Citation
A Comparative analysis of the economic organization of gas and electric supply

Material Information

Title:
A Comparative analysis of the economic organization of gas and electric supply
Creator:
Anderson, John Angus, 1943-
Publication Date:
Language:
English
Physical Description:
x, 115 leaves. : illus. ; 28 cm.

Subjects

Subjects / Keywords:
Customers ( jstor )
Economic competition ( jstor )
Electric utilities ( jstor )
Gas utilities ( jstor )
Marginal cost pricing ( jstor )
Marginal costs ( jstor )
Prices ( jstor )
Standard deviation ( jstor )
Utilities costs ( jstor )
Utility rates ( jstor )
Dissertations, Academic -- Economics -- UF
Economics thesis Ph. D
Electric utilities -- Finance ( lcsh )
Gas manufacture and works -- Finance ( lcsh )
City of St. Petersburg ( local )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis -- University of Florida.
Bibliography:
Bibliography: leaves 152-155.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
This item is presumed in the public domain according to the terms of the Retrospective Dissertation Scanning (RDS) policy, which may be viewed at http://ufdc.ufl.edu/AA00007596/00001. The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator(ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
022837702 ( ALEPH )
14145426 ( OCLC )

Downloads

This item has the following downloads:


Full Text











A COMPARATIVE ANALYSIS OF THE
ECONOMIC ORGANIZATION OF GAS
AND ELECTRIC SUPPLY










BY


JOHN ANGUS ANDERSON


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY







UNIVERSITY OF FLORIDA 1973















ACKNOWLEDGMENT S


The writer wishes to express his deepest appreciation to the members of his supervisory committee. A special thanks is offered to Dr. Milton Z. Kafoglis, Chairman, who suggested the subject and devoted countless hours in editing the manuscript. Without his help and constant motivation, this dissertation would not have been completed. Drs. Charles W. Fristoe, Norman G. Keig and John H. James each encouraged the writer when progress was slow and offered invaluable suggestions throughout the study. Finally, I wish to acknowledge the loyalty of my wife, Peggy, who willingly sacrificed during the preparation of this paper.















TABLE OF CONTENTS



ACKNOWLEDGMENTS ...............

LIST OF TABLES .. .. . . . . . ... . . .

LIST OF FIGURES . .............

ABSTRACT ........ ...................


I. INTRODUCTION ... ..........
Economic Theory and Public
Economics .....
Outline of the Study . . .
Economic Issues . . .
Related Research . . .
Aggregative Analysis A Regional Analysis.
Policy Implications.

II. ECONOMiC ISSUES .......
Economies and Diseconomies
Joint Costs .....
Load Factors......
Competitive Stimulus Costs of Divestiture
Summary .......
Marginal Cost Pricing. . �


Utility


* . . . vii

. . . . viii



2 6 6 7 9 9
* . . . 10


The Marginal Cost Pricing Principle Marginal Costs and the Energy
Industry ...........
Pricing in the Energy Industry . . Sunrnary ...... ......


III. EVALUATION OF EARLIER WORK .. ...........
The Study by Bruce M. Owen .. ........
The Study by Patrick C. Mann .........
The Study by William H. Collins .
The Study by the N.E.R.A ... .........
Summary of Owen, Mann, Collins and
the N.E.R.A. . . . . . . . . . . .
The Study by Joe D. Pace ... .........
Summary of Earlier Work ........








Page


AGGREGATIVE ANALYSIS .... ...........
The Data ..... ...............
Methodology ..... ..............
Table Notations ... ...........
Aggregative Characteristics of the
Firms ...... ...............
Findings of the Aggregative Analysis
Economies and Diseconomies . . .
Marginal Cost Pricing .........
Summary ..... ..............


V. THE REGIONAL ANALYSIS ........
The Regional Distribution ....
Aggregative Characteristics of the
Firms ..... ..............
Findings of the Regional Analysis.
Economies and Diseconomies . .
Marginal Cost Pricing .......
Summary of the Regional Analysis .


VI. POLICY IMPLICATIONS ..........
Data and Methodology ... .........
The Aggregative Analysis ........
The Regional Analysis .......
Comparison of Expected Benefits and
Costs . . . . . . . . . . . . .
Alternative Policy Suggestions

APPENDICES

A. RESULTS OF THE STATISTICAL COMPUTATIONS:
THE AGGREGATIVE ANALYSIS ........

B. RESULTS OF THE STATISTICAL COMPUTATIONS:
THE REGIONAL ANALYSIS .......

BIBLIOGRAPHY ................


. . .. . 87
. . .. . 88

. . .. . 91
.. 4.. . 94
. . .. . 94
. . .. . 104
110














LIST OF TABLES


Table Page

2.1 A MEASUREMENT OF THE SUBSTITUTABILITY OF GAS AND ELECTRICITY .. ........... .18

2.2 RELATIONSHIP BETWEEN GAS-ELECTRIC PRICE AND SALES RATIOS FOR THE INDUSTRIAL
SECTOR, 1962 AND 1967 .. .......... .20

4.1 THE RESULTS OF TIE SURVEY OF THE UTILITIES 55

4.2 AGGREGATE CHARACTERISTICS OF THE UTILITIES:
DEMOGRAPHIC AND SALES VARIABLES ..... 62

4.3 AGGREGATE CHARACTERISTICS OF THE UTILITIES:
EXPENSE, PLANT AND TAX VARIABLES . . . . 63

4.4 AGGREGATE CHARACTERISTICS OF THE UTILITIES: EMPLOYMENT AND FINANCIAL VARIABLES . . . 66 4.5 AVERAGE CONSUMPTION AND PRICE . ........ 68 4.6 AVERAGE COSTS ................70

4.7 AVERAGE WAGES ..... ................ 71

4.8 PHYSICAL PLANT ..... ................ .73

4.9 FINANCIAL ...... .................. 75

4.10 RESULTS OF LINEAR REGRESSION WHERE THE
SLOPE (b) ESTIMATES THE MARGINAL
COST ...... .................. 79

4.11 AVERAGE PRICE AS A PERCENT OF LONG RUN
MARGINAL COST .... ............. .82

5.1 THE GEOGRAPHICAL DISTRIBUTION OF THE UTILITIES ..... ................ .90








Table Page

5.2 AGGREGATE CHARACTERISTICS OF THE UTILITIES: DEMOGRAPHIC AND SALES VARIABLES ..... 92

5.3 AGGREGATE CHARACTERISTICS OF THE UTILITIES:
EXPENSE, PLANT AND TAX VARIABLES . . . . 93

5.4 AGGREGATE CHARACTERISTICS OF THE UTILITIES: EMPLOYMENT AND FINANCIAL VARIABLES . . . 95 5.5 AVERAGE CONSUMPTION AND PRICE . ....... 97 5.6 AVERAGE COSTS ..... ................ 99

5.7 AVERAGE WAGES ..... ............... 101

5.8 PHYSICAL PLANT ..... ................ .102

5.9 FINANCIAL ....... ................ .103

5.10 RESULTS OF LINEAR REGRESSIONS WHEN THE SLOPE
ESTIMATES THE MARGINAL COST ....... 105

5.11 AVERAGE PRICE AS A PERCENT OF LONG RUN
MARGINAL COST .... .............. 108















LIST OF FIGURES



Figure Page

1. THE SHORT RUN/LONG RUN MARGINAL COST
RELATIONSHIP .... .............. .28

2. PRICING UNDER CAPACITY CONSTRAINTS:
THE CASE OF A SINGLE PEAK CONSUMER . . . 30

3. PRICING UNDER CAPACITY CONSTRAINTS:
THE CASE OF BOTH PEAK AND OFF-PEAK
CONSUMERS .... ............... .31

4. ESTIMATES OF MARGINAL COSTS .......... .80

5. FEDERAL POWER COMMISSION NATIONAL POWER SURVEY REGIONS .... ............. .. 89















Abstract of Dissertation Presented to
the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy



A COMPARATIVE ANALYSIS OF THE ECONOMIC ORGANIZATION OF GAS AND ELECTRIC SUPPLY


by

John Angus Anderson

August, 1973


Chairman: Milton Z. Kafoglis Major Department: Economics


A competitively organized economy is expected to result in a more efficient allocation of resources and more efficient production. The public utility sector possesses unusual or exceptional characteristics that make both perfect competition and unregulated monopoly undesirable. Senator Metcalf has introduced a bill in the U. S. Senace (S. 403) that seeks to create a blend of competition and regulation in the energy sector. This bill would divorce combination gas/electric utilities, presumably increasing the degree of competition in the energy industry; however, it would not abandon traditional regulation. This thesis examines the hypothesis implicit in the Metcalf bill that separate ownership of gas and electricity supplies will in fact generate economies associated with competition.


viii








Certain economies of dual or joint operation are available to combination utilities that are not available to single service utilities. To the extent that these economies are significant, they are expected to lowr the costs of the combination firms relative to the single service firms. On the other hand, joint utilization of resources encompasses only a small portion of total operating expenses. Thus, average total costs may not be significantly affected if the dual utilization of the general resources is lost. Separate ownership of the sources of energy is also predicted to promote a competitive stimulus that sharpens the decision making process of single service finns. However, managers of separate firms may be prompted to charge discriminatory prices in an attempt to gain sales at "destructively" low rates. Previous studies have conflicting conclusions and policy recommendations. It is this conflict that establishes the need for this thesis.

Data are gathered from reports submitted directly by

each utility. These data are more detailed and comprehensive than those available in existing publications. The data are subjected to "z" and "t" tests which are widely known and commonly accepted as a basis for evaluating differences in variables. Few differences are observed in the relative sizes of the single service utilities and the electric operations of combination utilities; the single service gas utilities appear significantly larger than the gas operations of combination utilities.

When the data are analyzed on an aggregative basis,

rates are lower and the volume of sales higher for single service firms than for combination gas/electric firms. Sales expenses are








greater, although average total costs are lower, for single service firms. Single service firms discriminate between types of customers to a greater extent'than combination firms. Thb findings conform to those reached in earlier studies which incorporated similar data and methodology.

However, the distribution of the two types of utilities is not geographically uniform. Specifically, single service utilities are concentrated in the south and in the west where climate causes high per-customer consumption. In response to this spatial difference, the data are subdivided and studied on a geographic basis. Most of the differences that are observed as significant on a national basis are not significant on a regional basis. The findings of the national analysis that appeared to indicate that competitive forces in the energy industry are substantial now appear simply to indicate geographical or demographical characteristics peculiar to the individual firms and regions.

On the other hand, the legal and financial costs of

divestiture may be significant. Since the costs are expected to outweigh the benefits, the hypothesis examined is rejected. More stringent enforcement of present legislation that considers each case on its own merits is more realistic.












CHAPTER I

INTRODUCTION



The energy industry is characterized by different industrial organizations in various geographical areas of the United States. In some areas the supply of both electricity and natural gas is provided by a single firm (a "combination" utility); in other areas the supply of electricity is provided by one firm while the supply of natural gas is provided by another firm (a "single service" or a "straight" utility). Certain potential economies are predicted where combination utilities exist, but in a regulated atmosphere there may be no motive to attain fully these economies. On the other hand there is a greater element of "potential" competition where single service utilities exist, but this organization may frustrate the attainment of certain economies. This thesis examines the comparative efficiency of alternative organizational arrangements that have been proposed recently regarding the production and the distribution of electric power and natural gas.








Economic Theory and Public Utility Economics

In an enterprise economy, a multitude of privately owned and managed busineas enterprises are expected to play a central role in the determination of the course and the character of economic activity. Economic theory predicts that a competitively organized economy results in an efficient allocation of resources in the sense that the standard Paretian conditions are fulfilled. The encouragement of competition is a stated objective of public policy in the United States. However, certain segments of the economy, especially the public utility industries, possess unusual or exceptional characteristics and have had distinctly monopoly policies applied to them. Public utilities produce widely used "necessities" for which there are no close substitutes and are, to some degree, natural monopolies. These industries are held to be natural monopolies because (a) advantages to large-scale firms are such that unrestricted competition would lead to exclusive occupancy of individual markets by one or a very few firms, and/or

(b) unrestricted competition among several or many firms in a market would, because of the technological characteristics of the

-industries involved, have deleterious effects on the welfare of ,buyers. The energy industry has traditionally been considered

--a public utility and special regulatory policies have been applied to this sector of the economy. Legislative bodies have granted ,exclusive market regions and have blocked entry in order to


1J. S. Bain, Industrial Organization (New York: John Wiley and Sons, Inc., 1968), page 581.





-3

guarantee monopoly organization, presumably on the assumption that available economies of scale and other characteristics of production and distribution are inconsistent with competitive organization. Protection of the consumer has been shifted to the regulatory agencies.

Criticism of the effectiveness of regulation abounds.
2
Joe Bain, for example, has summarized the deficiencies as follows:

(1) Commissions have not been able to consistently impose rates at a competitive level.

(2) Regulatory commissions in general do not have the power to determine broadly the rate of output or the scale of operations of the regulated firms, thus abandoning important areas to managerial discretion.

(3) Regulation seriously reduces the usual profit and efficiency incentives which exist in the private sector.

(4) By implication, regulation has the responsibility of supporting a rate structure that will secure for firms a rate of return as determined by statutes, courts and commissions. This policy may perpetuate the provision of redundant services and may forestall economically desirable adjustments in the relative volumes of services supplied or facilities maintained by different components of the industry.

Though regulation has not been ideal, unrestricted competition does not appear to be an attractive alternative.


J. S. Bain, Industrial Organization, pages 640-645.





-4

A number of economists have maintained that unregulated enterprise in the public utility sector is likely to be as, if not more, effective than regulated enterprise. For example, George Stigler and Claire Friedland found that regulation of electric utilities
3
was redundant and ineffective for two reasons. First, individual utilities do not possess significant long-run monopoly power. Instead, they face competition from other energy sources in a large porportion of their products uses, and they face competition from within their own industry. In the long run, their industrial users (and hence many of their domestic users) have sufficient mobility to ensure a competitive outcome. Second, regulatory bodies are incapable of forcing the utility to operate at a specific (presumably optimal) combination of output, price and cost. Stigler and Friedland state: "The theory of price regulation must, in fact, be based on the tacit assumption that, in its absence, a monopoly has exorbitant power."4 These authors conclude that electrical utilities do not possess such power.

Since neither of the organizational extremes appears to be feasible, alternative suggestions incorporating a blend of competition and regulation have been proposed. Thus, the transportation industry is regulated but in an environment of "balanced competition" where entry is controlled but not closed. It has been argued that a similar blend of regulation and competition should be


3G. J. Stigler and C. Friedland, "What Can Regulators Regulate? The Case of Electricity," The Journal of Law and Economics, Volume 5 (1962), pages 15-16.

G. J. Stigler and C. Friedland, "What Can Regulators Regulate? The Case of Electricity," pages 15-16.








applied to the energy sector. In response to this view, Senator Metcalf introduced a bill (S.403, 92nd Congress, First Session) that ammends the Federal Power Act. In this bill it is declared that: ". . . it is in the national interest to promote interenergy competition between electricity and gas whenever possible." 5 This bill makes it unlawful to own or to operate facilities used in the production, generation or distribution of both electricity and natural gas. It would divorce combination gas and electric firms, presumably increasing the degree of competition in the energy industry, but would not abandon traditional regulation. It would seek to add a "free enterprise" competitive dimension to the allocative process. To the extent that electricity and natural gas do in fact compete, the intensified competition would result in superior resource allocation.

This thesis examines the hypothesis implicit in the

Metcalf bill that: (a) separate ownership of gas and electricity supplies will in fact promote competition, and (b) this increased competition will encourage efficiency and lead to an improved rate structure. If the hypothesis is affirmed, the policy embodied in the Metcalf bill can be recommended. If the hypothesis cannot be affirmed or is refuted, this policy cannot be recommended. Our criterion, of course, is economic efficiency. Any conclusions developed must, therefore, be further evaluated in terms of the


5S. 403, "A bill to prohibit certain combinations and control between electric and gas utilities," as presented in Combination Utility Companies, Hearings before the Subcommittee on the Judiciary, United States Senate, Washington: U. S. Government Printing Office, 1971, page 1.





-6

broader objectives of social policy relative to economic organization and equity. Though the author's impressions and opinions on these matters are occasionally expressed, the scope of the substantial work is limited to a test of the hypotheses described above.


Outline of the Study

Economic Issues

In order to evaluate the hypotheses, a framework suitable for statistical testing must be established. This is the primary objective of Chapter II where economic theory is employed to predict possible outcomes under the alternative organizations. In Chapter II the expected economies and diseconomies of the "straight" utilities are compared with the relative economies and diseconomies of the "combination" utilities. Combination firms are predicted to have lower common costs due to joint utilization of assets and lower fuel costs due to higher load factors. These factors are predicted to lower the cost curves of combination companies relative to single service utilities.

