Fortnightly
Competitive Efficiency: A Ranking of U.S. Electric Utilities
June 15, 1997
By Hossein Haeri, M. Sami Khawaja, and Matei Perussi
Do mergers and "critical mass" really make a difference? The answer, it seems, is yes.
To become more competitive, U.S. electric utilities have embarked on a quest in recent years to improve operational efficiency and factor productivity. The question is: Are utilities making progress? And, which companies have gained a competitive edge? Which have not?
Industry analysts have long argued that given the structure of the markets they serve and their cost-based, rate-setting procedures, electric utilities tend toward monopolistic behavior. Consequently, they are prone to wasteful applications of resources, especially overcapitalization. Without proper incentives, the argument went, utility managers have little motivation to cut costs or improve efficiency. As Hicks has argued, they would be more likely to exploit their market power by not bothering to approach maximum efficiency. "The best of monopoly profits," Hicks suggests, "is a quiet life."
These arguments, however, are waning quickly as the bang and clatter of competition disturbs the utility manager's "quiet life." Prompted by the discipline imposed by competitive markets and the demands of incentive regulation, utilities are paying increasing attention to the economic fundamentals of electricity production and delivery.
An examination of efficiency improvements at U.S. utilities, as measured by megawatt-hours per employee, reveals a modest increase (0.5 percent per year) between 1990 and 1995, mostly after 1993. This has led to moderately lower average system rates (see Figure 1). Variable expenses have declined in nearly all categories of operation and maintenance, fuel and labor. Price stability in the oil markets and better procurement practices also have helped control fuel input costs. In fact, labor productivity has shown steady annual improvements of more than 6 percent per annum, increasing from 4,670 MWh per employee (1990) to 6,420 MWh per employee (1995).
We have estimated the operational efficiencies for 94 U.S. electric utilities from 1990 to 1995 using conventional statistical techniques. As might be expected, the patterns that emerge appear to show some link between operational performance and geographic location. Also, to lend credence to the current "merger mania," we found that size of operation (and the fact of the merger itself) does appear to act as a significant determinant of overall efficiency.
Measures and Models
One measure of operational efficiency is productivity \(em the ratio of outputs to inputs. Productivity among firms can vary due to several factors, however, such as
differences in production technologies, environments in which production takes place and efficiencies of the production processes. A firm is efficient if it cannot increase its output without adding more inputs; or, conversely, if it cannot decrease the quantity of its inputs without reducing its output.
Productive efficiency has two components: technical and allocative. The technical component marks the ability to produce as much output as possible with available inputs, or using as little input as possible to produce the same level of output. The allocative component tracks the ability to combine inputs and outputs in optimal proportions under prevailing prices. In other words, it is the flexibility to adjust the mix of inputs as their prices change. Here, we measure overall operational efficiency without breaking it into components. %n1%n
Methods for measuring efficiency can be divided into two families, each comprising several specific techniques. One group of measurement techniques relies on mathematical programming. Using observed outputs and inputs for a group of firms, the algorithm calculates a measure of how efficient each firm is in converting inputs into outputs. This calculation is done by constructing a production "frontier" and measuring each firm's distance from it. %n2%n The other family is econometric. This family involves applying regression techniques to calibrate a production function that compiles information on inputs, outputs and other production characteristics of a group of firms over one or more periods. Each firm's efficiency is measured by comparing it with other firms in the group.
In general, efficiency is almost always measured in relative terms, comparing one firm with another firm or with an industry average (benchmarking). A firm can also be compared with itself at different times (trend analysis), or its performance can be evaluated against its goals (goal or "gap" analysis). The difference between efficiency levels under the operationally best possible resource
allocation and the actual resource allocation is the degree of x-inefficiency \(em the familiar concept introduced by Harvey Leibenstein in 1966.
