THE RESIDENTIAL MARKET STANDS AS THE NEXT FRONTIER for natural gas unbundling. In California, Illinois, Maryland, Massachusetts, New Jersey, New York, Ohio, Pennsylvania and elsewhere, states have introduced pilot programs and other unbundling efforts to target residential gas consumers. %n1%n
These efforts are hardly surprising. The residential market, presently dominated by the regulated local distribution companies, appears lucrative. In 1995, the residential sector of the U.S. natural gas industry consumed 4,736 trillion Btu of natural gas or 32 percent of all natural gas delivered by LDCs in that year. %n2%n U.S. residential consumers accounted for $28.7 billion or 59 percent of the gas utility industry's total revenues. %n3%n
Nevertheless, despite all the enthusiasm industry representatives have recently expressed in trade publications and public forums, the creation of a competitive residential market may prove a very slow process. Marketers appear cautious in taking the responsibility of serving residential consumers, %n4%n and for very good reasons. Gaining a sizable portion of this market requires substantial investment in mass marketing, development of name recognition, acquisition of appropriate technology and employment of skillful personnel. Moreover, residential customers do not behave rationally in a "neoclassical" economic sense. They react not only to a price but to several qualitative factors that have yet to be studied by LDCs and marketers.
It appears natural to assume that marketers can take share away from the LDCs by offering commodity price discounts. Some LDCs have sought to turn that assumption to their own advantage, predicting that lower commodity prices will boost gas use and overall throughput, thus increasing LDC revenues from distribution and transportation service. %n5%n But risk lies behind every assumption. Marketers may discover that discounts do not pay off as expected. LDCs may find little opportunity to piggyback. Marketers invest in discounts, but will those expenditures pay off as expected?
This article offers results from creating a software program and model %n6%n that answer two basic questions: (1) What share of the residential natural gas market can be realistically captured by non-regulated suppliers? (2) Will residential unbundling increase total throughput for gas utilities? If so, by how much?
The results presented in this study are illustrative. The models' inputs reflect characteristics of a "typical" residential energy market with "typical" end uses, customer profiles and energy prices for the Northeast region. The data taken for an actual LDC can produce a different outcome (at some extent). The data used here to calibrate the model of consumer behavior are not based on empirical studies, but represent judgment and estimates.
This model of a typical utility market makes three assumptions:
1. The market is circumscribed by the service territory of a single incumbent LDC that continues to provide merchant services to residential customers on its system;
2. Each residential consumer has a choice of taking merchant and transportation services from the incumbent LDC or taking merchant services from a marketer and relying on the LDC for transportation services, which are available at an unbundled tariff rate; and
3. All marketers operating in the LDC's service territory act as one "typical" marketer. They compete with the LDC for providing merchant services.
A model based on these assumptions appears representative for the natural gas industry during a transitional period. %n7%n During this period, the "success" of competition will depend on whether marketers can capture a sizable portion of the residential market.
Market Share: Price vs. Service
The model tests the effect of discounts on market share in two ways: first, by size of discount, without any reference to service quality or intangible factors (so-called "pure" price competition); second, with the additional consideration of qualitative factors such as reliability and convenience. In this way, the model shows that marketers might just as easily gain market share by improving their qualitative comparison vis-a-vis LDCs, as by increasing the size of the discount.
PURE PRICE COMPETITION. The first "what-if" starting point assumes that the only difference between marketers and the LDC (as a natural gas merchant) is the supply price. The model assumes that consumers entertain no preferences toward an LDC or marketers as their potential suppliers.
To study this pure price competition, the model tested two scenarios, Scenario 1 and Scenario 2, under which marketers offer their supplies at price discounts of 10 percent and 20 percent, respectively. Note that these discounts are applied only to commodity prices. The distribution rates charged by a local gas utility do not depend on a supplier and stay constant in both scenarios. Thus, if a supply cost represents 50 percent of a residential customer's bill, the total bill reduction will only be 5 percent and 10 percent under scenarios 1 and 2, respectively.