On the other hand, single service firms have both the ability and the incentive to increase their load factors as much as do the combination gas/electric firms and the joint utilization of-resources may encompass only a small portion of the total operating expense. Moreover, it is sometimes asserted that regulatory agencies do not recognize technological developments in the energy industry and thus do not force the realization of potential efficiencies; competition, on the other hand, may force








such efficiencies. To the extent that these factors are significant, single service utilities are predicted to have lower costs and greater efficiencies than are the combination gas/electric utilities.

Furthermore, Chapter II establishes a basis for evaluating and comparing the rate structures of combination utilities and straight utilities. The marginal cost pricing principle is described, criticized and finally defended as an ideal criterion. However, to the extent that indivisibilities, externalities, and decreasing costs are significant, this criterion must be modified. The "second-best" theory is briefly discussed and accepted although, as a practical matter, a strict application of second-best is also unattainable. Finally, a realistic and attainable guide, developed in the light of both theory and practice, is established. Though crude, it appears to be consistent with the goal of welfare maximization.


Related Research

Previous studies have investigated the hypothesis with which this study is concerned. These works are characterized by conflicting conclusions and policy recommendations. Studies by B. M. Owen, P. C. Mann, W. H. Collins, Jr., the National Economic Research Associates and J. D. Pace are analyzed, evaluated and
6
compared in Chapter III. Owen, Mann, Collins and the NERA utilized data for the entire United States in an effort to compare the average values of relevant variables under the alternative


6Citations for these studies are in Chapter III.






-8

organizational arrangements. The conclusions of several authors are similar: straight utilities were found to charge lower average prices and sell larger quantities of electricity per customer than did the combination utilities. These findings have provided a basis for recommending that the combination utilities be divorced or separated into two competing units. This proposed divorcement of the ownership of the supplies of energy is expected to increase competition. The implication is that the advantages of competitive organization are greater than the advantages of joint operation and production.

Differences in geographical locale may cause differences in measured variables. Under cursory examination these measured differences may appear to be associated with competitive forces; in actuality they may be the result of geographical concentration and thus demographic influences. For example, if single service utilities are concentrated in warm regions, climate is expected to cause high per-customer consumption regardless of the organizational structure. A statistical analysis that segments data only by industrial organization ignores these geographical influences. Owen, Mann and Collins fail to consider the geographical location of the firms and thus fail to recognize the resulting demographic characteristics that each firm faces.

The Pace study did recognize the impact of demographic characteristics upon the findings of the national studies. This study utilized a complex interrelationship of demographic variables and found that the "clear" indications of beneficial competition were in fact insignificant. The conclusions of the Pace study






-9

were in conflict with those of Owen, Mann and Collins. It seems clear that additional research is necessary in order to reconcile these conflicting conclusions.


Aggregative Analysis

The analysis of Chapter IV provides a basis for comparing this study with other work. Certain variables are examined in an attempt to measure the degree to which combination utilities do in fact compete with straight utilities. The national data and the statistical methodology incorporated in this chapter do not differ significantly from those employed by other authors. Since the aggregative analysis incorporates methods and conclusions similar to those of previous studies, the data of this thesis are assumed to be similar to those of the previous works.

The findings in Chapter IV support the view that the

combined gas and electric utilities possess greater market power than the single service utilities. Competition between two straight firms, especially with respect to the supply of electricity, appears to force a greater degree of efficiency than the regulatory process itself can achieve. To this point, the policy proposal to require divestiture of gas and electric operations appears to have support. Certainly there does not appear to be justification for a policy that would allow the formation of any new combination utilities.


A Regional Analysis

In Chapter V the statistical methodology of Chapter IV

is applied to the same series of data, but on a geographical basis.






-10

The objective of the geographical break-down is to isolate the effects that demographic differences have on utilities in various parts of the United States. For example, as explained above, on a national basis single service utilities have lower prices and higher quantities. If single service utilities are concentrated in warm regions, climate is expected to increase consumption of energy. The larger consumption would, because of the step-down nature of the pricing structure, reduce the average revenue per unit of output. The conclusion reached above that the single service utilities have higher quantity sales and lower average prices is not altered. But this conclusion is now explained on a totally different basis.

If the conclusions of the regional analysis are not

significantly different from the national analysis, competition would appear to be a significant force. If, on the other hand, significant differences exist between the national and the regional studies, the existence of demographic differences must be acknowledged. Antitrust as a policy recommendation may not bring a reduction in price, increase in output or increase in efficiency since differences are caused by demographic characteristics that are independent of market structure and ownership. Chapter V supports the latter interpretation.


Policy Implications

Chapter VI brings the analysis together by comparing the expected costs with the expected benefits. The benefits of divestiture appear to be relatively insignificant; the costs appear to be






-11

potentially substantial. The hypothesis examined in this thesis is therefore not supported. The primary hypothesis that requires divestiture is rejected; the formation of new combination utilities, however, is not categorically accepted. Each case should be considered on its own merits.












CHAPTER 11

ECONOMIC ISSUES



The hypothesis that is investigated in this study suggests that the performance of an industry that competes (even though imperfectly) is superior to the performance of an industry that is monopolistic. In this chapter alternative types of organization are discussed and relevant economic theory is presented in order to develop a basis for the measurement of the economic performance of the various organizations.

A perfectly competitive economy with no external economies, joint supplies or indivisibilities achieves a Pareto optimal equilibrium. In a Pareto optimal equilibrium

(1) the total cost of the optimal output is at a minimum and

(2) the price is equated with marginal cost.1 Although rough, these criteria are utilized in this study as feasible yardsticks by which alternative organizations of the energy industry are compared. Each of these criteria are examined in turn.


IV. Pareto, Manuel d'Economigue Politique (2nd Edition, Giard, Paris, 1927). The marginal conditions necessary to achieve welfare maximization have been extensively developed since this original work and are beyond the scope of this study. See for example: N. Kaldor, "Welfare Propositions of Economics," Economic Journal, XLIX (September, 1939); P. Samuelson, Foundations of Economic Analysis (Cambridge: Harvard University Press, 1948); and especially M. A. Reder, Studies in the Theory of Welfare Economics (New York: Columbia University Press, 1947). These references are only a representative sample of numerous citations available.


-12-





-13

Economies and Diseconomies

Economic efficiency requires the minimization of costs in producing the optimal output. Certain economies and diseconomies are expected under alternative organizations of the energy industry. These economies are joint costs, load factors and the competitive stimulus; the diseconomy is the cost of divestiture.


Joint Costs

Combination gas/electric utilities are potentially

able to operate more efficiently with respect to their resources because they are able to utilize their administrative, general and other fixed resources in the distribution of both electricity and gas. For example, the customer accounting process, including meter-reading, bill preparation and handling of payments can be accomplished more efficiently by one company than by two. The reasons for the increase in efficiency are obvious: only one meter reader per customer (instead of two) is required; only one bill per customer (instead of two) is prepared; and only one payment per customer (instead of two) is received.2 Additionally,


2Frederick T. Searls, Vice President and General Counsel, Pacific Gas and Electric Company, estimated that it would cost an additional $15 million a year to carry on the entire customer accounting process, including meter-reading, bill preparation and the handling of payments by two separate companies in the place of his one company. (See Combination Utility Companies, Hearings Before the Subcommittee on Antitrust and Monopoly, 92nd Congress, Ist Session, 1971, Washington, U. S. Government Printing Office, at page 64.) Pacific Gas and Electric Company supplies gas and electricity to about 2 million customers which requires a mailing of 24 million bills each year. If this utility were divided into two companies, it would be necessary to increase the annual






-14

there exist potential economies in a single supervisory and administrative organization and a single set of offices, warehouses, service centers, and the like, instead of two. One application for service is more convenient for the customer as well as economical for the company, which can send one man, in one truck, to turn on both meters and initiate both services. In new installations, a single trench can be utilized for both electric and gas lines. Both gas and electric operations can be controlled from a single energy control center that contains weather forecasting, dispatch groups, and other related control functions. All these economies are expected to reduce average costs of production and increase the utilization of fixed facilities, permitting the utility to operate in a more optimal manner.

It is possible, however, that the competitive stimulus encouraged by separate ownership would produce greater overall gains in efficiency. If the joint costs are a significant proportion of the total costs, the net advantage is expected to accrue to the combination utilities. If the joint costs are only a small proportion of the total costs and the competitive stimulus is a significant force, the net advantage is expected to accrue to the straight utilities. mailing to 48 million bills. At the present rate of 8 cents for first-class mail, this would mean an additional cost of almost $2 million in postage for the mailing of the bills alone. Mr. Searls further stated that there are substantial economies to be realized in having a single supervisory and administrative organization and service facilities. He concluded, "We are confident that the actual savings are a very substantial amount." (at page 65)






-15

Soloman Freedi an, Director of the Division of Corporate Regulation, Securities and Exchange Commission, studied the expected savings due'to the joint utilization of resources as a percent of the total expenses and concluded:3

..... the bulk of the expenses of operating an electric
utility or gas utility is entirely unrelated to any
combined functions of the two.... It is in only a relatively narrow category that a combined gas and
electric company can effect savings in operating
expenses.. .the claimed "loss of economies" has not
been found, by the SEC, to be substantial.

This statement is in obvious contradiction with the statement cited in the footnote above. Mr. Searls asserted that the savings would be of a "substantial" magnitude; Mr. Freedman suggests that the economies have not been found. This contradiction will

-be examined in Chapters IV and V. Load Factors

Additional economies are potentially available to combined gas/electric utilities since the combined gas requirements for final customer use and for electric generation are under a common control. The load factor is defined as the average use of facilities as a percentage of the maximum use.4 Combination gas/electric 'utilities are predicted to have greater load factors, thus lower fuel costs and lower average total costs, than single service utilities. For example, assume that the temperature in an energy .system is presently cold, but expected to drop sharply for a short


3Combination Utility Companies, page 408.

4Paul J. Garfield and Wallace F. Lovejoy, Public Utility
Economics (Englewood Cliffs, New Jersey: Prentice-Hall, Inc., 1964), page 153.






-16

period. The gas controller estimates the increase in gas use that will result and calculates the amount of curtailment in gas fuel for electric generation that will be needed so that enough gas will be available to meet gas customer requirements. This information is given to the electric dispadther who determines whether to increase hydroelectric production or to burn oil in place of gas. It is predicted that full use is made of gas transmission facilities with this organization which results in a more optimal use of the system's resources.

On the other hand, single service electric utilities are able to negotiate interruptable contracts with (single service) gas companies at low rates. The gas companies increase their load factors (by the interruptable nature of the contract) and the electric companies lower their fuel costs. The potential significance of this economy therefore lies not in the market structure, but in the willingness and the ability of the single service utilities to negotiate fuel contracts.

An indication of this efficiency is the average dual cost per BTU of heat. In Chapters IV and V, fuel costs per BTU are compared as a measure of this economy. Competitive Stimulus

If gas and electricity are substitutable products, single service electric utilities compete with single service gas utilities for energy customers. The competition may force a sharpening of the-decision-making process and therefore a more efficient overall operation.






-17

When the Public Utility Holding Company Act and the

Federal Power Act were passed in 1935, the competitive interface between gas and electricity was far more limited than it is today. Additionally, both industries were more local in scope. After World War II, transcontinental pipelines and high voltage transmission lines transformed these two industries into truly interstate operations. Over recent decades, technological changes have heightened the potential competition between gas and electricity; the cross elasticity of demand for electricity and gas has increased.

One study that indicates the substitutability of gas

and electricity relates the monthly charge for 500 KWH in various metropolitan areas to the percentage of homes in each area using certain electric appliances. The higher the monthly charge for electricity, the lower is the proportion of homes with electrical appliances. Table 2.1 presents representative data from this study. The author stated: "Using a statistical sample of over eighty S.M.S.A.'s and the residential gas rate, I found that the cross elasticity value for water and space heating was in excess of unity."5 To a significant extent the gas and electric industries presently provide substitutable outputs and therefore effective competition between the suppliers of gas and electricity is anticipated.


5Dr. John W. Wilson, Department of Social Sciences, U. S. Military Academy, as quoted in Combination Utility Companies, page 99.






Table 2.1


A Measurement of the Substitutability of Gas and Electricity


Monthly
Price for Percentage of Homes with Electric S.M.S.A. 500 KWH: Ranges Water Heat Space Heat


Seattle $ 5.00 84.2 79.3 9.56 Portland 6.40 80.6 74.5 10.10 Nashville 6.90 74.7 65.7 34.48 Atlanta 8.60 33.6 14.2 .58 Birmingham 8.73 39.2 16.9 .88 Pittsburg 9.70 21.9 8.1 .14 Washington 11.59 21.6 8.0 .30 Baltimore 12.85 13.4 8.2 .35 Boston 13.56 25.5 8.2 .12 New York 14.24 7.6 1.5 .15


Source: Federal Power Commission, Typical Electric Bills 1963 and U. S. Department of
Commerce, Bureau of the Census, 1960 Census of Housing, as reported in
Combination Utility Companies, Hearings Before the Subcommittee on Antitrust
and Monopoly, United States Senate, 92nd Congress, First Session, 1971, page 99.






-19

David Smith recently studied the substitutability of gas and electricity by correlating the price ratios of gas and elec6
tricity with their respective sales ratios. Table 2.2 presents the findings of this study with regard to industrial customers. The elasticity of product substitution between gas and electricity exceeds unity at all relevant gas/electric price ratios and is
7
higher for the 1967 data than for the 1962 data. Dr. Smith concludes "that the increase in the elasticity of substitution between these two energy sources over this period of time can be attributed to intense efforts on the part of both gas and electric industries to penetrate each other's market in the industrial
8
sector."

Hubert H. Nexon, Vice President, Commonwealth Edison

Company, asserted: " . . competition itself tends to sharpen the decision making process. And conducting eithe the gas or the electric business is sufficiently complex to require undivided "9
management attention. Don Cook of the American Electric Power Association stated: "It's hard enough to run one company in one business well without trying to run two that are natural

6
Dr. D. B. Smith, "Inter-Energy Competition Between Gas
and Electricity in the Industrial Market: A Case for Deregulation?" Business Studies, North Texas State University, XII, No. 1, pages 8-9.
7The data were fitted to the hyperbolic form I/Y = a + bX, and the elasticity of product substitution was defined as: Eps= dY . X bX
X j a + bX
8D. B. Smith, pages 8-9.

9Combination Utility Companies, page 488.






Table 2.2


Relationship Between Gas-Electric Price and
Sales Ratios for the Industrial Sector, 1962 and 1967


1962
- .733**
.543


Coefficient of Correlation (R Coefficient of Determination (R) Elasticity of Product Substitution (Es,/e) with a Gas-Electric Price Ratio (Gp/Ep)* of: 14.00 10.00 8.00 6.00 5.00 4.00 3.50 3.00 2.50 2.00 1.75 1.50


1967
- .629
.396


-1.065
-1.093
-1.120
-1.166
-1.206
-1.272
-1.323
-1.399
-1.520
-1.747
-1.956
-2.225


-1.087
-1.126
-1.162
-1.229
-1.288
-1.388
-1.419
-1.594
-1.808
-2.266
-2.766
-3.920


*Gas measured in therms and electricity in kilowatt hours.
**Significant at the 1 percent level.


Source: D. R. Smith, "Inter-Energy Competition Between Gas and Electricity in the Industrial Market: A Case for Deregulation?" Business Studies, North Texas State University, XII, No. 1, pages 8-9.





-21-


competitors."'I0 The implication of these statements is that the competitive stimulus provided by separate operations forces increased efficiencies and lower costs.

Regulatory commissions have had the implied responsibility to support rates which will secure for firms a normal return on their investments; they have not had the encouragement to recognize technological developments that increase competition if the effect 11
is to lower the rate of return. If technological developments heighten potential competition, divestiture may increase efficiency. Regulation may stifle competition and therefore decrease efficiency. Indications of this economy are examined in Chapters IV and V by comparing the costs and the cost curves for alternative organizations. Costs of Divestiture

A diseconomy that will be incurred if society is to benefit from competition is the legal and financial costs of divestiture. If the gas operation of an existing combination firm is sold so that two single service utilities are created, existing debt and equity must be refunded. Present interest rates are significantly higher than the rates of the old bonds; the interest expense of the new (gas) utility will be greater than the interest expense of the gas operations of the old (combination) utility. Mr. Eugene Meyer, 12
Vice President, Kidder, Peabody and Company, stated:


10Ibid., page 6.

llJoe S. Bain, Industrial Organization (New York: John Wiley and Sons, Inc., 1968), page 642.

12Combination Utility Companies, page 231.






-22

Whatever the arguments about the benefits of
competition or of combination utilities so far as operating costs are concerned, it is apparent that the financial costs attendant
to breaking up the combination companies
would be devastating for the consumer under
current market conditions.

Mr. Meyer studies the proposed divestiture as it would be applied and concluded that it would cost $5,315,404,000 even assuming lower than market rates, sales prices at original costs, and no costs to 13
finance common plant items.