Our Approach
Utilities use technology to transform capital, labor, energy and materials into electricity. The physical relationship between the amounts of each input and electricity produced can be expressed as a production function. In our analysis, we used a simple formulation of the production function known as the Cobb-Douglas. Under this formulation, output, measured in MWh, depends on capital, labor, fuels and materials used by utilities. A load factor variable was included to account for idle capacity. A trend variable was used to capture the time-varying effect of technology. %n3%n
Except for the Producer Price Index, which came from the Bureau of Labor Statistics, all other data came from Edison Electric Institute's Uniform Statistical Reports. Data were gathered on each variable from 1990 through 1995. We chose the holding company as the analysis unit rather than the operating company. Mergers during the data period were aggregated into single holding-company level. The analysis began with the complete database for all EEI member utilities. Only utilities with complete data for all variables in all six years were kept. This criterion left 94 observations for use in the analysis.
Output was measured as total physical production in MWh sold to all accounts (Schedule 14). Input variables were capital, labor, fuel, operating expenses and load factors. Fuel inputs were total outlays for all fuels in real dollars (Schedule 14). Operating expenses were the sum of all expense accounts and included operation, maintenance, depreciation, depletion, amortization and property losses, excluding local taxes (Schedule 2). Annual load factors were obtained from Schedule 17. All monetary variables were expressed in real terms, deflated by the PPI.
Leaders and Laggards
The statistical results from calibrating the production function showed that all included variables affected output and, together, explained more than 99 percent of its variations. %n4%n Estimated efficiency rankings and percentage changes in overall relative efficiency from 1990 to 1995 for the 94 companies are listed in Table 1. From 1990 to 1995, American Electric Power, Washington Water Power, and Southwestern Public Service Co., followed narrowly by Allegheny Power and PacifiCorp, led other utilities in the group in average efficiency.
Bangor Hydro-Electric Co., Upper Peninsula Energy, and Maine Public Service Co. scored the lowest, lagging the leaders nearly 22 percent. In interpreting the figures, it should be noted these are normalized scores and represent relative rankings rather than absolute efficiencies. In other words, scores of 100 and 99 for AEP and PacifiCorp, respectively, should not be construed as the actual operational efficiencies for the two utilities. Instead, the figures mean that over the five-year period, Idaho Power has been, on average, 1 percent more efficient than PacifiCorp.
Comparing the top three performers with the bottom three, marked differences emerge between the groups regarding location and size, as measured in MWh sales. The differences in rates are most striking. During the five years of the analysis period, the average system rates for the bottom three utilities were almost exactly double the average rates of the top three. The best performers are much larger than the worst, and are concentrated in the Northwest. Marked differences between the two groups are apparent in several important dimensions, including labor productivity, average operating expenses and, especially, percentage of purchased power.
Six of the 10 top performers are in the Pacific Northwest; eight of the 10 bottom performers come from the Northeast. The data show that, compared with the top three utilities, on average, the bottom three utilities lag in sales per employee by nearly a 3-to-1 margin, and purchase a far greater portion of their power from outside sources. The bottom group also has slightly higher proportions of residential customers. No apparent differences emerge between the two groups regarding wages (Table 2).
Close examination of utility efficiency scores reveal several important patterns, as shown in Table 3. Size of the operation is a significant determinant of efficiency and matters considerably in overall rankings. It shows a strong relationship with efficiencies due to economies of scale. The results suggest as much as a 5-percent difference in efficiency between utilities in the largest group and those in the smallest group.
Contributions of economies of scale to efficiency are also apparent when we consider company structure (individual operating company vs. holding company). For example, holding companies show slightly higher efficiencies than individual operating companies. More important, five of the six holding companies resulting from mergers during 1990-1995 show above-average efficiency gains. The one-half of the utilities in the sample that are combined operations show slightly higher efficiencies, resulting possibly from economies of joint production.
Northwest utilities lead in overall efficiency. Southeastern, Southern and North-Central utilities follow the Northwest by a high 5-percent margin. A utility's reliance on nuclear generation, measured as nuclear fuel outlays, also shows a strong negative correlation with efficiency; the higher the share of nuclear fuel costs, the lower the operational efficiency. Inversely, we find a strong relationship between operational efficiency and the share of hydroelectric power in a utility's generation mix.
