand for the drug. In that way, "pioneer" drugs, those protected by a patent, provide a setting similar to the introduction of retail electric competition in which the conditions of market entry and competition change radically on a given date set by law or regulatory policy. %n5%n
Shared industry characteristics between patented drugs and regulated electric markets are strikingly similar:
Consist of many firms with products serving exclusive markets;
Exchange large up-front investments for exclusive right to sell into a well-defined market;
Products offered by the incumbent and potential entrants are physically identical (branded drugs versus generic).
Production capacity is largely fungible, with considerable latitude about which facilities produce the product.
Product price is much more than its variable cost.
Manufacturing and distribution do not have to be integrated.
Apart from the protection afforded by law or regulation, the cost of market entry is small.
Cost of product failure is extremely high.
There is likely to be some demand-side resistance to new entrants, given the goodwill accumulated by the pioneer while the market was protected.
Full Competition: Small Loss at
First, then Accelerating
Evidence following competitive entry in the drug market presents both strategic and empirical issues for electric utilities. Strategic implications vary over time. First, as expected, the loss of an exclusive franchise results in immediate competitive entry. Later, however, entrants must invest in brand promotion and offer customers significant price discounts to capture market share. Thus, as competitive entry continues during the first year or two following the loss of an exclusive franchise, the demand curve facing sellers gradually "kinks." Entrant's products tend to occupy the price-sensitive region of the kinked demand curve, while the pioneer drug occupies the price inelastic (insensitive) region of the curve.
Evidence regarding competitive entry in retail generation comes from a combination of experience from the drug industry and recent empirical evidence from electric industry restructuring. To illustrate this case, we used data from the Pennsylvania Public Utility Commission's final order in the PECO Energy restructuring proceeding. That order provides a basis for translating what happens to patented drugs into what is likely to happen to utility market shares.
In PECO's case, the commission ordered that retail customers be given "shopping credits," which exceeded PECO's estimated market-clearing price for competitive generation. But when combined with rates for T&D and competitive transition charges, the credit still met the rate cap. In this way, the potential price reduction seen by customers following the introduction of retail competition will reflect the difference between the market-clearing price (estimated at 2.8 cents) and the shopping credit. The larger the shopping credit, the larger the potential price difference the customer would see reflected in his electricity bill.
This estimated price difference between a hypothetical range for the shopping credit and the predicted market-clearing price was combined with the elasticity estimates from the drug industry to produce this paper's forecast of market share loss with the introduction of retail competition. The results appear in Table 2.
Recall that the design of this analysis suggests the estimates of lost market share are maximum losses. Under the conditions used to produce Table 2, the incumbent utility is assumed to charge the maximum rate (the formerly regulated rate or rate cap), while entrants are expected to charge a rate equal to the market-clearing price. Further, the incumbent utility is not expected to engage in any form of advertising.
The findings show that if a utility takes a passive stance and does not match entrant prices, the loss in market share over the first two years following retail competition ranges from 13 percent to 21 percent of the utility's residential load, 14 percent to 21 of its commercial load, and 40 percent to 62 percent for industrial customers, depending on the size of the shopping credit. While these estimates of lost market share are well more than the pre-retail competition estimates shown in Table 1, the data indicate that at least some of an incumbent utility's existing customers may be retained following retail competition, depending on the magnitude of the shopping credit.
Besides estimates of market share, the study also suggests the rate at which competitors are likely to "eat away" at an incumbent generator's market share following retail competition. If the incumbent does nothing to boost brand awareness, the predicted reductions in market share following the loss of an exclusive franchise are 20 percent immediately, reaching 75 percent of eventual erosion by the end of the first year. The remaining 25 percent market share loss occurs during the second year of of retail competition. %n6%n So, although the incumbent takes an initial hit, the rate of market share loss tends to slow rather than accelerate over time.
Scott T. Jones is CEO and Matthew B. Krepps is principal at The
Economics Resource Group Inc., a consulting firm in Cambridge, Mass.
1 The sensitivity of sales to price changes is directly related to the number of close substitutes for electricity generated by the incumbent utility. The dramatic increase in the closeness and number of substitutes available under competition implies that historical demand elasticities provide only a lower bound on true demand elasticities under open retail competition.
2 For example, see Robert Halvorsen, Econometric Models of U.S. Energy Demand, 1975. For a survey of the empirical literature, see Douglas R. Bohi. Analyzing Demand Behavior, A Study of Energy Elasticities, 1981.
3 NiMo put testimony before the New York State Public Service Commission (Case 96-E-0134) in 1996 that included data used to generate estimates of short-to-intermediate term demand elasticities (up to three years forward) across rate classes.
4 Bohi, op. cit., and John E. Kwoka Jr., Power Structure, Ownership, Integration and Competition in the U.S. Electricity Industry, 1996.
5 A selection of the most relevant articles from the drug industry include the following: G. Urban, et al., "Market Share Rewards to Pioneering Brands: An Empirical Analysis and Strategic Implications," Management Science, Vol. 32, No. 6, June 1986, pp. 645-659; R. Caves, et al., "Patent Expiration, Entry and Competition in the U.S. Pharmaceutical Industry," Brookings Papers on Economic Activity, 1991, pp. 1-66.; H. Grabowski and
J. Vernon, "Brand Loyalty, Entry and Price Competition in Pharmaceuticals after the 1984 Drug Act," The Journal of Law & Economics, October 1992, pp. 331-350; S. Ellison, et al., "Characteristics of Demand for Pharmaceutical Products: An Examination of Four Cephalosporins," RAND Journal of Economics, Vol. 28, No. 3, Autumn 1997, pp. 426-446; R. Frank and D. Salkever, "Generic Entry and the Pricing of Pharmaceuticals," NBER Working Paper Series, Working Paper 5306, October 1995; R. Frank and
D. Salkever, "Pricing, Patent Loss and the Market for Pharmaceuticals," Southern Economic Journal, July 1992, pp. 165-179.
6 H. Grabowski and J. Vernon, op. cit., pp. 335-337, and R. Caves, M. Whinston and M. Hurwitz, "Patent Expiration, Entry and Competition in the U.S. Pharmaceutical Industry," Brookings Papers on Economic Activity, Microeconomics, 1991, pp. 19, 35-37.
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.