At least two proposals have been suggested as a means by

which this diseconomy can be minimized. First, if the state commission announced, in advance, that the new gas utility was not able to charge rates in excess of those justified by the original (low) interest, the sale price of the divested assets would reflect the previous (low) rate. Interest cost would not go up; book value of the assets would go down.14 This does not imply a loss in actual value, however, as these assets are presently earning the low rates.

Second, most of the utility bonds with low rates were

issued many years ago and will mature soon. As they are paid off with new money, the average interest rate of combination utilities, 15
as well as single service utilities, will rise.


13Ibid., page 233.

14Dr. John W. Wilson, in Combination Utility Companies, page 85.
15This point was discussed by Robert H. Willis, President, Connecticut Natural Gas Corporation, in Combination Utility Companies, page 383.






-23-


Summary

A framework has been established that compares the economies and the diseconomios of alternative organizations in the energy industry. Combination firms are assumed to have lower common costs due to joint utilization of assets and lower fuel costs due to higher load factors. Additionally, regulation is expected to require efficiency in the overall operation of all utilities, including combination gas/electric firms. These economies are predicted to lower both the average costs and the cost curves of combination companies relative to single service utilities.

On the other hand, it is predicted that joint utilization of resources encompasses only a small portion of total operating expenses and that single service firms have both the ability and the incentive to increase their load factors. It is further asserted that regulatory agencies do not recognize technological developments in the energy industry and thus do not force the realization of potential efficiencies. To the extent that these factors are significant, single service utilities are predicted to have lower costs and greater efficiencies than combination gas/ electric utilities.

Given these economies and diseconomies, it seems that

there does not exist one attainable organization that is predicted to be superior to all others. Measurements of the economies and of the diseconomies are undertaken in Chapters IV and V to determine the relative significance of each.






-24

Marginal Cost Pricing

Economic efficiency requires not only the minimization

of costs in producing the optimal output, but also the equilibrating of price and marginal cost. The minimization of costs represents efficient production; the equating of price and marginal cost represents efficiency in the allocation of resources. The Marginal Cost Pricing Principle

It is impossible, and in fact inappropriate, to develop the general theory of marginal cost pricing in this paper. This task has been vividly and lucidly fulfilled by many authors.16 The principle states that where prices reflect marginal (opportunity) costs, consumers will purchase quantities of the goods thg reflect an optimum allocation of resources, an optimum level of production and an optimum scale of plant. If some buyers are charged more than marginal (opportunity) costs for particular commodities, they will buy less than the optimum quantities. "Consumers who would willingly have had society allocate to its production the incremental resources required, willingly sacrificing


16See for example: K. J. Arrow, Social Choice and Individual Values (New York: Wiley & Sons, 1951); J. R. Hicks, "The Foundations of Welfare Economics," Economic Journal (1939); N. Kaldor, "Welfare Propositions in Economics and Interpersonal Comparisons of Utility," Economic Journal (1939); I. M. D. Little, "Review of T. Scitovsky, 'Welfare and Competition'," Econometrica (1952); M. W. Reder, Studies in the Theory of Welfare Economics (Oxford: Oxford University Press, 1947); or T. Scitovsky, "The State of Welfare Economics," American Economic Review (1951). These specific references are intended as a starting place only for reading and research. The literature literally abounds with references regarding the statement, development, criticism and refinement of the marginal cost principle.






-25

the alternative goods and services that those resources could have produced, will refrain from making those additional purchases because the price to them exaggerates the sacrifices." 17

Two anomalies to the principle must be recognized:

(1) Prices must reflect all the (marginal) costs of

production and the benefits of consumption for the principle to be valid. If a portion of the costs are borne by society or a portion of the benefits are received by society, the effect is 18
an underallocation of resources and a less than optimal output.

(2) The rule does not produce optimal results if it is

applied only partially. If portions of the economy are constrained in such a manner that price is not equated with marginal costs, the second-best theorem indicates that increases in welfare would not be assured by forcing other portions of the economy to equate 19
price with marginal cost.

E. J. Mishan, among others, developed the marginal cost principle into a practical policy proposal in spite of the problems

A. E. Kahn, The Economics of Regulation (Volume I, New York: Wiley & Sons, 1970), pages 66-67.

18See especially E. J. Mishan, "Reflections on Recent Developments in the Concept of External Effects," in Welfare Economics: Ten Introductory Essays (New York: Random House, 1969), page 180.

19The "Second-Best" theory was formulated by R. G. Lipsey and K. Lancaster in "The General Theory of Second Best," Review of Economic Studies (Volume XXIV, 1957). This article set off an intense controversy regarding the validity and importance of the marginal cost principle as a relevant and cogent policy proposal for use in an economy pervaded by imperfect competition, monopoly, externalities and government intervention.






-26-


20
of the second-best. Mishan proposes welfare will increase if some prices are equated with marginal costs even if other prices (in the constrained sectors) cannot be so equated. He suggests this policy will result in a greater increase in welfare the smaller the constrained sectors relative to the remaining ones, and the larger the initial discrepancies in the price/marginal cost ratios of the free sectors as compared with the constrained sectors. Since it is unlikely that at any instant the real world will ever attain a Pareto optimum, "(w)e can only hope to be moving in that direction most of the time and not to be too far away from an optimum for any prolonged period."21

The marginal cost pricing principle is a guide that,

while far from perfect, aims the economy in this desirable direction. Marginal Costs and the Energy Industry

Marginal cost is the change in costs when one additional unit of output is produced; alternatively, it is the cost that would not be incurred if the marginal unit is not produced. Since marginal cost is completely independent of fixed or sunk costs, the only cost relevant in deciding how much to produce in an existing plant is the variable cost of operating that plant.

The longer the time perspective of the costing process, the greater the proportion of costs that become variable. If a



20
E. J. Mishan, "Second Thoughts on Second Best," in Welfare Economics, page 141.

21Ibid, page 156.






-Z7

long enough time perspective is observed, all costs are variable and thus all costs are reflected in the marginal. The short run marginal cost reflects the social opportunity cost of providing the additional unit that buyers are, at any given time, trying to decide whether to buy or not to buy. The application of the marginal cost pricing principle requires that price be related 22
to short run marginal cost. Pricing at short run marginal cost does not necessarily mean price will be less than average cost. If the average total cost curve rises (or is constant) at any point along the output scale, it is because the marginal cost exceeds (or is at least equal to) the average total cost. If price is equated to marginal cost, which is greater than (equal to) average total cost, price must be remunerative to the firm.

Most resources are fixed and most costs are not represented in the short run marginal cost of electric and gas utilities. Significant differences are thus predicted between the levels of short run and the long run marginal costs up to the capacity of the system. At capacity, infinite changes in costs would not enable the utility, in the short run, to increase output; at capacity, short run marginal cost is predicted to become vertical. Figure I represents a graphical illustration of the predicted relationship between short run and long run marginal cost. SRMC is the short run marginal cost and it is vertical at capacity (output 0A). LRMC is the long run marginal cost indicating constant returns to scale over the relevant output range. In this paper, constant

22
A. E. Kahn, The Economics of Regulation, pages 70-73.

















SRMC


LRMC=LRAC


A

Figure #1 The Short Run/Long Run Marginal Cost Relationship


i





-29

returns is assumed as representative of the costs of the energy industry. 23

Pricing in the Energy Industry

Figure 2A indicates an energy system with capacity of 0QI. If price is equated with SRNC (at OP3), price would exceed LRAC (=LRMC) indicating economic profits. The profits indicate that a larger capacity is necessary.

In Figure 2B the system capacity is OQ2. If price is equated with SRMC (at OP), it would be less than LRAC and thus force economic losses. The losses indicate that a smaller capacity is necessary.

In Figure 2C, the system capacity of OQ2. If price is

equated with SRMC (at OP2), it is also equated with LRMC and LRAC. Resources are allocated in an optimal manner.

If the same capacity serves more than one class of customer with more than one time period considered, capacity costs should be charged only to those users that use the entire capacity. For example, Figure 3A presents a vertical summation of peak and off-peak demands.24 D1 represents the peak use and


23Constant returns does not imply that two firms could
supply the same output as one firm with the same level of average costs, for the duplication of resources that are required in the distribution of the energy would greatly increase the per-customer Cost,-Constant returns implies that one firm can increase its system capacity and have the same per-unit costs. For a discussion of this assumption please see A. E. Kahn, pages 152-165.

24For a discussion of the vertical summation technique see: P. 0. Steiner, "Peak Loads and Efficient Pricing," Quarterly Journal of Economics (November, 1957), pages 585-610; 0. E. Williamson, "Peak-Load Pricing and Optimal Capacity Under Indivisibility Constraint," American Economic Review (September, 1966), pages 810-827; and A. E. Kahn, The Economics of Regulation, pages 89-103.



























SRMC


LRMLC=LRAC






- Q




(B)





LRMC=LRAC


Figure #2


Pricing Under Capacity Constraints: The Case of a Single Peak Consumer


-30-


SRMC






-31-


SRMC SRMC'
'' (A)


DC



P3



P2 _ _ _LRMC=LRAC








O QI Q2 Q3 Q





Q
(B)
DC

SRMC SRMC'







P5
! _ __...._LRMC=LRAC P4



P2 D2 DI



0 Q2 Q3

Figure #3

Pricing Under Capacity Constraints:
The Case of Both Peak and Off-Peak Consumers





-32

D2 represents the off-peak use. Existing capacity is OQ2. Optimum pricing requires the peak user to pay price OP3 (=SRMC) which will ration the peak quantity to that level of capacity (OQ2). The offpeak user will pay price OP1 (=SRMC) and excess capacity (Q2 - QI) exists during the off-peak period. Only the peak user pays any part of the fixed (capacity) costs. Since OP3 exceeds LRMC, the system capacity is not optimal. In the long run, a capacity of OQ3 should be built. The peak price should then be OP2 (=SRMC=LRMC), the peak quantity OQ3, and the off-peak price OP1 (=SRMC), the offpeak quantity OQI. The marginal cost pricing principle thus forces the optimal scale of plant and allocation of resources.

In Figure 3B there is no excess capacity in the off-peak period. Existing capacity is OQ2. Peak price is OP5; off-peak price is OP Since OP3 + OP5 is greater than LRAC, system capacity is less than optimal. If the capacity is expanded to OQ3 (where DC is equated with LRMC), peak price should be OP4, off-peak price should be OP2 and both would use the entire capacity (OQ2). Each class contributes to capacity costs, the contribution from the peak user (P4 - SRMC) is greater than the contribution from the off-peak user (P2 - SRMC).

Summary

Policies are based on generalizations which do not always hold. Frequently, these generalizations must be quite broad. Since we cannot develop perfect policy rules, we must do the best we can and deal with exceptions in an ad hoc manner. Rules must be tested against the available alternatives and not rejected when they do not conform to the purely theoretical criterion. A






-33


set of simple, specific policy rules is essential and must be selected from the best of those alternatives that are available.

The margirial cost rule appears to be superior to its only apparent rival, the average cost rule. The marginal cost rule can be stated: Prices should be set at short run marginal costs when possible, and when not possible, marginal costs should be recognized as both a target toward which prices should be 25
directed and a floor above which prices should always exceed.

The optimal system capacity equilibrates the combined demands (Dc) with the long run marginal cost (LRMC). Each class of customer must pay a price at least equal to the short run marginal cost (SRMC), plus a contribution (in proportion to its usage at the peak period) to capacity costs. Under conditions of constant costs (LRMC=LRAC), the equilibrating of price and SRMC results in a profit or a loss only if the system capacity is other than optimal (resources are not allocated in an optimal manner).

The energy industry presently utilizes peak-responsibility 26
pricing to some degree. Electric and gas utilities employ a twopart tariff for sales to wholesale and to industrial customers in which an "energy" charge (SRMC) is added to a "demand" or "capacity" charge. The proper measure of the demand charge is the proportionate share of the peak demand, placed by each customer, on the system peak. The different rates charged to different classes of customers

25
For an excellent summary of the application of the marginal cost rule see M. J. Farrell, "In Defense of Public Utility Price Theory," in Public Enterprise, R. Turvey, ed. (Baltimore: Penguin Books, 1968), page 47.
26A. E. Kahn, The Economics of Regulation, pages 95-100.





-34


is not discriminatory if the difference in rates reflects a difference in cost. Discrimination consists of price differences not corresponding to cost differences. The off-peak user should not be charged any capacity cost if capacity would be there whether or not the off-peak user made demands on it.

Average prices are compared to both short run and to long run marginal costs in Chapters IV and V. In those comparisons, prices are shown as a percentage of marginal costs (P = X*" MC). The percentage mark-up figure for the industrial customers is subtracted from the percentage mark-up figure for the residential customers. The difference is a rough estimate of discrimination.













CHAPTER III

EVALUATION OF EARLIER WORK



The competitiveness of single service and combination utilities has been studied by several authors. In this chapter these studies are reviewed and evaluated. The policies suggested by these studies are in conflict. Indeed, it is this conflict that establishes the need for the work pursued in this thesis.


The Study by Bruce M. Owen

Bruce M. Owen was the first to analyze the economic
I
impact of the common ownership of the sources of energy. This research is a valuable contribution for it stimulated interest and further writings. Owen bases his study on theoretical reasoning that suggests that combination utilities should have greater economic power than single service companies. If regulatory commissions have successfully prevented combination firms from exploiting this market power, there should be no significant difference between prices and outputs due to the market organization. Thus, the effectiveness of regulation may be measured by observing the relationship of prices and outputs between the types of utilities.


lB. M. Owen, "Monopoly Pricing in Combined Gas and
Electric Utilities," The Antitrust Bulletin, Volume XV (Winter, 1970), pages 713-726.


-35-






-36

Owen classifies any utility that has both electric and

gas revenues as a combination firm and includes only utilities with revenues of $1 mill-ion or more in 1967. The data are derived from the Federal Power Commission's Statistics of Privately Owned Electric Utilities and Brown's Directory of North American Gas Companies. Owen defines price as average revenue per kilowatthour or m-therm sold to final customers. All figures are studied on a national basis. An econometric model employing simultaneous equations is utilized.

Owen finds that prices are significantly higher and

output significantly lower for combination gas/electric utilities than for single service electric utilities. He concludes, therefore, that the results establish the ineffectiveness of regulation. More vigorous enforcement of antitrust policy is proposed.

The results of the study can be questioned on several grounds. First, the lumping together of all utilities that have both gas and electric revenues with no regard to the relative significance of the components creates a group of firms that are not comparable. For example, Southern California Edison Company had electric revenues of over $720 million and gas revenues of less than $150,000 in 1970. Pacific Gas and Electric Company had electric revenues of over $704 million and gas revenues of over $474
3
million in 1970. These two companies are both classified as

2
Ibid., page 722.
3Federal Power Commission, Statistics of Privately
Owned Electric Companies in the United States, Washington: U. S. Government Printing Office (1970), page 103.





-37-


combination utilities and not distinguished in any manner even though one has gas revenues that comprise only .02 percent of its electric revenues and the other has gas revenues that comprise 67 percent of its electric revenues. Such inconsistencies within the groups of firms examined could lead to erroneous conclusions. A second criticism involves geographical differences and their effects upon energy consumption. Given that the marginal rates decline as use by a given customer increases, average revenue may not be an appropriate measure of price. Instead, the average revenue received by a particular utility from any given customer class will be determined jointly by the actual design of the rate structure as well as the quantity consumption per customer. If there are significant differences in quantity of consumption per customer among geographical areas of the country, and if certain types of utilities are found concentrated in specific geographical localities, the results of a study based on aggregated national data must be questioned.


The Study by Patrick C. Mann

Patrick C. Mann compares the economic performance of

single service electric utilities with the electric operations of combination gas/electric utilities. Only firms with revenues greater than $20 million in 1967 are considered. Mann uses data from the Federal Power Commission's Statistics of Privately Owned Electric Utilities. The method utilized is regression analysis.

4
P. C. Mann, "The Impact of Competition in the Supply of Electricity," Quarterly Review of Economics and Business, Volume 10, Number 4 (Winter, 1970), pages 37-49.






-38

Combination utilities are defined as firms with revenues of any magnitude from both gas and electric sales. Mann recognizes the problem this classification creates and includes a variable in his analysis (the ratio of gas to electric revenues) to take the problem into account. He correlates this gas/electric revenue ratio with the level of average electric revenue per residential kilowatt-hour sold. Other variables are the rate of return earned on net electric plant, distribution expense per kilowatt-hour, distribution expense per customer, administrative-general expense per kilowatt-hour, sales expense per customer, cost per kilowatthour of steam generated electricity, and cost per kilowatt-hour consumption per residential customer. Mann finds no significant relationship between the ratio of gas to electric revenues and the price of residential electricity.

Mann concludes that joint operations of gas and electric operations d o not affect residential markets. "(R)esidential price differentials between combination and straight electric firms are caused primarily by factors independent of the dual service nature of the combination utility."5 Additionally, Mann concludes that there is evidence that the higher commercial and industrial prices associated with combination firms are "partly a function of competition between gas and electricity suppliers."6

Mann does make a significant improvement to Owen's

statistical methodology by including the ratio of gas to electric

5Ibi., page 48.