The Incentive to Improve
The efficiency by which a utility uses its resources directly influences its profitability. In fact, increased productivity may be the most important determining factor in a utility's operations for both regulated and competitive markets. Judging by current trends, there is little doubt that those functions of electric utilities that remain regulated will be subject to incentive ratemaking in one form or another. In all forms of incentive regulation, retained earnings are largely decided based on their specific factor productivities (partial incentive mechanisms) or overall efficiency gains (price cap formulas). It seems, therefore, reasonable to expect utilities will have every incentive to improve their efficiency by closely monitoring operations and controlling costs.
Efficiency also bears directly on price, determining the utility's ability to compete in commodity markets. Our study suggests a close association between
efficiency rankings and average system rates for utilities in the sample (Figure 2). In fact, the results suggest efficiency scores account for more than 60 percent of the variations in average system rates.
As John Kenneth Galbraith has said, "Things that are measured tend to improve." Operational efficiency has never been more important for electric utilities than it is today, as they embark on the new era of retail access and
competition. As competition
intensifies, market pressures will inevitably force prices toward marginal costs, leading to shrinking margins and a greater demand for operational efficiency. Productive efficiency will emerge as the survival condition in a competitive environment.
Hossein Haeri, M. Sami Khawaja and Matei Perussi and are economists in the Portland, Ore., offices of Barakat & Chamberlin Inc., a consulting firm that provides technical and strategic services to the utilities industry.
Table 1. Relative Efficiency Rankings for 94 Electric Utilities
Relative
Efficiency
Relative Change from
Rank Utility Efficiency '90 to '95
1 American Electric Power Co. 100.00% 1.62%
2 Washington Water Power Co. 99.99% -2.26%
3 Southwestern Public Svc. Co. 99.53% 2.33%
4 Allegheny Power System 99.36% -2.02%
5 PacifiCorp 99.21% 0.96%
6 Idaho Power Co. 99.17% 1.41%
7 Kentucky Utilities Co. 98.50% 4.11%
8 Portland General Electric Co. 97.50% -0.74%
9 Puget Sound Power & Light Co. 97.41% -0.89%
10 Minnesota Power 96.84% 0.89%
11 Southern Co. 96.67% 1.11%
12 Northern States Power Co. 96.54% 0.80%
13 Montana Power Co. 96.50% -0.79%
14 Louisville Gas and Electric Co. 96.11% 1.59%
15 Cincinnati Gas & Electric Co.,
Cinergy Corp. * 95.94% 4.38%
16 Union Electric Co. 95.93% 5.65%
17 Central and Southwest Corp. 95.76% 2.41%
18 Texas Utilities Co. 95.72% 1.92%
19 Duke Power Co. 95.06% 3.22%
20 Ipalco Enterprises 95.03% 2.27%
21 Kansas Power and Light Co.,
Western Resources * 94.55% 1.70%
22 Oklahoma Gas and Electric Co. 94.53% 1.32%
23 So. Indiana Gas & Electric Co. 94.44% 2.01%
24 Houston Lighting & Power Co. 94.29% 2.21%
Relative
Efficiency
Relative Change from
Rank Utility Efficiency '90 to '95
25 Scana Corp. 93.68% 1.16%
26 Entergy Corp. 93.67% 0.98%
27 Virginia Electric and Power Co. 93.50% 2.50%
28 Wisconsin Power and Light Co. 93.44% 2.44%