The results of pure price competition modeling are presented in Figure 1. These results suggest that non-regulated suppliers will dominate the residential natural gas market under pure price competition. Their shares will fall around 55 percent (Scenario 1, 10 percent discount) and 61 percent (Scenario 2, 20 percent).
Nevertheless, these results should not be interpreted as a statement that marketers could easily capture 55 percent or 61 percent of the residential market. On the contrary, slashing prices is a very expensive way of obtaining additional load. As we all know, natural gas is a very volatile commodity. Consistently giving residential customers even a 10-percent savings could prove extremely difficult and costly at $40 to $60 per customer.
QUALITATIVE FACTORS. The case of pure price competition reveals very little about reality. However, the results of two
discount-only scenarios serve as a benchmark here for further analysis. Pure price competition does not reflect the complexity of the consumer's decision-making process; many qualitative factors influence customers. These factors include, but are not limited to, the reliability of gas supply, service quality, billing convenience, location (local versus out-of-state) and name recognition.
A recent study by Cambridge Reports/Research International shows that most residential customers (70 percent) place higher importance on having guaranteed delivery of natural gas rather than a reduced monthly bill. Moreover, more than 20 percent of customers feel that "price does not matter as much." %n8%n
For years, LDCs have provided reliable gas service to their customers. They have visible local presence and well-organized customer service departments. Allowing for this, we assumed that natural gas consumers perceive a local utility as a more favorable source of gas supply: higher reliability, solid customer service and billing convenience. %n9%n
This study introduces a qualitative factor ("Q") reflecting consumer perceptions toward non-price characteristics in gas supply services. In the model, the Q factors for utilities (Ql) and for marketers (Qm) serve as input parameters for the choice models.
The model tests the impact of qualitative factors on market share. It compares the market share for unregulated natural gas merchants (under each of the two scenarios for price discounts) with a value given as DQ, which represents the difference in qualitative factors between LDCs and marketers (DQ = Ql - Qm). The analysis indicates (see Figure 2) that DQ drastically affects the outcomes of competition between a utility and non-regulated entities. If consumers perceive marketers as non-reliable suppliers with poor service quality, their share can dip well less than 10 percent.
The model reveals that by increasing its comparative standing in qualitative factors (thereby reducing DQ) marketers can increase share as effectively as they would by discounting price. Thus, Figure 2 shows that by a reduction in DQ by a factor of 0.2 (moving to the left along the horizontal axis), non-regulated suppliers can gain as much as market share as they might by slashing prices 10 percent. It appears that addressing service quality and/or improving other qualitative factors could make for a much more efficient strategy for marketers than to pursue any significant price reductions.
This observation begs a question: What is the most cost-effective method of gaining market share? Answering this question requires understanding the economics behind the Q factor. What does it mean and how much does it cost to increase the Q factor by a certain amount?
In this model, the Q factor represents the sum of a series of qualitative factors (qi) relating to different non-price characteristics of supply services and suppliers: Q = Sqi. Each qi has its own costs and benefits associated with service quality improvements. Addressing such cost/benefit issues goes way beyond the scope of this study. However, utilities and marketers can perform analysis to estimate values of qualitative factors and to develop and calibrate consumer choice models. They can also assess expenses related to service quality improvements and use the results in market simulation models.
Load Growth: Discounts vs. Rebates
Unbundling and competition will also have an impact on total volume of natural gas sold to residential customers. Some industry representatives believe if marketers reduce gas supply prices, residential consumers will find gas more attractive, which in turn will "boost" throughput levels and revenues for LDCs. The model estimates this "boost" and compares it with potential effects of such traditional marketing tools as rebates.
Consider four scenarios:
1. Business-as-usual (base case). Customers see no reductions in natural gas rates and receive no rebates.
2. Discounts only. Marketers offer a 20-percent discount. Consumers have no other preferences between a utility and marketers. (In this case, as was shown above, in Figure 1, marketers control 61 percent of gas sales.)