6
Ibid, page 49.






-39

revenues. However, his conclusions are questionable for at least two reasons. First, Mann attempts to test the differences in efficiencies, as measured by costs, between the types of utilities. In a regulated monopoly with zero economic profits, total revenue equals expenses (production, transmission, distribution and customer accounts) plus taxes and the rate of return. Mann analyzes variations in average revenues after accounting for all cost differences except differences in hydroelectric production expenses, transmission expenses and taxes. These factors are largely determined by the geographical location of the individual firm and not by the ownership of the utility. A firm will utilize hydroelectric facilities only if there exists a sufficient movement and fall of water in the service area. Transmission expenses are largely a function of the density of the population. Taxes are related to the areas the plant and equipment are located. Mann has included most of the potential sources of cost variations in his dependent variables. The remaining sources of cost variation are a function of geographical location. It is not surprising that he finds few differences in the level of costs.

A second deficiency in Mann's study results from the

lumping together of commercial and industrial customers. According to Federal Power Commission statistics for all Class A and B investor-owned utilities, the average revenue received from all commercial customers was 2.0812 cents per kilowatt-hour in 1970; the'average revenue received from all industrial customers






-40

was 1.0175 cents per kilowatt-hour.7 The much lower rate prevailing for industrial customers reflects the fact that such customers tend to take more power at higher load factors and higher voltages than commercial customers. In addition, demand elasticities are thought to be much lower in the case of commercial customers than in the case of industrial customers. When commercial and industrial customers are grouped, the average revenue of the combined commercial and industrial sales is affected by the number (and the usage) of commercial customers and the number (and the usage) of industrial customers. A utility with a high number of commercial customers (relative to the number of industrial customers) is expected to have a higher average revenue than a company with a low number of commercial customers (relative to the number of industrial customers). The ratio of commercial to industrial customers and the level of use of each is highly correlated to geographical factors. This geographical distribution is not investigated, by Mann.


The Study by William H. Collins

Further work has been pursued by William H. Collins.8

He sought to identify and empirically test the social desirability of combination gas/electric utilities. The analysis is based on 1967 data from the Federal Power Commission's Statistics of Privately

7Federal Power Commission, Statistics of Privately Owned
Electric Utilities in the United States (Washington: U. S. Government Printing Office, 1970), Table 4, page XVII.
8W. H. Collins, Jr., "Combination Gas-Electric Utilities," Ph.D. dissertation, Southern Illinois University (197). See also W. H. Collins' testimony in Combination Utility Companies (Washington" U. S. Government Printing Office, (69-612), 1971), pages 441-448.







"41"

Owned Electric Utilities in the United States. The sample includes 52 combination utilities, 89 straight electric utilities and 62 straight gas utilities. Collins employs "z" tests, Student "t" tests, Mann-Whitney U tests, and Wilcoxon Matched Pairs Signed Ranks tests in his computations of the data.

Collins finds that there is no significant difference between single service gas utilities and the gas operations of combination utilities. On the other hand, the results of the electric computations indicate that the performance of the electric operations of combination utilities is significantly below that of single service electric utilities. This conclusion is based primarily on the fact that when the performance of the single service electric firms is compared with the performance of the combination firms, the combination group has significantly higher levels of average revenue per kilowatt-hour for each customer class and significantly lower levels of kilowatt-hour consumption per residential customer. In addition, Collins finds that average salaries and wages, average distribution expenses, average operating revenue (less operation and maintenance expense) and average taxes are significantly higher on a per kilowatt-hour basis for combination utilities than for single service utilities. In contrast, sales and advertising expenses are found to be higher for single service utilities.

Collins' study is a valuable contribution to the development of the literature concerning the competitiveness of public utilities in that he establishes a statistical methodology that






-42

identifies differences between variables indicating differences in consumption and efficiencies. Collins found that combination companies charge higher prices and have lower quantity sales than single service electric firms. Economic theory predicts that the greater the monopoly power, the greater the ability to raise prices and (thus) restrict sales. The conclusions formulated by Collins appear to support the hypothesis that a more efficient allocation and utilization of resources results when competition between the sources of electricity and gas is increased. As a policy matter, the Collins study thus supports the idea that antitrust, if applied to combination gas/electric firms, will benefit the consumer.

The findings of the Collins study can be questioned principally because of too strong an assumption concerning the homogeneity of the groups of utilities. Collins states that he statistically tested the groups of data and found no significant difference with regard to such factors as utility size, relative importance of different types of customers, type of generation, percent purchased power, degree of urbanization, population density, geographic location and distance from fuel sources. However, his methods are questionable. Although he states the groups are not significantly different, he does not describe how he tested the homogenety in geographic location. If the combination utilities are concentrated in geographical localities that are characterized by higher fuel costs, wage rates, state and local tax burdens, generating unit costs, and other such cost items, a priori reasoning suggests that such firms will charge higher prices





-43-


(and thus sell smaller outputs) regardless of the market organization. Conversely, if single service electric utilities are concentrated in geographical localities where the climate is warm, the high use of air conditioning will increase the monthly consumption per customer. This increase in consumption reduces average revenue since marginal rates are inversely related to consumption. The lower average revenues (average prices) may thus be due to the higher consumption; the higher consumption may not be due to lower average revenues.

Collins uses a logical methodological procedure. However, he leaves his conclusions open to criticism by not analyzing the geographical location of the utilities and by neglecting the impact that a highly uneven distribution of the types of firms might have on the statistical findings.


The Study by the N.E.R.A.

The National Economic Research Associates, Inc. (NERA),

analyzed various financial and operating data pertaining to electric and gas distribution companies in a report prepared for the Long Island Lighting Company.9 Data were gathered from a variety of sources including: Moody's Public Utilities Manual, Standard and Poor's Compustat tape, the Federal Power Commission's Statistics of Privately Owned Electric Utilities, and Brown's Directory of North American Gas Companies. Single service utilities are defined

-99
National Economic Research Associates, Inc., "Combination Companies: A Comparative Study," as printed in Combination Utility Companies, pages 275-363.






-44

as any company with more than 90 percent of its revenue derived from the sale of one source of energy. The methodology utilized by the NERA involves the computation of the arithmetic means of the variables studied. These means are compared as indicators of the relative similarities and differences of the utilities. For example, the capital structure of the firms was studied. As an example of the methodology, it was stated that equity comprised 37.1 percent of the straight electric utilities' total capital, 36.7 percent of the combination utilities' total capital, and 42.2 percent of the straight gas utilities' total capital. From these measurements the conclusion was reached that ". straight gas companies have relatively more equity in their capital structure . . . than either straight electric or combination companies."

The'linancial" data section examines the growth rate in earnings per share, the rate of return on invested capital and the rate of return on common equity (as defined by Moody's Public Utilities Manual). The study concludes that differences in the financial measures (stated above) between the groups of utilities are very small. Specifically the study states . while there is a close similarity in the averages for the straight electric and combination companies . . . the straight electric would appear to have a slight overall edge over the combination companies. Similarly, except for return on invested capital, the combination companies have out-performed the straight gas

10
Ibid., page 282





-45

group."11 The observation is made that straight gas companies have relatively more equity in their capital structure (as stated above), than either the straight electric or the combination companies.

In the "expense and plant" data section the study investigates (1) customer accounts, (2) administrative, general, and sales expenses, (3) operating and maintenance expenses and (4) gross plant. Combination companies are found to have lower average customer expense than the straight firms although when these expenses are related to quantity sales, the relationship was reversed. Similar findings are observed with respect to the administrative, general and sales expense figures. The single service utilities have higher expense figures per customer although no difference is observed on a per-unit comparison. Operation and maintenance expense per customer is higher for both single service electric and single service gas utilities than for combination utilities. However, operation and maintenance expense per unit output is lower for single service electric utilities than combination firms, but higher for single service gas utilities than combination firms.

With respect to the "consumption and customer" variables, the study finds that single service electrics have higher growth rates and percentage increases in customers than the electric side of combination companies. The gas side of combination companies has higher growth rates and percentage increase in customers than the single service gas companies.

The "miscellaneous" data investigated include employees per dollar of revenue created, load factors, BTU heat rate, average

11Ibid., page 298.






-46

revenue from residential electric sales and average revenue from residential gas sales. The study found that all three groups have similar employees per dollar revenue and that load factors are not significantly different between the firms. The average revenue from residential electric sales is lower for the single service electric than for the combination companies, while the average revenue from residential gas sales is lower for the combination utilities than for the straight companies.

As an overall conclusion regarding the study, Jules Joskow, Vice President, National Economic Research Associates,
12
Inc., stated:

The results of our study show that while for certain variables one group of companies may have, on average, outperformed one or both of the other groups, the data, when viewed on an
overall basis, do not suggest that any one group
has a statistical advantage over either of the other two. Thus, we further conclude that the
data do not indicate that a conclusion can be
reached either in favor of, or against, combination companies based on their performance as
herin measured.

The findings of fact of the above study d o not

differ significantly from the studies of Owen, Mann and Collins. The policy implications do differ from the earlier studies. Although similar data are studied, and similar findings discovered, the authors of the NERA study state that a conclusion could not be reached regarding overall performance. A question of data interpretation is thus raised.


12Ibid., page 275.






-47

The NERA study analyzes 47 straight electric, 40 combination and 13 straight gas utilities. The findings are based on a simple comparison of the arithmetic means of each group of firms and no generally accepted statistical technique is utilized to measure the significance of the findings. The analysis of only a portion of the firms with no explanation as to the basis for inclusion or exclusion of firms and the inadequate statistical method raises questions concerning the validity of the analysis. Differences between arithmetic means may be caused by actual differences in the variables compared; alternatively, differences between arithmetic means may be caused by sampling. This study does not have a statistically sound base and is theoretically weak. Thus, the findings are, in and of themselves, not analytically important. The data and the statistical findings of the NERA analysis are similar to those of previous studies. The fact that the conclusions are different is important.

Summary of Owen, Mann, Collins and the NERA

Differing samples of utilities, various definitions of

the degree of competition, and a variety of cost-influencing variables have been used in the studies reviewed above. It is thus not surprising that somewhat different conclusions have been drawn by the various authors. However, despite the differences and the conflicts, several points stand out. The combined control of gas and electric operations is associated with lower levels of kilowatthour consumption and higher average revenues per kilowatt-hour. The studies also indicate that the costs of electrical operation are






-48

higher for combination gas/electric utilities than for single service utilities. No differences between combination and single service gas utility performance are apparent.

The policy recommendations each author proposes are likewise similar although each has differences in his application. Since each of the studies found higher prices and lower outputs associated with combination gas/electric utilities, each concluded the separation of ownership of the sources of energy would increase the overall welfare of the consumers. The one exception is the NERA study which states that no conclusion can be reached.

In conducting these studies, the authors have consistently failed to consider the geographical distribution of the utilities and the resulting variation in the demographic factors that would thus appear. Dissimilar demographic characteristics are expected to influence the level of each utility's rates and thus affect the quantities consumed by customers in the service area. Differences between combination gas/electric utilities and single service electric utilities that are isolated by the authors reviewed above may thus be caused by the geographical (and therefore demographical) differences between the groups of firms compared.


The Study by Joe D. Pace

Joe D. Pace studied the differences between types of utilities that have been isolated by the earlier studies.13

13j. D. Pace, Senior Economist, National Economic Research Associates, Inc. The study referred to in this thesis is found in Combination Utility Companies, pages 518-553.






-49

His analysis is based on the fact that less electricity is consumed. in areas served by combination companies. He states that previous authors failed to probe deeply enough into possible relationships between this lower consumption and various demographic factors which may differentially characterize the service areas of single service electric utilities as opposed to those of combination companies. He utilizes multiple regression analysis to conduct an exhaustive statistical study of national data. Pace limits his study to urban areas in the United States that contained 25,000 or more housing units in 1970 and were served by investorowned utilities. Because of a "lack of gas industry data comparable 14
in detail and uniformity to those of the electric industry" the study includes only the electric component. Data were obtained from the Annual Reports of Class A and B electric utilities to the Federal Power Commission. The study includes 194 urban areas of which straight electric utilities serve 108 (or 55.7 percent) and combination companies serve 86 (or 44.3 percent).

Pace establishes the fact that various types of utilities

are not evenly distributed throughout the United States. Combination utilities are found concentrated in urban areas which are larger, more densely populated and characterized by higher fuel costs and higher state and local taxes. Larger numbers of apartments exist in combination areas. On theother hand, because average temperatures are higher in the areas served by single service electric utilities, the potential demand for air conditioning is greater in areas served


14Ibid., page 538.






.50.

by single service electric utilities. Pace concludes that it is not surprising to find that average residential electric revenues are lower in areas-served by combination companies since consumption is greater and step-rates are charged. "Whether the geographic and demographic conditions explain all or only a small part of the observed usage and rate differences. . ." is the question that
15
Pace studies.

The variables that Pace studies are divided into six

areas. Measures of the "housing" variables are (a) the number of apartment units as a percent of total housing units and (b) the number of seasonal and migratory units per 1,000 housing units. "Urbanization" is measured by the number of housing units in each community, population density and the ratio of residential electric customers in each urban area to the number of housing units in each community. To study "income" variables, Pace segments the percent of all housing units without all sound plumbing facilities, the percent of total housing units occupied by more than 1.51 persons per room, apartment units renting for less than $100.00 per month as a percent of total housing units, per capita income, effective buying income estimates (per household) and the percent of households with cash incomes of less than $3,000 per year. To investigate "weather" conditions Pace introduces as independent variables both the number of cooling degree days for June, July and August, and an income-adjusted cooling degree day measure.


15
Ibid., page 519.






-51

"Locational" factors include geographic indicators based on the section of the country the utility is located. The final subdivision relates to-"cost-influencing" factors. Included as variables are state and local taxes per dollar revenue, fuel costs, hydroelectric generation as a proportion of total generation, purchased power as a proportion of total sales, fuel used in generation, system size, density and statewide net electric operating revenue as a percent of net electric plant.

Pace found no significant difference between the average revenues of combination utilities and the average revenues of single service firms when the demographic variables are included in the analysis. "In plain language, this means that while demographic, geographic and underlying cost conditions very strongly affect the average residential electric revenue required, under current circumatances, it matters little, on average, whether a combination company or a single service electric utility renders service in a given community."16 Pace thus concludes that there is ". . .clear and substantial support to the finding that combination companies have no adverse effect, as a general matter, on residential electric rates .,17

The conclusions and the policy implications of the Pace study thus contradict those of previous authors and represent a significant contribution to the literature. As with the other studies, however, questions can be raised.

16
Ibid., page 533.
17Ibid., page 537.






-52

First, Pace examines only the electric operations of utilities. Further, Pace limits his study by examining only operations in large, urban areas. This restriction in the scope of the study may obscure significant differences that do exist. Additionally, the large number of items that Pace includes in his regressions introduce questions as to the interdependence between variables. Do the variables Pace examines explain the variations or are there more general, or basic, explanations?


Summary of the Earlier Work

Previous studies have resulted in conflicting policy recommendations. Owen, Mann and Collins found that combination utilities charge higher prices, sell smaller outputs and operate less efficiently than single service utilities. Their conclusions are based on national studies that largely ignore demographic differences between geographic areas. Pace concludes that there is no significant difference in the performance of utilities based on the type of ownership. High prices are determined by demographic characteristics which restrict consumption. It is this conflict that establishes the need for the study contained in Chapters IV and V.













CHAPTER IV

AGGRECATIVE ANALYSIS



The objective of this thesis is to test the hypothesis that separate ownership of gas and electric supplies will promote competition and that this separate ownership will result in improved efficiency and lower rates. Data are gathered from reports submitted directly by each utility. These data are more detailed and comprehensive than those available in existing publications. The data are subjected to statistical tests that are widely known and commonly accepted as a basis for evaluating differences in variables. The data are analyzed on a national, or aggregative, basis in this chapter.


The Data

Earlier studies obtained data from Federal Power Commission publications Statistics of Privately Owned Electric Utilities and
1
Statistics of Privately Owned Gas Utilities. These publications provide basic information for each combination and single service utility operating in the United States.

The data this thesis analyzes are gathered from publications entitled "Uniform Statistical Report" which are prepared by each utility and are filed with the Federal Power Commission. All Class A and B single service electric, single service gas, and combination


iWashington: U. S. Government Printing Office.


-53-






-54

gas/electric utilities were sent letters requesting a copy of this
2
report. The results of the survey of the utilities are summarized in Table 4.1.

The differences between the number of requests and the number in the sample reflect several factors. First, some of the firms are generation companies only and thus have no retail customers. In order to isolate differences associated with competition between utilities, retail sales must be examined; these generation utilities are therefore eliminated from the study. Second, many of the companies are part of holding companies and filed consolidated statements. As far as possible, each company is treated as an individual and independent entity, but where it appears that a holding company exerts policy influence at the subsidiary level, they are grouped together and treated as one company. Thus, individual utilities in the "Southern" holding system are studied independently while the "New England Electric System" and the "New England Gas and Electric Association" are studied as single firms.