29 Iowa Power, Midwest Power,
MidAmerican * 93.38% 5.71%
30 Dayton Power and Light Co. 93.05% 2.08%
31 Carolina Power & Light Co. 92.91% 2.32%
32 Wisconsin Public Service Corp. 92.81% 3.31%
33 Empire District Electric Co. 92.65% 0.68%
34 Kansas City Power & Light Co. 92.51% 8.31%
35 Public Service Co. of Colorado 92.48% 1.53%
36 Gulf States Utilities Co. 92.39% 4.22%
37 Pennsylvania Pwr. & Light Co. 92.33% 2.84%
38 Cipsco,
Central Illinois Public Service * 92.32% 9.05%
39 Potomac Electric Power Co. 92.06% -0.27%
40 Interstate Power Co. 91.90% 0.98%
41 Illinois Power Co. 91.87% 8.34%
42 Florida Power Corp. 91.66% 0.81%
43 Iowa-Illinois Gas & Electric Co. 91.59% 1.48%
44 Consumers Power Co. 91.57% 0.90%
45 Nevada Power Co. 91.51% -0.18%
46 Otter Tail Power Co. 91.50% 4.50%
47 Detroit Edison Co. 91.25% 2.93%
48 Tampa Electric Co. 91.10% -0.58%
Relative
Efficiency
Relative Change from
Rank Utility Efficiency '90 to '95
49 Commonwealth Edison Co. 91.01% 2.83%
50 Ohio Edison Co. 90.95% 1.77%
51 Baltimore Gas and Electric Co. 90.74% 6.79%
52 Central Illinois Light Co. 90.64% 3.21%
53 Central Louisiana Electric Co. 90.57% 2.32%
54 Delmarva Power & Light Co. 90.41% 4.35%
55 NIPSCO Industries 90.10% 3.98%
56 St. Joseph Light & Power Co. 89.71% 4.25%
57 Utilicorp United 89.65% 3.45%
58 Iowa Electric Light & Power Co.,
IES Utilities* 89.04% 12.13%
59 New York State
Electric & Gas Corp. 88.83% 1.35%
60 Philadelphia Electric Co.,
PECO Energy Co. * 88.75% 6.73%
61 General Public Utilities Corp. 88.60% 0.98%
62 Public Svc. Enterprise Group 88.42% 1.32%
63 Arizona Public Service Co. 88.27% 1.73%
64 Niagara Mohawk Power Corp. 87.75% 0.52%
65 MDU Resources Group 87.23% 1.34%
66 Centerior Energy Corp. 86.86% 5.50%
67 Duquesne Light Co. 86.84% 4.30%
68 Pacific Gas and Electric Co. 86.77% -0.67%
69 Sierra Pacific Power Co. 86.71% 0.32%
70 Northwestern Public Svc. Co. 86.33% 5.90%
71 Public Svc. Co. of New Mexico 86.26% 9.38%
Relative
Efficiency
Relative Change from
Rank Utility Efficiency '90 to '95
72 Cent. Hudson Gas & Elec. Corp. 85.98% 0.61%
73 Tucson Electric Power Co. 85.82% 6.44%
74 So. California Edison Co. 85.78% 1.37%
75 El Paso Electric Co. 85.77% 4.87%
76 New England Electric System 85.45% 0.45%
77 Commonwealth Energy System 85.25% 5.02%
78 San Diego Gas & Electric Co. 85.22% 1.54%
79 Green Mountain Power Corp. 85.18% -2.17%
80 Northeast Utilities 85.13% 3.38%
81 Rochester Gas & Electric Corp. 84.99% 1.37%
82 Black Hills Corp. 84.90% 0.67%
83 Long Island Lighting Co. 84.60% -3.68%
84 Cent. Vermont Public Svc. Corp. 84.30% 4.73%
85 United Illuminating Co. 83.88% 3.54%
86 Orange and Rockland Utilities 83.41% 7.63%
87 Consolidated Edison Co.
of New York 83.25% 3.25%
88 Boston Edison Co. 82.97% 1.75%
89 Central Maine Power Co. 82.87% 2.08%
90 Hawaiian Electric Co. 81.31% -1.78%
91 Eastern Utilities Associates 80.85% 5.42%
92 Maine Public Service Co. 80.08% 0.83%
93 Upper Peninsula Energy Corp. 78.44% 1.39%
94 Bangor Hydro-Electric Co. 78.32% -1.29%
Average Efficiency 90.49% 2.47%
*Companies merged.
Table 2. Comparison of Top and Bottom Performers
Variable Top 3 companies Bottom 3 companies
Upper Bangor
Washington Southwestern Maine Peninsula Hydro
Water Public Public Energy Electric
AEP Power Service Service Corp. Co.