3. Small rebate. Consumers receive a $200-rebate from the LDC for installing gas-fired space-heating equipment.
4. Large rebate. Consumers receive a $500-rebate from the LDC for installing gas-fired space-heating equipment.
Assume also that the "discounts only" case includes no other extra marketing activity. Such an assumption does not usually hold during residential pilots since utilities and marketers usually publicize programs. (Companies frequently offer cash incentives.) However, modeling outcomes of the pilot programs was not the goal here. The model is designed to measure the "pure" impact of competition on natural gas sales to residential customers.
The simulation shows competition increases annual natural gas sales additions by 15 percent (see Figure 3). Thus, if the residential market would normally grow by 2 percent a year, competition will increase the growth to 2.3 percent annually. This growth appears modest, however, when considering that marketers offering a 20-percent discount supplied 61 percent of the natural gas. In contrast, the "old-fashioned" rebates for installation of gas-fired equipment could raise throughput additions by 30 percent and 90 percent for scenarios 3 and 4, respectively. The message is very simple. Utilities that hope to get a free ride on marketing activities of competing third-party suppliers may be misguided; marketing will remain an important part of LDC operations. Quantitative models of competitive local energy markets will help utilities and marketers survive in the new competitive environment. Building such models requires studying consumers' behavior and quantifying consumers' perceptions toward prices and non-price factors. Pilot programs are a perfect opportunity to analyze consumers' behavior and to use that information to tune up market models. Energy companies cannot operate in the dark. They cannot afford to rely on guesses. Errors could prove too costly.
Alexander Lonshteyn, Ph.D., is a senior consultant with Boston Gas Co., and a partner at RudLon Associates Inc., a consulting and software development firm to energy companies. The firm has developed U-Analyst, a software system for modeling competitive energy markets. The views expressed in this article are not necessarily those of Boston Gas Co.
1. AGA, Providing New Services To Residential Customers: A Summary of Pilot Programs and Unbundling Initiatives 1997 Update, 1997. Available at www.aga.com/gio/ib97-03.html.
2. AGA, 1996 Gas Facts, A Statistical Record of the Gas Industry, 1995 data.
4. S. Lawrence Paulson, "Gas Restructuring: Can Distributors Repeat the Success of Pipelines?" Public Utilities Fortnightly, Oct. 15, 1997.
5. LDCs said they viewed marketers as "allies" working to boost throughput, and reported that activity by marketers led to the sign-up of additional LDC customers. See, The Future of the Natural Gas Industry (Staff Position Paper), appended to Notice Inviting Comments on Staff Report, Case 97-g-1380, Sept. 4, 1997 (N.Y.P.S.C.).
6. U-Analyst is a software system for modeling competitive energy markets. It was built around an end-use model simulating economic behavior of energy consumers. It is an unbundled model and assumes burner-tip prices paid by consumers are sums of supply and distribution prices. Supply prices are for delivering gas or electricity to city gates (em the job of brokers, marketers and utilities. Distribution prices are for fuel/energy distribution and transmission (em monopolistic domain of public utilities.
Reliance on an unbundled model allows U-Analyst to simulate both: (1) consumers' choices of energy suppliers (marketers, brokers and utilities) and (2) choices of fuels for equipment (traditional application of end use models). The system employs logit (decision choice) functions to estimate choice probabilities. In the first case, it takes into account information about energy costs associated with different supply alternatives and qualitative characteristics of suppliers (e.g., reliability). In the second case, the model considers capital investments pertaining to each fuel selection alternative, for instance, the cost of installing a gas water-heater versus an electric water-heater. It also evaluates related operational expenses and impact of such qualitative factors.
7. We recognize that the industry, regulators and other stakeholders need to resolve many technical and policy issues regarding how residential unbundling should take place in reality. Should LDCs exit the merchant business? Who should serve as a supplier of the last resort? However, despite unknowns, it is likely that LDCs will continue to provide merchant services for at least three to five years.
8. Boston Gas, Long-Range Resource and Requirements Plan, Aug. 1, 1997, Mass. DPU 97-81, Attachment D, Boston Gas Attribute Trade-Off/Deregulation Study.
9. There is some anecdotal evidence that in some pilot programs customers have shown a negative attitude toward an incumbent LDC, but we regard this as an exception rather than a rule.
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.