For our purposes, a "straight" company is defined as one that derives not less than 95 percent of its revenue from the sale of one type of energy. The Federal Power Commission defines a combination utility as a firm with any revenue from the sale of both gas and electricity. As a third adjustment to the statistical reports, each firm's revenues are examined and combination firms


2The requests are based on a list of all the utilities in the Federal Power Commission publications mentioned above.









Table 4.1


The Results of the Survey of the Utilities


TOTAL NUMBER OF REQUESTS


TOTAL NUMBER IN FINAL SAMPLE


% SAMPLE REPRESENTS
OF TOTAL


Straight Electric Combination Gas/Electric Straight Gas


56%


66%


49%






-56

are reclassified as single service firms if one type of energy constitutes less than 5 percent of the total dollar sales.

A final factor involving the difference between those firms from which information was requested and those firms in the final sample involves the comprehensiveness of the data received. All firms must fill out the report and make it available to the public when it is requested; however, there are instances where portions of the data are not complete. Only those firms that supplied complete information are included in the final analysis.

Because of the adjustments outlined above, the utilization of information from the individual firm's reports in place of information from the Federal Power Commission represents a trade-off: the data are more detailed but they are not as complete. In the opinion of this writer, the additional detail is worth the loss in completeness, for this detail allows comparisons that are impossible with published data.


Methodology

A major intent of this thesis is to reconsile the contradictory conclusions of other studies; the methodology thus closely parallels that established in these studies. A cross-sectional approach (which freezes time and searches the data for indications of differences) is utilized as it is in all studies reviewed in Chapter III.

Detailed information is examined to ascertain the parameters that differ significantly due to the effects of competition. The study involves two similar but separate aspects of production: that of electricity and that of gas. Within each aspect of production,






-57

there are two types of utilities that are to be investigated: the electric figures of combination firms must be compared with the corresponding figures of straight electric companies and the gas figures of combination firms must be compared with the corresponding figures of straight gas companies. Due to the lack of a common measuring device, few relevant comparisions can be made between electric and gas figures. The test thus involves the question: Do statistics observed in two samples differ? This test must be posed independently with respect both to electric and to gas figures.

To accomplish the statistical test, means and standard

deviations are computed. The question then asked is: Can differences between two sample means be attributed to chance errors or to differences in the populations from which the samples were drawn? A statistical approach commonly used to compare means involves the computation of a "z" statistic (for large samples) or a
3
"Student's t" statistic (for small samples). The null hypothesis in both cases asserts that the means of the population parameters being tested are equal. This assumes that if a large number of samples were taken from one population (or two identical populations), the differences between the sample means would give a normal distribution centered around zero. If samples were taken from different populations, the differences between the sample means would not average zero but would, instead, equal the differences in the means of the population parameters. For example, the

3
See for example M. Hamburg, Statistical Analysis for Decision Making (New York: Harcourt, Brace and World, 1972), pages 330-348; Ya-lun Chou, Statistical Analysis (New York: Holt, Rinehart and Winston, Inc., 1969), pages 387-395; Smith and Williams, Statistical Analysis for Business: A Conceptual Approach (Belmont, California: Wadsworth Publishing Company, Inc., 1971), pages 393-429.






-58

assumption is made that there is no difference between the total population served by the straight electric companies and the electric operations of the combination utilities. The average population size of the sample of straight electric companies is then compared to the average population size of the sample of the combination utilities. The test is whether the divergence indicates differences in the populations served by the firms. Specifically, a statistic showing the difference between the sample means (xI - x2) is established and if this statistic differs significantly from zero, it is concluded that the population parameters are indeed different.

To arrive at the significance factor, the "standard error
4
of the difference between two means" is estimated as:


s _ - = sl/nI + /n2
(x -x2)

The test is two-tailed because the hypothesis of equal populations would be rejected if (xl - x2) differed significantly (either above or below) from zero.

For large samples the test can be viewed as a comparison of the specific difference of sample means to a frequency distribution obtained by drawing repeated pairs of samples and observing the differences in these samples. If the distribution is normal, 69 percent of all observations should lie within plus-or-minus one standard deviation, 95 percent within plus-or-minus two standard deviations and 99 percent within plus-or-minus three standard deviations.


4M. Hamburg, Statistical Analysis for Decision Making, page 332.






-59

It is then assumed that if xI - x2 is greater than three standard errors, the population means differ at the 99 percent confidence level.

The computations of these comparisons is simplified by
5
computing the "z" ratio which is defined as: x I - x 2
Z =

(xI - x2)


The "z" statistic measures the number of standard errors between the means. For example, a "z" of 3 indicates that xI - x2 (the difference between the means) is three times the value of the standard erros (s_ _ ). By selecting a level of confidence, a
x1 - x2
a value of"z'is determined which, if exceeded, indicates that the means differed significantly from each other.

For small samples, the"Student's t"is utilized. The

theory of this test is similar to that of the"I'test. The estimate
6
of the standard error is:

V (nl - l).(sP) + (n2 - l)'(si)


(xl - x2) n1 + n2 2

and the"ttstatistic is defined as:7

I " x2
S
( - T2)

5
Ibid., page 333.

6Ibid., page 347.

71bid., page 347.






-60

The critical values of "t" are based on the size of the

sample. Tables are included in the statistical appendices presenting the values of both'"z" and of "t" for various levels of confidence.


Table Notations

In order to simplify the presentation of the many computations upon which the variables are to be compared, confidence levels of 95 percent and 99 percent are selected as critical. The data are analyzed and the corresponding "z" or "t" statistic is computed and presented in the statistical appendices. Those variables that are significantly different are presented in the body of Chapters IV and V. For example, the average "Total Gas Population Served" by the straight gas utilities is compared to the average "Total Gas Population Served" by the gas operations of the combination gas/electric utilities. The values are 1,901,721 for the straight gas companies and 984,968 for the combination gas/ electric utilities. The "z" statistic computed for this variable is -2.26 which indicates that the straight gas companies serve a significantly larger population (at the 95 percent confidence level) than the gas operations of the combination firms. The negative sign indicates that the straight utility figure is larger than the combination utility figure. The average population figures and the "z" statistics are found in Appendix C of Chapter IV. In Table 4.2, a "G" under the gas column of "Total Population Served" indicates that the straight gas company is larger (at the 95 percent confidence level) than the gas operations of the combination utilities. If the "z" is less than -2.8, the degree of confidence would be 99 per-






-61

cent and the character in Table 4.2 would be "G*". A dash ("-") indicates that there is no significant difference between the parameters examined.


Aregate Characteristics of the Firms

Variables are selected that measure aggregate characteristics of each utility. These figures are presented in Tables 4.2 through 4.4.

Table 4.2 presents demographic and sales characteristics. No significant differences are observed in the electric computations; thus, it is concluded that the electric operations of the combination utilities and the straight electric utilities are of similar size. However, differences are observed in the gas computations. Single service gas companies serve a larger population (although not with a higher density), sell larger quantities to industrial customers and receive larger revenues from industrial customers. Single service gas companies are larger than the gas operations of combination gas/electric utilities.

Table 4.3 compares cost figures, plant size figures, interest charges, taxes, and efficiency and fuel cost measures (for electric firms). Few differences are observed in the electric computations. The differences that are observed indicate that straight electric companies spend more on sales efforts and have greater dollar investments in general plant than combination gas/ electric companies.

Significant differences are evident with regard to the

gas expense figures. The indication again is that the straight gas







Table 4.2


Aggregate Characteristics of the Utilities: Demographic and Sales Variables


Electric Gas Computations Computations

1. Total Population Served G

2. Square Miles Served

3. Population/Square Mile

4. Number of Customers:
Residential Customers
Commercial Customers Industrial Customers

5. Revenues Received:
Residential Revenues
Commercial Revenues
Industrial Revenues - G

6. Quantity Sales:
Residential Sales Commercial Sales
Industrial Sales - G


Notation: See text, pages 60-61.






Table 4.3


Aggregate Characteristics of the Utilities: Expense, Plant and Tax Variables

Electric Gas
Computations Computations

1. Expenses:
Operating Expense - G
Maintenance Expense
Depreciation Expense - G

Production Expense - G
Distribution Expense Customer Expense
Sales Expense E* G* C'
Administration and General Expense - G

Total Operating Expense G

2. Plant:
Production Plant - G
Distribution Plant
General Plant E* G*

3. Total Interest C*

4. Taxes:
State and Local Tax
Income Tax - G

5. Efficiency (Average Cost/BTU)

6. Fuel Cost (BTU/KWH)

Notation: See text, pages 60-61.


I






-64-


companies are larger than the gas operations of the combination gas/electric firms. The higher production plant for single service gas companies reflects the fact that combination firms buy from gas transmission companies whereas single service gas companies tend to be vertically integrated. The higher interest charges for the combination utilities is partially due to the nature of the data: the data do not differentiate between electric or gas operations of combination firms. The aggregate "Interest Charged" thus reflects total interest (for both electric and gas operations) for the combination utility and reflects only the interest charges of the single operations of the straight utility. The significant point is that there are differences in the gas computations and there are not significant differences in the electric computations. This fact indicates that single service electric companies have interest charges as large as the combined operations of gas/ electric companies and, thus, are of somewhat equal size. Single service gas companies are smaller than combination companies and, thus, smaller than single service electric companies.

State and local taxes do not differ in the gas computations, but single service gas companies pay higher income taxes relative to the payments of the gas operations of combination companies. These figures will be discussed in detail at a point later in the thesis; however, it should be noted that while straight gas companies have larger expenses than the gas operations of combination firms, they pay more in income taxes but do not pay more in state and local taxes.






-65

The efficiency of a utility's production depends to a

large extent upon the state of technology; the organization of the industry is not expected to affect this efficiency. It is thus not surprising that differences in efficiency and fuel costs in the electric computations are not observed. Similar comparisons could not be made for gas operations as the reports do not contain figures of a corresponding nature for gas operations.

Table 4.4 indicates that single service gas companies employ more workers and pay larger total wages than do the gas operations of combination firms. These facts are expected since the single service gas companies are larger than the gas operations of the combination firms. The fact that the data are not divided between fringe benefits for gas employees and fringe benefits for electric employees creates the (false) appearance that combination utilities pay larger dollar fringe benefits. This computation is discussed later in the study.

The financial figures do not compare equal items since both the "Source of Funds" and the "Application of Funds" for combination firms represent the combined gas/electric operations. As with the interest charges, the interesting point to notice is that there are no differences between the combined gas and electric financial figures and the straight electric figures, yet there are significant (at the 99 percent level) differences in the gas computations. The single service electric companies are about the same size as the electric operations of the combination companies. The gas operations of the combination firms are small in relation to the whole company.









Table 4.4


Aggregate Characteristics of the Utilities: Employment and Financial Variables


Electric Gas
Computations Computations

1. Employment:
Number of Employees G Total Wages G
Pension Benefits C '

2. Financial:
Source of Funds:
a. From Outside - C* b. From Inside - C*

Application of Funds:
a. Gross Additions to Plant - C* b. Dividends on Preferred Stock - C* c. Dividends on Common Stock - C* d. Total Application of Funds - C*


Notation: See text, pages 60-61.






-67

In t his section, certain aggregate characteristics of the firms have been analyzed to ensure that the data are homogeneous and that the observations made from the data are correctly interpreted. Few differences are observed between single service electric companies and the electric operations of the combination gas/electric utilities. The gas operations of the combination utilities are small relative to the operations of the single service gas companies. Relevant comparisons may therefore be expected with respect to electric computations, but caution must be maintained when attempting comparisons with gas figures. Differences that at first appear to be caused by competition, may be caused by size differences.


Findings of the Aggregative Analysis

In the national study, certain variables are studied as measures of "Economies and Diseconomies" and of "Marginal Cost Pricing." Both are discussed in turn. Economies and Diseconomies

Competitively organized industries, relative to monopolisticly organized industries, are expected to benefit the consumer by forcing lower prices and more optimal outputs. Table 4.5 illustrates comparisons of average revenues, average consumption and the average
8
yearly dollar bill per customer. The electric operations of combination companies charge higher prices and have smaller average consumption than single service electric companies. These price and

8
In Table 4.5, average revenue is defined as total dollar revenue divided by quantity sales; average consumption is defined as quantity sales divided by the total number of customers; and the average yearly bill is defined as the total dollar revenue divided by the number of customers.








Table 4.5


Average Consumption and Price


Electric Gas
Computations Computations

1. $ Revenues / Quantity Sales: (Average Price)
Residential C* Commercial C* Industrial C*

2. Quantity Sales / # of Customers:
Residential E* 00
Commercial E*
Industrial

3. $ Revenues / # of Customers:
Residential Commercial Industrial


Notation: See text, pages 60-61.


I






-69

consumption differences are not significant in the gas computations. This observation suggests that competition is a significant factor at least with regard to the electric aspects of the industry. Since the "competitive" indicators are not observed in the gas computations, it is concluded that there is a lesser degree of competition in the supply of gas. These findings are anticipated due to the size differential observed in the previous section. In the electric computations, the single service and the combination companies are more nearly equal in size and the effects of competition are thus more obvious.

Competition is expected to sharpen the decision-making

process and therefore expected to lower the average costs. Table 4.6 illustrates average cost computations for the utilities. Average total costs, average depreciation costs, average distribution costs, and average customer costs are significantly higher for the electric operations of combination utilities than for the straight electric companies. This observation suggests that competition forces single service firms to be more efficient than (monopoly) combination gas/ electric firms. Similar findings are not observed in the gas computations. These findings are not expected since competition is weaker in the supply of gas than in the supply of electricity.

Table 4.7 presents figures that compare average wages.

No significant differences are observed in the electric computations. Single service electric utilities are about equal in size to the electric operations of the combination utilities and each pays comparable wages. Pension benefits are slightly larger as a percentage of the total wages for the straight electric companies,







Table 4.6


Average Costs


Electric Gas
Computations Computations

1. Total Operating Revenue (Total Cost) / Quantity Sales C* 2. Total Operating Expense / Quantity Sales C*

3. Operating Expense / Quantity Sales
4. Maintenance Expense / Quantity Sales
5. Depreciation Expense / Quantity Sales C*
-4
6. Production Expense / Quantity Sales 0 .
7. Distribution Expense / Quantity Sales C 8. Customer Expense / Quantity Sales C 9. Sales Expense E* 10. Administration and General Expense / Quantity Sales


Notation: See text, pages 60-61.








Table 4.7


Average Wages


Electric Gas
Computations Computations

1. Total Wages / # of Employees C

2. Total Wages / Total Operating Revenue
(Wages as a % of Total Operating Revenue)

3. Pension Benefitsf # of Employees C* 4. Pension Benefits / Total Wages E C*
(Pension Benefits as a % of Total Wages)

5. Total Operating Revenue / # of Employees


Notation: See text, pages 60-61.






-72

although on a per-worker basis, there is no significant difference between the types of utilities.

In the ga's computations, combination companies pay higher wages per employee and have higher pension benefits. The gas operation of combination utilities is smaller than the electric operations of these firms; this size difference may cause the wage differential.

Table 4.7 further indicates that the average productivity is not significantly different in either the electric or the gas
9
computations. This productivity constitutes a rough measure of efficiency; higher average productivity causes greater outputs and increased revenues. This study indicates that competition does not increase the productivity of the employees.

Combination gas/electric utilities use their office

buildings and other items of general plant in both gas and electric operations. Table 4.8 presents computations of the use and the importance of plant items. These figures indicate that combination companies have greater revenues per dollar general plant, and incur higher general expenses per dollar general plant. Single service utilities spend more as a percentage of total plant on general plant. These facts support the idea that combination companies are able to use their general plant in a joint manner. The relative importance of this economy is not great however, as the differences are not significant with respect to total plant computations.

9
The average productivity is measured as the revenue generated per employee, or the total revenue divided by the number of workers employed. This productivity figure is presented as a rough estimate of the value of the marginal product.






Table 4.8


Physical Plant

Electric Gas Computations Computations
1. Total Operating Revenue / ( ) Plant:
Total Operating Revenue / Production Plant
Total Operating Revenue / Distribution Plant
Total Operating Revenue / General Plant C* C*
Total Operating Revenue / Total Plant
2. Quantity Sales / ( ) Plant:
Quantity Sales / Production Plant
Quantity Sales / Distribution Plant E*
Quantity Sales / General Plant C* c*
Quantity Sales / Total Plant E

3. ) Plant / Total Plant:
Production Plant / Total Plant
Distribution Plant / Total Plant C* General Plant / Total Plant E* G*

4. Relationship of Expenses to Plant:
Distribution Expense / Distribution Plant
Administration and General Expense / General Plant C* C*

Notation: See text, pages 60-61.






-74

Single service electric utilities have higher sales per

customer and therefore greater sales per dollar of (fixed) distribu10
tion plant. The greater consumption per customer indicates that the distribution plant is used more efficiently by the single service electric utilities than by the electric operations of the combination firms. Table 4.8 indicates that the straight electric utilities do in fact use their distribution plant more efficiently since the quantity sales per dollar distribution plant is greater for the straight electric companies than for the electric operations of the combination firms.