Total Sales (MWh) 116,196,875 10,558,467 19,084,259 664,623 808,215 1,725,870
% Residential Sales 24% 29% 13% 26% 31% 30%
% Industrial Sales 36% 15% 39% 20% 28% 51%
Average System Rate 0.05 0.04 0.04 0.09 0.07 0.10
Salary per employee 45,755 48,301 43,412 37,006 46,468 40,278
Total Sales (MWh)/Employee 6,408 10,335 9,403 3,675 1,495 3,469
Plant in Service ($1000s)/MWh 0.16 0.14 0.12 0.12 0.20 0.16
Percent Purchased Power 4% 42% 2% 84% 81% 81%
Operating Expense ($1000s)/MWh 0.03 0.03 0.03 0.07 0.06 0.08
Load Factor 0.63 0.60 0.63 0.64 0.71 0.76
Table 3. Comparison of Efficiency by Various Categories
Number
of utilities Average
Variable Category in category Efficiency
Size (Average MWh) Small = Quartile 1 23 86.7%
Medium= Quartile 2 24 90.8%
Large = Quartile 3 23 92.0%
Very Large = Quartile 4 24 92.4%
Region Northwest 4 98.5%
West 16 89.6%
North-central 21 92.2%
Central / Midwest 9 90.5%
South / Southeast 13 93.9%
East / Northeast 31 87.3%
Nuclear fuel: 0% 39 91.1%
Percent of total fuel cost 1-10% 23 91.0%
10-20% 19 89.7%
20-30% 7 88.6%
30-40% 5 89.0%
over 40% 1 88.7%
Purchase Power - <25 % 52 92.3%
% of total sales 25-50% 32 89.6%
50-75% 5 86.8%
over 75% 5 81.3%
Utility with gas sales Yes 47 90.7%
No 47 90.3%
Percent Industrial 0-20% 22 89.1%
20-40% 60 91.0%
over 40% 12 90.0%
Holding Company Yes 30 91.6%
No 64 90.0%
Hydro electric % of sales 0 39 90.0%
0-10% 47 91.0%
over 10% 8 90.0%
1Several econometric techniques have been developed for obtaining the measurement of each component. The computational procedures, however, are complex and inexact.
2One study employing this technique was published in PUBLIC UTILITIES FORTNIGHTLY. (See, "The Efficient Utility: Labor, Capital, & Profit," by D. Thomas Taylor and Russell G. Thompson, Sept. 1, 1995, p. 25.) That study used Data Envelopment Analysis, a mathematical programming technique, to estimate relative efficiencies of 13 investor-owned utilities. Some of that study's flaws and certain weaknesses of its methodology were later noted by Matthew Morey and L. Dean Hiebert. (See, "Measuring Utility Efficiency: A New Frontier" [letter to editor], PUBLIC UTILITIES FORTNIGHTLY, Jan. 1, 1996, p. 7.)
3The estimated equation was formulated as:
Ln(Yit) = Siai +SjbjLn(Xijt) + LFit + Îit where Ln(Yit) is the natural logarithm of total output in megawatt hours, Ln(Xijt) is natural logarithm of a set of j inputs (labor, capital, fuel and material), LF is the load factor, and T is a trend variable with values of 1 to 6 representing each year of data from 1990 to 1995. Index i refers to utilities, and index t refers to time periods.
Ît is an error term representing two elements: statistical noise (vit) and inefficiency (ui): Îit = vit + ui). The decomposition of the error term into its two components may be done in several ways. The fixed effects approach assumes differences in the efficiency of different utilities are captured in their respective intercepts by the term (ai) in the above equation. That is, had all utilities used the same amount of each input, all differences in output levels would be represented in the intercept.
In estimating the efficiency level associated with each utility, the most efficient utility would be defined as the one with largest intercept. In other words, the most efficient utility represents 100-percent efficiency, and all other utilities are compared to it.
4The data, estimation results and summary statistical properties in SAS output format are available from the authors by request.
26
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.
Public
Utilities Fortnightly & Spark