In sum, combination utilities use general plant items more optimally than the single service utilities, yet straight utilities use their distribution plant more efficiently than do the combination firms. The net advantage of this trade-off is small but weighted in favor of the single service firms since straight utilities have greater quantity sales per dollar total plant.

Table 4.9 compares the aggregate utilization of funds

between the two types of utilities. The data from the combination firm's "Uniform Statistical Reports"are not differentiated between electric or gas operations. This table thus compares the combined gas/electric operations of combination utilities with the single operations of straight companies. The significance of this comparison becomes clear when the gas data are investigated. The combined gas/ electric utilization of funds is significantly larger than the single service gas utilization of funds although straight gas firms are


10See Tables 4.5 and 4.8.









Table 4.9


Financial


Electric Gas
Computations Computations

1. Use of Funds as a % of Total Operating Revenue:

Gross Additions to Plant / T. Op. Rev. C* C* Dividends on Preferred Stock / T. Op. Rev. C* Dividends on Common Stock / T. Op. Rev. E* C* Funds for Retirement of Securities / T. Op. Rev. C*


Notation: See text, pages 60-61.






-76
11
significantly larger than the gas operations of combination companies. The data in Table 4.9 indicate that the utilization of funds of the gas operations of the combination utilities is significantly larger than the utilization of funds of the single service gas companies; this comparison probably indicates relative size differences and not the effect of competition.

Competitive industries are expected to experience greater risk than monopolistic industries. Firms that experience greater risk are expected to have lower debt/equity ratios than firms that experience smaller risks. The data in Table 4.9 indicate that single service electric utilities pay significantly larger dividends per dollar revenue to common stock owners than the combined gas/ electric operations of the combination firms. This observation indicates that single service electric utilities have lower debt/ equity ratios than combination gas/electric firms and may indicate that single service utilities experience greater risk. Marginal Cost Pricing

Marginal cost is the change in total cost when output is increased one unit; marginal cost is the first derivative of the total cost function. A long run cost function reflects changes in costs when all resources are varied. A short run cost function reflects changes in costs when capacity is held constant. Long run marginal cost is the extra operating cost plus the extra capacity cost as output is increased; short run marginal cost is the extra operating cost as output is increased.


11See Table 4.3.






-77-


In the energy industry, "Total Operating Revenue" equals total costs since profits are regulated. For this study, total operating revenues-are defined as long run costs. "Operating Expenses" includes fuel purchased energy, storage (where applicable), transmission, distribution, customer accounts, sales, and administrative and general expenses. "Operating expenses" does not include maintenance, depreciation, depletion, taxes, interest or dividend expenses. For this study, operating expenses are defined as short run costs. Each of these costs is related to quantity (in KWH and M. Therms) output as approximations of the long run and the short 12
run cost functions.

Various regressions were computed in order to investigate

the specific characteristics of the cost functions. For instance, both logarithmic and semi-logarithmic forms of the data were investigated. Second degree equations are reduced to linear form if the logs of the data are regressed. In each case, the fit was no better using the logarithmic form than using the original data. This observation indicates that a linear relationship adequately explains the cost functions.


12A detailed study of the nature of the costs in the energy industry is a thesis in itself and is beyond the scope of this study. The definitions of costs that are used are intended only as rough, relative measures and not as precise, absolute measures of marginal cost. Numerous preliminary relationships were investigated by the author before selecting the definitions as stated. For example, a stepwise linear regression (program "BMD-O2R" of the Biomedical Computer Program package of the University of California) was utilized to sort out possible independent/dependent variable relationships. Independent variables measuring quantity output, fuel costs, efficiency, average wages, pension benefits, quantity consumption, taxes, size of plant, and a dummy indicating organization were regressed with each of the dependent variables. In each case, he quantity variable was entered first. No noticeable increase in R was observed as other independent variables were added. Although crude, the assumption was made that a single independent variable (quantity produced) determines






-78

Further verification of the linear relationship involves
2
a polynomial regression which computes Y = a + blx + b2x +
k 13
bk 0 + e. This program ran six steps and the error sum of squares compared to the error sum of squares of the linear regression. There was no noticeable reduction in the error term. In fact, for the single service electric firms there was a significant reduction in the fit of the curvilinear function.

Since both the short run and the long run cost functions are linear, both the short run and the long run marginal costs (the first derivatives of the total cost functions) are constant. Table

4.10 presents the results of the regressions; Figure 4 illustrates graphically the general form of the cost functions. The constant long run marginal cost is constant to plant (or system) capacity and then it is undefined. The short run marginal cost is lower than the long run marginal cost in each case. Both the short run marginal cost and the long run marginal cost are higher in the combination computations than in the single service computations. The cost curves of the single service utilities are thus lower than the cost 14
curves of the combination utilities. The single service utilities utilize resources more efficiently than the combination utilities.



the value of the dependent variable (cost). This relationship is summarized as: Cost = f (quantity).

13BMD-05R of the Biomedical Computer Program.
14
It was previously verified that the costs were also higher for the combination companies. See especially Table 4.6.







Table 4.10


Results of Linear Regressions Where the Slope (b)
Estimates the Marginal Cost
(Form: Y = a + bX)


CLASS OF COMPANY


DEPENDENT VARIABLES:


Operating Expense:
R 2



Total Operating Revenue:
2
R


ELECTRIC SIDE

Straight Electric Combination


.06848 .08115 (.9219) (.8810) .16495 .20275 (.9426) (.9117)


GAS SIDE

Straight Gas Combination


.02892 .03962 (.7322) (.9411) .04347 .05490 (.7481) (.8644)









SMC


I
Q .203


.1651 1 TMC


.081


Straight Electric


SMC


_ I_ - LMC


Combination


SMC


.055
.040


4LMC


Straight Gas


SMC


LMC


Combination


Figure #4

Estimates of Marginal Costs


.068


Q








.043 .029






-81

Profit maximization requires lower prices in markets with higher price elasticities and higher prices in markets with greater price inelasticitids. A vertical summation of peak and off-peak demands indicates that prices for off-peak customers may be below the long run marginal cost and still be optimal as long as they are above short run marginal cost. A utility may discriminate between classes 15
of customers in an attempt to maximize profits. For example, industrial customers have a greater potential to utilize alternative sources of energy than residential customers. The industrial demand is thus expected to be more price elastic than the residential demand. Additionally, however, industrial customers consume significantly larger quantities of energy than residential customers and thus have lower fixed costs per unit output and lower fixed costs per customer. In an attempt to attract industrial customers, a utility may lower the industrial rate and raise the residential rate. Lower rates to industrial customers which have lower costs are not discriminatory per se, since they may reflect the lower costs. But if the rate differential is greater than that suggested by the cost difference, the utility has discriminated with respect to rates. This discrimination is caused by the desire of any utility (whether a combination or a single service utility) to maximize profits and not by the competition between the suppliers of energy.

Table 4.11 suggests a measure of this discrimination.

The industrial average price is divided by the residential average


15Both the vertical summation of the demands and the concept of discrimination were discussed in Chapter II.





Table 4. 11

-Average Price as a Percent of Long


ELECTRIC OPERATIONS GAS OPERATIONS


oil


p' U U



COMBINATION UTILITIES:

Residential 2.4064 119% 11.894 202%
- - - 51% - - 59% - - - 44% - - 114% Industrial 1.2132 60% 5.191 88%



SINGLE SERVICE UTILITIES:

Residential 2.1698 132% 12.833 295%
- -- 49% 67% - - - 44% - - 165% Industrial 1.0672 65% 5.632 130%






-83

price (I.A.P./R.A.P.) to determine the industrial rate as a percentage of the residential rate. In both the electric and in the gas computations, similar differentials exist between industrial and residential rates relative to the class of customer. The industrial rate is (approximately) the same percentage of the residential rate for both the single service utilities and for the combination utilities.

Further investigation of Table 4.11 indicates that the difference in rates may be greater than the differences in costs. The long run marginal cost is compared to the average price for each class of customer. A figure is computed that indicates the ratio of average price to marginal cost (A.P./LRMC). In each instance the industrial ratio is less than the residential ratio simply indicating that the industrial rate is less than the residential rate. A second measure of the rate "spread" is computed by subtracting the industrial A.P./LRMC ratio from the residential A.P./LRMC ratio. In both the electric computations and in the gas computations, the "spread" between the residential rates and the industrial rates is greater for single service utilities than for combination utilities. This difference in "spread" is attributable to differences in (marginal) costs between single service firms and combination firms. This difference in "spread" between industrial and residential prices indicates single service utilities have greater differences in rates relative to (marginal) costs (though not relative to classes of customers) than combination gas/electric utilities. This observation may suggest that a single service utility builds a more discriminatory rate structure than a combination utility in






-84

an attempt to take (industrial) customers from their competitors. Alternatively, this observation may suggest that a single service utility builds a more optimal rate structure as it is forced, by competition, to charge prices that more nearly reflect marginal costs. This discrepancy indicates that additional research is necessary on this subject.

Summary

In this chapter, certain variables have been studied as indications of the degree to which combination utilities compete with single service utilities. Where straight electric companies compete with straight gas companies, rates are lower and sales are larger than where combination companies control the supply of both sources of energy. Sales expenses are greater for separate firms indicating greater non-price competition. Single service companies have lower average costs which may be the result of competitive pressures forcing greater efficiencies. This greater efficiency is further indicated by higher KWH sales per dollar plant and a larger output per dollar input for the single service utilities even though these firms are unable to use their general plant in a joint manner. Single service electric companies spend a larger proportion of revenues on dividends to common stock owners, thus indicating that they experience greater risk.

Comparisons relating to the gas operations are not as

conclusive for several reasons. First, the gas operations of combination utilities are small relative to the total operations. Second, some data are presented in a combined manner from the combination






-85

firms and are not divided between the gas or the electric operations. Since the combined gas/electric operations of combination companies are larger than thd single gas operations of straight gas companies, the lack of detailed information makes relevant comparisons difficult, if not impossible. Taking these limitations into consideration, some comparisons can be made. Price competition does not appear as significant a factor in the gas computations as in the electric computations. No significant differences are observed in average costs even though wages are higher and pension benefits are greater for the combination companies. The inability to utilize general plant for both electric and for gas operations creates greater average expenditures for general plant for single service utilities relative to combination utilities.

This study further indicates that not only are average

costs lower, but the entire set of cost curves is lower, for single service utilities relative to combination utilities. The differences in costs create a potential discriminatory situation. Absolute rate differentials are not significantly different between single service and combination utilities. Relative rate differentials are significantly different however: single service utilities charge their industrial customers a rate that constitutes a smaller percentage of marginal cost than combination utilities charge their industrial customers.

These findings support the theory that combined gas and

electric utilities possess greater market power than separate ownership of the two sources of energy allows. Competition, especially





-86

with regard to the supply of electricity, appears to lower costs and may cause discriminatory pricing. These findings are in general agreement with the studies conducted by Owen, Mann and Collins.













CHAPTER V

THE REGIONAL ANALYSIS



Certain differences were identified in the previous chapter that appear to be associated with competition between single service and combination utilities. The analysis to this point indicates that competition between single service and combination utilities results in lower prices and higher outputs. Straight utilities seem to enjoy greater efficiency, as measured in terms of both average and marginal cost. These utilities appear to assume greater risk as measured by payments to common stockholders. Furthermore, it is observed that in the geographic areas where the ownership of the sources of energy are separated, sales expense is larger. There also appears to be more widespread rate discrimination between classes of customers in those areas where ownership is separate.

It was not proven in Chapter IV that these differences

are caused by the separation of ownership. It is possible that they are attributable to geographic or demographic factors which may differentially characterize the service areas of single service utilities as opposed to those of combination companies. As a crosscheck on the results of Chapter IV, this chapter classifies the firms into geographic regions and again tests the hypotheses stated in that chapter. If the differences found in Chapter IV are significant in the regional analyses, the hypotheses are supported. If, however, the differences found in Chapter IV are not significant in the regional analyses, the hypotheses are rejected.
-87-






-88

Statistical techniques similar to those employed in the aggregative analysis are utilized in this chapter. Since the samples are small,'a "Student t" text is employed to compare sample means.


The Regional Distribution

The data are divided into the Federal Power Commission's "National Power Survey Regions"; these regions are presented in Figure #5. Most firms are located entirely within one region although some have sales in overlapping areas. Every attempt is made to place these companies in the region in which most of their sales are made. To expedite the statistical procedure, each geographical region is assigned a number. Thus, the Northeast is called Area #1, the East Central is called Area #2, etc.

Table 5.1 shows both the absolute and the relative composition of firms within each geographical region. It can be seen that 29 percent of all combination gas/electric companies, 28 percent of all single service electric companies and 29 percent of all straight gas companies are located in Area #1 (the Northeast) where population and industry are concentrated. The uniform distribution between the classes of utilities that prevails in the Northeast does not hold in other areas. For example, single service electric companies outnumber combination gas/electric companies fourteen to five in Area #2, yet combination utilities outnumber single service electric companies fourteen to three in Area #3. Furthermore, two-thirds of all the combination utilities (66%) are located in the northeastern part of the United States (Areas #1, #2 and #3); only one-third of











FED-RAL O oArea #1: NATIONAL POWER SURVEY REGIONS Northeast Area #3: West Central
., , Area #2: East Central

--1~ F ~ WST e N T RAL


Area #4:.. . West ,- T - .,Q EST 32 ,0


Area #6: Southeast


Figure #5




Table 5.1


AREA #1: Northeast


The Geographic Distribution of Utilities
AREA #2: East Central AREA #3: West Central


29% 28% 29%


5



10% 19% 12%


AREA #4: West AREA #5: South Central 16


AREA #6: Southeast


18% 15% 18%


12% 21%


4% 13% 12%


Legend:
L = COMINATION


L j = STRAIGHT ELECTRIC P7 = STRAIGHT GAS


5



4% 15%


27%




Full Text

PAGE 1

A COMPARATIVE ANALYSIS OF THE ECONOMIC ORGANIZATION OF GAS AND ELECTRIC SUPPLY BY JOHN ANGUS ANDERSON A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1973

PAGE 2

ACKNOWLEDGMENTS The writer wishes to express his deepest appreciation to the members of his supervisory committee. A special thanks is offered to Dr. Milton Z. Kafoglis, Chairman, who suggested the subject and devoted countless hours in editing the manuscript. Without his help and constant motivation, this dissertation would not have been completed. Drs. Charles W. Fristoe, Norman G. Keig and John H. James each encouraged the writer when progress was slow and offered invaluable suggestions throughout the study. Finally, I wish to acknowledge the loyalty of my wife, Peggy, who willingly sacrificed during the preparation of this paper.

PAGE 3

TABLE OF CONTENTS ACKNOWLEDGMENTS ii LIST OF TABLES v LIST OF FIGURES vii ABSTRACT viii I. INTRODUCTION 1 Economic Theory and Public Utility Economics 2 Outline of the Study 6 Economic Issues 6 Related Research 7 Aggregative Analysis 9 A Regional Analysis. 9 Policy Implications 10 II. ECONOMIC ISSUES 12 Economies and Diseconomies 13 Joint Costs 13 Load Factors 15 Competitive Stimulus 16 Costs of Divestiture 21 Summary 23 Marginal Cost Pricing 24 The Marginal Cost Pricing Principle. . . 24 Marginal Costs and the Energy Industry 26 Pricing in the Energy Industry 29 Summary 32 III. EVALUATION OF EARLIER WORK 35 The Study by Bruce M. Owen 35 The Study by Patrick C. Mann 37 The Study by William H. Collins 40 The Study by the N.E.R.A 43 Summary of Owen, Mann, Collins and the N.E.R.A 47 The Study by Joe D. Pace 48 Summary of Earlier Work 52 iii

PAGE 4

Page IV. AGGREGATIVE ANALYSIS 53 The Data 53 Methodology 56 Table Notations 60 Aggregative Characteristics of the Firms 61 Findings of the Aggregative Analysis .... 67 Economies and Diseconomies 67 Marginal Cost Pricing 76 Summary 84 V. THE REGIONAL ANALYSIS 87 The Regional Distribution 88 Aggregative Characteristics of the Firms 91 Findings of the Regional Analysis. . . * . . 94 Economies and Diseconomies 94 Marginal Cost Pricing 104 Summary of the Regional Analysis 110 VI. POLICY IMPLICATIONS 112 Data and Methodology 113 The Aggregative Analysis 113 The Regional Analysis 114 Comparison of Expected Benefits and Costs 117 Alternative Policy Suggestions 119 APPENDICES A. RESULTS OF THE STATISTICAL COMPUTATIONS: THE AGGREGATIVE ANALYSIS 121 B. RESULTS OF THE STATISTICAL COMPUTATIONS: THE REGIONAL ANALYSIS 126 BIBLIOGRAPHY 151 iv

PAGE 5

LIST OF TABLES Table Page 2.1 A MEASUREMENT OF THE SUBST ITUTABILITY OF GAS AND ELECTRICITY 18 2.2 RELATIONSHIP BETWEEN GAS-ELECTRIC PRICE AND SALES RATIOS FOR THE INDUSTRIAL SECTOR, 1962 AND 1967 20 4.1 THE RESULTS OF THE SURVEY OF THE UTILITIES . . 55 .4.2 AGGREGATE CHARACTERISTICS OF THE UTILITIES: DEMOGRAPHIC AND SALES VARIABLES 62 4.3 AGGREGATE CHARACTERISTICS OF THE UTILITIES: EXPENSE, PLANT AND TAX VARIABLES .... 63 4.4 AGGREGATE CHARACTERISTICS OF THE UTILITIES: EMPLOYMENT AND FINANCIAL VARIABLES ... 66 4.5 AVERAGE CONSUMPTION AND PRICE 68 4.6 AVERAGE COSTS 70 4.7 AVERAGE WAGES 71 4.8 PHYSICAL PLANT 73 4.9 FINANCIAL 75 4.10 RESULTS OF LINEAR REGRESSION WHERE THE SLOPE (b) ESTIMATES THE MARGINAL COST 79 4.11 AVERAGE PRICE AS A PERCENT OF LONG RUN MARGINAL COST 82 5.1 THE GEOGRAPHICAL DISTRIBUTION OF THE UTILITIES 90 v

PAGE 6

Table Page 5.2 AGGREGATE CHARACTERISTICS OF THE UTILITIES: DEMOGRAPHIC AND SALES VARIABLES 92 5.3 AGGREGATE CHARACTERISTICS OF THE UTILITIES: EXPENSE, PLANT AND TAX VARIABLES .... 93 5.4 AGGREGATE CHARACTERISTICS OF THE UTILITIES: EMPLOYMENT AND FINANCIAL VARIABLES ... 95 5.5 AVERAGE CONSUMPTION AND PRICE 97 5.6 AVERAGE COSTS 99 5.7 AVERAGE WAGES 101 5.8 PHYSICAL PLANT 102 5.9 FINANCIAL 103 5.10 RESULTS OF LINEAR REGRESSIONS WHEN THE SLOPE ESTIMATES THE MARGINAL COST 105 5.11 AVERAGE PRICE AS A PERCENT OF LONG RUN MARGINAL COST 108 vi

PAGE 7

LIST OF FIGURES Figure Page 1. THE SHORT RUN/LONG RUN MARGINAL COST RELATIONSHIP 28 2. PRICING UNDER CAPACITY CONSTRAINTS: THE CASE OF A SINGLE PEAK CONSUMER ... 30 3. PRICING UNDER CAPACITY CONSTRAINTS: THE CASE OF BOTH PEAK AND OFF-PEAK CONSUMERS 31 4. ESTIMATES OF MARGINAL COSTS 80 5. FEDERAL POWER COMMISSION NATIONAL POWER SURVEY REGIONS 89 vii

PAGE 8

Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy A COMPARATIVE ANALYSIS OF THE ECONOMIC ORGANIZATION OF GAS AND ELECTRIC SUPPLY by John Angus Anderson August, 1973 Chairman: Milton Z. Kafoglis Major Department: Economics A competitively organized economy is expected to result in a more efficient allocation of resources and more efficient production. The public utility sector possesses unusual or exceptional characteristics that make both perfect competition and unregulated monopoly undesirable. Senator Metcalf has introduced a bill in the U. S. Senace (S. 403) that seeks to create a blend of competition and regulation in the energy sector. This bill would divorce combination gas/electric utilities, presumably increasing the degree of competition in the energy industry; however, it would not abandon traditional regulation. This thesis examines the hypothesis implicit in the Metcalf bill that separate ownership of gas and electricity supplies will in fact generate economies associated with competition. viii

PAGE 9

Certain economies of dual or joint operation are available to combination utilities that are not available to single service utilities. To the 'extent that these economies are significant, they are expected to lower the costs of the combination firms relative to the single service firms. On the other hand, joint utilization of resources encompasses only a small portion of total operating expenses. Thus, average total costs may not be significantly affected if the dual utilization of the general resources is lost. Separate ownership of the sources of energy is also predicted to promote a competitive stimulus that sharpens the decisionmaking process of single service firms. However, managers of separate firms may be prompted to charge discriminatory prices in an attempt to gain sales at "destructively" low rates. Previous studies have conflicting conclusions and policy recommendations. It is this conflict that establishes the need for this thesis. Data are gathered from reports submitted directly by each utility. These data are more detailed and comprehensive than those available in existing publications. The data are subjected to "z" and "t" tests which are widely known and commonly accepted as a basis for evaluating differences in variables. Few differences are observed in the relative sizes of the single service utilities and the electric operations of combination utilities; the single service gas utilities appear significantly larger than the gas operations of combination utilities. When the data are analyzed on an aggregative basis, rates are lower and the volume of sales higher for single service firms than for combination gas/electric firms. Sales expenses are ix

PAGE 10

greater, although average total costs are lower, for single service firms. Single service firms discriminate between types of customers to a greater extent' than combination firms. The findings conform to those reached in earlier studies which incorporated similar data and methodology. However, the distribution of the two types of utilities is not geographically uniform. Specifically, single service utilities are concentrated in the south and in the west where climate causes high per-customer consumption. In response to this spatial difference, the data are subdivided and studied on a geographic basis. Most of the differences that are observed as significant on a national basis are not significant on a regional basis. The findings of the national analysis that appeared to indicate that competitive forces in the energy industry are substantial now appear simply to indicate geographical or demographical characteristics peculiar to the individual firms and regions. On the other hand, the legal and financial costs of divestiture may be significant. Since the costs are expected to outweigh the benefits, the hypothesis examined is rejected. More stringent enforcement of present legislation that considers each case on its own merits is more realistic. x

PAGE 11

CHAPTER I INTRODUCTION The energy industry is characterized by different industrial organizations in various geographical areas of the United States. In some areas the supply of both electricity and natural gas is provided by a single firm (a "combination" utility); in other areas the supply of electricity is provided by one firm while the supply of natural gas is provided by another firm (a "single service" or a "straight" utility). Certain potential economies are predicted where combination utilities exist, but in a regulated atmosphere there may be no motive to attain fully these economies. On the other hand there is a greater element of "potential" competition where single service utilities exist, but this organization may frustrate the attainment of certain economies. This thesis examines the comparative efficiency of alternative organizational arrangements that have been proposed recently regarding the production and the distribution of electric power and natural gas. -1-

PAGE 12

-2Economic Theory and Public Utility Economics In an enterprise economy, a multitude of privately owned and managed busineas enterprises are expected to play a central role in the determination of the course and the character of economic activity. Economic theory predicts that a competitively organized economy results in an efficient allocation of resources in the sense that the standard Paretian conditions are fulfilled. The encouragement of competition is a stated objective of public policy in the United States. However, certain segments of the economy, especially the public utility industries, possess unusual or exceptional characteristics and have had distinctly monopoly policies applied to them. Public utilities produce widely used "necessities" for which there are no close substitutes and are, to some degree, natural monopolies. These industries are held to be natural monopolies because (a) advantages to large-scale firms are such that unrestricted competition would lead to exclusive occupancy of individual markets by one or a very few firms, and/or (b) unrestricted competition among several or many firms in a market would, because of the technological characteristics of the industries involved, have deleterious effects on the welfare of buyers.* The energy industry has traditionally been considered a public utility and special regulatory policies have been applied to this sector of the economy. Legislative bodies have granted ^exclusive market regions and have blocked entry in order to J. S. Bain, Industrial Organization (New York: John Wiley and Sons, Inc., 1968), page 581.

PAGE 13

-3guarantee monopoly organization, presumably on the assumption that available economies of scale and other characteristics of production and distribution are inconsistent with competitive organization. Protection of the consumer has been shifted to the regulatory agencies. Criticism of the effectiveness of regulation abounds. Joe Bain, for example, has summarized the deficiencies as follows: (1) Commissions have not been able to consistently impose rates at a competitive level. (2) Regulatory commissions in general do not have the power to determine broadly the rate of output or the scale of operations of the regulated firms, thus abandoning important areas to managerial discretion. (3) Regulation seriously reduces the usual profit and efficiency incentives which exist in the private sector. (4) By implication, regulation has the responsibility of supporting a rate structure that will secure for firms a rate of return as determined by statutes, courts and commissions. This policy may perpetuate the provision of redundant services and may forestall economically desirable adjustments in the relative volumes of services supplied or facilities maintained by different components of the industry. Though regulation has not been ideal, unrestricted competition does not appear to be an attractive alternative. — J' S. Bain, Industrial Organization , pages 640-645.

PAGE 14

-4A number of economists have maintained that unregulated enterprise in the public utility sector is likely to be as, if not more, effective than regulated enterprise. For example, George Stigler and Claire Fried land found that regulation of electric utilities 3 was redundant and ineffective for two reasons. First, individual utilities do not possess significant long-run monopoly power. Instead, they face competition from other energy sources in a large porportion of their products uses, and they face competition from within their own industry. In the long run, their industrial users (and hence many of their domestic users) have sufficient mobility to ensure a competitive outcome. Second, regulatory bodies are incapable of forcing tte utility to operate at a specific (presumably optimal) combination of output, price and cost. Stigler and Friedland state: "The theory of price regulation must, in fact, be based on the tacit assumption that, in its absence, a monopoly has exorbitant power. 1 * 4 These authors conclude that electrical utilities do not possess such power. Since neither of the organizational extremes appears to be feasible, alternative suggestions incorporating a blend of competition and regulation have been proposed. Thus, the transportation industry is regulated but in an environment of "balanced competition" where entry is controlled but not closed. It has been argued that a similar blend of regulation and competition should be 3~ G. J. Stigler and C. Friedland, "What Can Regulators Regulate? The Case of Electricity," The Journal of Law and Economics , Volume 5 (1962), pages 15-167 ~ 4 G. J. Stigler and C. Friedland, "What Can Regulators Regulate? The Case of Electricity," pages 15-16

PAGE 15

-5applied to the energy sector. In response to this view, Senator Metcalf introduced a bill (S.403, 92nd Congress, First Session) that ammends the Federal Power Act. In this bill it is declared that: ". . . it is in the national interest to promote interenergy competition between electricity and gas whenever possible.""' This bill makes it unlawful to own or to operate facilities used in the production, generation or distribution of both electricity and natural gas. It would divorce combination gas and electric firms, presumably increasing the degree of competition in the energy industry, but would not abandon traditional regulation. It would seek to add a "free enterprise" competitive dimension to the allocative process. To the extent that electricity and natural gas do in fact compete, the intensified competition would result in superior resource allocation. This thesis examines the hypothesis implicit in the Metcalf bill that: (a) separate ownership of gas and electricity supplies will in fact promote competition, and (b) this increased competition will encourage efficiency and lead to an improved rate structure. If the hypothesis is affirmed, the policy embodied in the Metcalf bill can be recommended. If the hypothesis cannot be affirmed or is refuted, this policy cannot be recommended. Our criterion, of course, is economic efficiency. Any conclusions developed must, therefore, be further evaluated in terms of the 5 S. 403, "A bill to prohibit certain combinations and control between electric and gas utilities," as presented in Combinat ion Utility Companies , Hearings before the Subcommittee on the Judiciary, United States Senate, Washington: U. S. Government Printing Office, 1971, page 1.

PAGE 16

-6broader objectives of social policy relative to economic organization and equity. Though the author's impressions and opinions on these matters are occasionally expressed, the scope of the substantial work is limited to a test of the hypotheses described above . Outline of the Study Economic Issues In order to evaluate the hypotheses, a framework suitable for statistical testing must be established. This is the primary objective of Chapter II where economic theory is employed to predict possible outcomes under the alternative organizations. In Chapter II the expected economies and diseconomies of the "straight" utilities are compared with the relative economies and diseconomies of the "combination" utilities. Combination firms are predicted to have lower common costs due to joint utilization of assets and lower fuel costs due to higher load factors. These factors are predicted to lower the cost curves of combination companies relative to single service utilities. On the other hand, single service firms have both the ability and the incentive to increase their load factors as much as do the combination gas/electric firms and the joint utilization ofresources may encompass only a small portion of the total operating expense. Moreover, it is sometimes asserted that regulatory agencies do not recognize technological developments in-the energy industry and thus do not force the realization of potential efficiencies; competition, on the other hand, may force

PAGE 17

-7such efficiencies. To the extent that these factors are significant single service utilities are predicted to have lower costs and greater efficiencies than are the combination gas/electric utilities Furthermore, Chapter II establishes a basis for evaluating and comparing the rate structures of combination utilities and straight utilities. The marginal cost pricing principle is described, criticized and finally defended as an ideal criterion. However, to the extent that indivisibilities, externalities, and decreasing costs are significant, this criterion must be modified. The "second-best" theory is briefly discussed and accepted although, as a practical matter, a strict application of second-best is also unattainable. Finally, a realistic and attainable guide, developed in the light of both theory and practice, is established. Though crude, it appears to be consistent with the goal of welfare maximization. Related Research Previous studies have investigated the hypothesis with which this study is concerned. These works are characterized by conflicting conclusions and policy recommendations. Studies by B. M. Owen, P. C. Mann, W. H. Collins, Jr., the National Economic Research Associates and J. D. Pace are analyzed, evaluated and compared in Chapter III. 6 Owen, Mann, Collins and the NERA utilized data for the entire United States in an effort to compare the average values of relevant variables under the alternative 6 Citations for these studies are in Chapter III.

PAGE 18

-8organizational arrangements. The conclusions of several authors are similar: straight utilities were found to charge lower average prices and sell larger quantities of electricity per customer than did the combination utilities. These findings have provided a basis for recommending that the combination utilities be divorced or separated into two competing units. This proposed divorcement of the ownership of the supplies of energy is expected to increase competition. The implication is that the advantages of competitive organization are greater than the advantages of joint operation and production. Differences in geographical locale may cause differences in measured variables. Under cursory examination these measured differences may appear to be associated with competitive forces; in actuality they may be the result of geographical concentration and thus demographic influences. For example, if single service utilities are concentrated in warm regions, climate is expected to cause high per-customer consumption regardless of the organizational structure. A statistical analysis that segments data only by industrial organization ignores these geographical influences. Owen, Mann and Collins fail to consider the geographical location of the firms and thus fail to recognize the resulting demographic characteristics that each firm faces. The Pace study did recognize the impact of demographic characteristics upon the findings of the national studies. This study utilized a complex interrelationship of demographic variables and found that the "clear" indications of beneficial competition were in fact insignificant. The conclusions of the Pace study

PAGE 19

-9were in conflict with those of Owen, Mann and Collins. It seems clear that additional research is necessary in order to reconcile these conflicting conclusions. Aggregative Analysis The analysis of Chapter IV provides a basis for comparing this study with other work. Certain variables are examined in an attempt to measure the degree to which combination utilities do in fact compete with straight utilities. The national data and the statistical methodology incorporated in this chapter do not differ significantly from those employed by other authors. Since the aggregative analysis incorporates methods and conclusions similar to those of previous studies, the data of this thesis are assumed to be similar to those of the previous works. The findings in Chapter IV support the view that the combined gas and electric utilities possess greater market power than the single service utilities. Competition between two straight firms, especially with respect to the supply of electricity, appears to force a greater degree of efficiency than the regulatory process itself can achieve. To this point, the policy proposal to require divestiture of gas and electric operations appears to have support. Certainly there does not appear to be justification for a policy that would allow the formation of any new combination utilities . A Regional Analysis In Chapter V the statistical methodology of Chapter IV is applied to the same series of data, but on a geographical basis.

PAGE 20

-10The objective of the geographical break-down is to isolate the effects that demographic differences have on utilities in various parts of the United States. For example, as explained above, on a national basis single service utilities have lower prices and higher quantities. If single service utilities are concentrated in warm regions, climate is expected to increase consumption of energy. The larger consumption would, because of the step-down nature of the pricing structure, reduce the average revenue per unit of output. The conclusion reached above that the single service utilities have higher quantity sales and lower average prices is not altered. But this conclusion is now explained on a totally different basis. If the conclusions of the regional analysis are not significantly different from the national analysis, competition would appear to be a significant force. If, on the other hand, significant differences exist between the national and the regional studies, the existence of demographic differences must be acknowledged. Antitrust as a policy recommendation may not bring a reduction in price, increase in output or increase in efficiency since differences are caused by demographic characteristics that are independent of market structure and ownership. Chapter V supports the latter interpretation. Policy Implications Chapter VI brings the analysis together by comparing the expected costs with the expected benefits. The benefits of divestiture appear to be relatively insignificant; the costs appear to be'

PAGE 21

-11potentially substantial. The hypothesis examined in this thesis is therefore not supported. The primary hypothesis that requires divestiture is rejected; the formation of new combination utilities, however, is not categorically accepted. Each case should be considered on its own merits.

PAGE 22

CHAPTER II ECONOMIC ISSUES The hypothesis that is investigated in this study suggests that the performance of an industry that competes (even though imperfectly) is superior to the performance of an industry that is monopolistic. In this chapter alternative types of organization are discussed and relevant economic theory is presented in order to develop a basis for the measurement of the economic performance of the various organizations. A perfectly competitive economy with no external economies, joint supplies or indivisibilities achieves a Pareto optimal equilibrium. In a Pareto optimal equilibrium (1) the total cost of the optimal output is at a minimum and (2) the price is equated with marginal cost. 1 Although rough, these criteria are utilized in this study as feasible yardsticks by which alternative organizations of the energy industry are compared. Each of these criteria are examined in turn. i V. Pareto, Manuel d'Economique Politique (2nd Edition, Giard, Paris, 1927). The marginal conditions necessary to achieve welfare maximization have been extensively developed since this original work and are beyond the scope of this study. See for example: N. Kaldor, "Welfare Propositions of Economics," Economic Journal , XLIX (September, 1939); P. Samuelson, Foundations of Economic Analysis (Cambridge: Harvard University Press, 1948); and especially M. A. Reder, Studies in the Theory of Welfare Economics (New York: Columbia University Press, 1947). These references are only a representative sample of numerous citations available. -12-

PAGE 23

-13Economies and Diseconomies Economic efficiency requires the minimization of costs in producing the optimal output. Certain economies and diseconomies are expected under alternative organizations of the energy industry. These economies are joint costs, load factors and the competitive stimulus; the diseconomy is the cost of divestiture. Joint Costs Combination gas/electric utilities are potentially able to operate more efficiently with respect to their resources because they are able to utilize their administrative, general and other fixed resources in the distribution of both electricity .and gas. For example, the customer accounting process, including meter-reading, bill preparation and handling of payments can be accomplished more efficiently by one company than by two. The reasons for the increase in efficiency are obvious: only one .meter reader per customer (instead of two) is required; only one bill per customer (instead of two) is prepared; and only one .payment per customer (instead of two) is received. 2 Additionally, Frederick T. Searls, Vice President and General Counsel, Pacific Gas and Electric Company, estimated that it would cost an additional $15 million a year to carry on the entire customer accounting process, including meter-reading, bill preparation and the handling of payments by two separate companies in the place of his one company. (See Combination Utility Companies . Hearings Before the Subcommittee on Antitrust and Monopoly, 92nd Congress, 1st Session, 1971, Washington, U. S. Government Printing Office, at page 64.) Pacific Gas and Electric Company supplies gas and electricity to about 2 million customers which requires a mailing of 24 million bills each year. If this utility were divided into two companies, it would be necessary to increase the annual

PAGE 24

-14there exist potential economies in a single supervisory and administrative organization and a single set of offices, warehouses, service centers, and the like, instead of two. One application for service is more convenient for the customer as well as economical for the company, which can send one man, in one truck, to turn on both meters and initiate both services. In new installations, a single trench can be utilized for both electric and gas lines. Both gas and electric operations can be controlled from a single energy control center that contains weather forecasting, dispatch groups, and other related control functions. All these economies are expected to reduce average costs of production and increase the utilization of fixed facilities, permitting the utility to operate in a more optimal manner . It is possible, however, that the competitive stimulus encouraged by separate ownership would produce greater overall gains in efficiency. If the joint costs are a significant proportion of the total costs, the net advantage is expected to accrue to the combination utilities. If the joint costs are only a small proportion of the total costs and the competitive stimulus is a significant force, the net advantage is expected to accrue to the straight utilities. mailing to 48 million bills. At the present rate of 8 cents for first-class mail, this would mean an additional cost of almost $2 million in postage for the mailing of the bills alone. Mr. Searls further stated that there are substantial economies to be realized in having a single supervisory and administrative organization and service facilities. He concluded, "We are confident that the actual savings are a very substantial amount." (at page 65)

PAGE 25

-15Soloman Freedman, Director of the Division of Corporate Regulation, Securities and Exchange Commission, studied the expected savings due "to the joint utilization of resources as a 3 percent of the total expenses and concluded: the. bulk of the expenses of operating an electric utility or gas utility is entirely unrelated to any combined functions of the two.... It is in only a relatively narrow category that a combined gas and electric company can effect savings in operating expenses ... the claimed "loss of economies" has not been found, by the SEC, to be substantial. This statement is in obvious contradiction with the statement cited in the footnote above. Mr. Searls asserted that the savings would be of a "substantial" magnitude; Mr. Freedman suggests that the economies have not been found. This contradiction will be examined in Chapters IV and V. Load Factors Additional economies are potentially available to combined gas/electric utilities since the combined gas requirements for final customer use and for electric generation are under a common control. The load factor is defined as the average use of facilities 4 as a percentage of the maximum use. Combination gas/electric utilities are predicted to have greater load factors, thus lower ^fuel costs and lower average total costs, than single service utilities. For example, assume that the temperature in an energy .system is presently cold, but expected to drop sharply for a short rc-r er.c-r: 3 Combination Utility Companies , page 408. 4?aul J. Garfield and Wallace F. Lovejoy, Public Utility Economics (Englewood Cliffs, New Jersey; Prentice-Hall, Inc., 1964), page 153.

PAGE 26

-16period. The gas controller estimates the increase in gas use that will result and calculates the amount of curtailment in gas fuel for electric generation that will be needed so that enough gas will be available to meet gas customer requirements. This information is given to the electric dispatcher who determines whether to increase hydroelectric production or to burn oil in place of gas. It is predicted that full use is made of gas transmission facilities with this organization which results in a more optimal use of the system's resources. On the other hand, single service electric utilities are able to negotiate interruptable contracts with (single service) gas companies at low rates. The gas companies increase their load factors (by the interruptable nature of the contract) and the electric companies lower their fuel costs. The potential significance of this economy therefore lies not in the market structure, but in the willingness and the ability of the single service utilities to negotiate fuel contracts. <-. An indication of this efficiency is the average dual cost per BTU of heat. In Chapters IV and V, fuel costs per BTU are compared as a measure of this economy. -. Competitive Stimulus If gas and electricity are substitutable products, single service electric utilities compete with single service gas utilities for energy customers. The competition may force a sharpening of the decision-making process and therefore a more efficient overall operation.

PAGE 27

-17When the Public Utility Holding Company Act and the Federal Power Act were passed in 1935, the competitive interface between gas and electricity was far more limited than it is today. Additionally, both industries were more local in scope. After World War II, transcontinental pipelines and high voltage transmission lines transformed these two industries into truly interstate operations. Over recent decades, technological changes have heightened the potential competition between gas and electricity; the cross elasticity of demand for electricity and gas has increased. One study that indicates the substitutability of gas and electricity relates the monthly charge for 500 KWH in various metropolitan areas to the percentage of homes in each area using certain electric appliances. The higher the monthly charge for electricity, the lower is the proportion of homes with electrical appliances. Table 2.1 presents representative data from this study. The author stated: "Using a statistical sample of over eighty S.M.S.A.'s and the residential gas rate, I found that the cross elasticity value for water and space heating was in excess of unity. " To a significant extent the gas and electric industries presently provide substitutable outputs and therefore effective competition between the suppliers of gas and electricity is anticipated . "*Dr. John W. Wilson, Department of Social Sciences, U. S. Military Academy, as quoted in Combination Utility Companies , page 99.

PAGE 28

4J CO (IJ r^vocMCT>vo<)-irivo -jOvtnaiHricnmN ooooi^-cococv4CMi-& S3 o u s C 3 U co SO -H CO fn CTi CO O £ W CO ffi 4-1 i-i m a) o u pq o u 4J o CO m 3 m o co pq C r a a» w c O U 60 CN CO 3 CTv i— 1 O -H Ci CO 1—1 a u CO tc CO o c H CO 4) a cr. in co Hi H C CO O 3 c; u CO u 3 a id O a; i •p i-i A3 o CO cr. +j a CO -r; •H <+4 > O 4-1 4J •rH •r-i T— 1 3 d •H 3) 4-J 3) 9 3 S pq 3 O O o Pw ~ CD u 0 i-i o C5 3 CO 3 J-i g Q) ^ ^3 E a a> o o 3 fa u co co o 3 O CO

PAGE 29

-19David Smith recently studied the substitutability of gas and electricity by correlating the price ratios of gas and elec6 tricity with their respective sales ratios. Table 2.2 presents the findings of this study with regard to industrial customers. The elasticity of product substitution between gas and electricity exceeds unity at all relevant gas/electric price ratios and is 7 higher for the 1967 data than for the 1962 data. Dr. Smith concludes "that the increase in the elasticity of substitution between these two energy sources over this period of time can be attributed to intense efforts on the part of both gas and electric industries to penetrate each other's market in the industrial 8 sector." Hubert H. Nexon, Vice President, Commonwealth Edison Company, asserted: "... competition itself tends to sharpen the decision making process. And conducting either the gas or the electric business is sufficiently complex to require undivided 9 management attention." Don Cook of the American Electric Power Association stated: "It's hard enough to run one company in one business well without trying to run two that are natural 6~~ Dr. D. B. Smith, "Inter-Energy Competition Between Gas and Electricity in the Industrial Market: A Case for Deregulation?" Business Studies , North Texas State University, XII, No. 1, pages 8-9. ^The data were fitted to the hyperbolic form 1/Y = a + bX, and the elasticity of product substitution was defined as: E = dY . X = _bX P dX Y a + bX Q D. B. Smith, pages 8-9.

PAGE 30

-20CJ\ o r» vO CM ON 00 co on CO \Q O CM ON CO CM O CM CO r-l 3 \o \0 CM o co O I— iH CM CM co . o> On t-I r-i r-i r-l r-l --I 1-1 t-I — i r-l fN CM cn 1 i 1 1 1 1 1 1 i 1 I 1 i CM CO CO m cn o o CM cn cr. o vO in CM VO CO CM VO o CM o. CM CM OA i-l • • I-l i-i r— 1 r-l r-l r-l rH T-I r-l r-l r-l r-l CM i t 1 1 1 1 1 1 1 1 1 1 r~W O O TO -j< o js a « r ft •h a CJ ^ •H CO 4J O 10 b&H CO CO 4J H H cj W w Pi w u D O rC u u C3 O • •H i-l u u cu a> i-4 4-1 c CD u u 6 r3

PAGE 31

-21competitors ," 1 ^ The implication of these statements is that the competitive stimulus provided by separate operations forces increased efficiencies and lower costs. Regulatory commissions have had the implied responsibility to support rates which will secure for firms a normal return on their investments; they have not had the encouragement to recognize technological developments that increase competition if the effect is to lower the rate of return. 11 If technological developments heighten potential competition, divestiture may increase efficiency. Regulation may stifle competition and therefore decrease efficiency. Indications of this economy are examined in Chapters IV and V by comparing the costs and the cost curves for alternative organizations. Costs of Divestiture A diseconomy that will be incurred if society is to benefit from competition is the legal and financial costs of divestiture. If the gas operation of an existing combination firm is sold so that two single service utilities are created, existing debt and equity must be refunded. Present interest rates are significantly higher than the rates of the old bonds; the interest expense of the new (gas) utility will be greater than the interest expense of the gas operations of the old (combination) utility. Mr. Eugene Meyer, 12 Vice President, Kidder, Peabody and Company, stated: Ibid. , page 6. n joe S. Bain, Industrial Organization (New YorkJohn Wiley and Sons, Inc., 1968), page 642. 12 Combination Utility Companies , page 231.

PAGE 32

-22Whatever the arguments about the benefits of competition or of combination utilities so far as operating costs are concerned, it is apparent that the financial costs attendant to breaking up the combination companies would be devastating for the consumer under current market conditions. Mr. Meyer studies the proposed divestiture as it would be applied and concluded that it would cost $5,315,404,000 even assuming lower than market rates, sales prices at original costs, and no costs to 13 finance common plant items. At least two proposals have been suggested as a means by which this diseconomy can be minimized. First, if the state commission announced, in advance, that the new gas utility was not able to charge rates in excess of those justified by the original (low) interest, the sale price of the divested assets would reflect the previous (low) rate. Interest cost would not go up; book value 14 of the assets would go down. This does not imply a loss in actual value, however, as these assets are presently earning the low rates. Second, most of the utility bonds with low rates were issued many years ago and will mature soon. As they are paid off with new money, the average interest rate of combination utilities, as well as single service utilities, will rise. 15 13 Ibid. , page 233. : 14 Dr. John W. Wilson, in Combination Utility Companies , page 85. 15^. This point was discussed by Robert H. Willis, President, Connecticut Natural Gas Corporation, in Combination Utility Companies , page 383.

PAGE 33

-23Summary A framework has been established that compares the economies and the diseconomies of alternative organizations in the energy industry. Combination firms are assumed to have lower common costs due to joint utilization of assets and lower fuel costs due to higher load factors. Additionally, regulation is expected to require efficiency in the overall operation of all utilities, including combination gas/electric firms. These economies are predicted to lower both the average costs and the cost curves of combination companies relative to single service utilities. On the other hand, it is predicted that joint utilization of resources encompasses only a small portion of total operating expenses and that single service firms have both the ability and the incentive to increase their load factors. It is further asserted that regulatory agencies do not recognize technological developments in the energy industry and thus do not force the realization of potential efficiencies. To the extent that these factors are significant, single service utilities are predicted to have lower costs and greater efficiencies than combination gas/ electric utilities. Given these economies and diseconomies, it seems that there does not exist one attainable organization that is predicted to be superior to all others. Measurements of the economies and of the diseconomies are undertaken in Chapters IV and V to determine the relative significance of each.

PAGE 34

-24Marginal Cost Pricing Economic efficiency requires not only the minimization of costs in producing the optimal output, but also the equilibrating of price and marginal cost. The minimization of costs represents efficient production; the equating of price and marginal cost represents efficiency in the allocation of resources. The Marginal Cost Pricing Principle It is impossible, and in fact inappropriate, to develop the general theory of marginal cost pricing in this paper. This task has been vividly and lucidly fulfilled by many authors. ^ The principle states that where prices reflect marginal (opportunity) costs, consumers will purchase quantities of the goods thst. reflect an optimum allocation of resources, an optimum level of production and an optimum scale of plant. If some buyers are charged more than marginal (opportunity) costs for particular commodities, they will buy less than the optimum quantities. "Consumers who would willingly have had society allocate to its production the incremental resources required, willingly sacrificing See for example: K. J. Arrow, Social Choice and Individual Values (New York: Wiley & Sons, 1951); J. R. Hicks, "The Foundations of Welfare Economics," Economic Journal (1939); N. Kaldor, "Welfare Propositions in Economics and Interpersonal Comparisons of Utility," Economic Journal (1939); I. M. D. Little, "Review of T. Scitovsky, 'Welfare and Competition'," Econometrica (1952); M. W. Reder, Studies in the Theory of Welfare Economics (Oxford: Oxford University Press, 1947); or T. Scitovsky, "The State of Welfare Economics," American Economic Review (1951). These specific references are intended as a starting place only for reading and research. The literature literally abounds with references regarding the statement, development, criticism and refinement of the marginal cost principle.

PAGE 35

-25the alternative goods and services that those resources could have produced, will refrain from making those additional purchases because the price t-o them exaggerates the sacrifices . "^ Two anomalies to the principle must be recognized: (1) Prices must reflect all the (marginal) costs of production and the benefits of consumption for the principle to be valid. If a portion of the costs are borne by society or a portion of the benefits are received by society, the effect is 18 an underallocation of resources and a less than optimal output. (2) The rule does not produce optimal results if it is applied only partially. If portions of the economy are constrained in such a manner that price is not equated with marginal costs, the second-best theorem indicates that increases in welfare would not be assured by forcing other portions of the economy to equate price with marginal cost. E. J. Mishan, among others, developed the marginal cost principle into a practical policy proposal in spite of the problems *^A. E. Kahn, The Economics of Regulation (Volume I, New York: Wiley & Sons, 1970), pages 66-67. 18 See especially E. J. Mishan, "Reflections on Recent Developments in the Concept of External Effects," in Welfare Economics: Ten Introductory Essays (New York: Random House, 1969), page 180. 19 The "Second-Best" theory was formulated by R. G. Lipsey and K. Lancaster in "The General Theory of Second Best," Review of Economic Studies (Volume XXIV, 1957). This article set off an intense controversy regarding the validity and importance of the marginal cost principle as a relevant and cogent policy proposal for use in an economy pervaded by imperfect competition, monopoly, externalities and government intervention.

PAGE 36

-2620 of the second-best. Mishan proposes welfare will increase if some prices are equated with marginal costs even if other prices (in the constrained sectors) cannot be so equated. He suggests this policy will result in a greater increase in welfare the smaller the constrained sectors relative to the remaining ones, and the larger the initial discrepancies in the price/marginal cost ratios of the free sectors as compared with the constrained sectors. Since it is unlikely that at any instant the real world will ever attain a Pareto optimum, "(w)e can only hope to be moving in that direction most of the time and not to be too far 21 away from an optimum for any prolonged period." The marginal cost pricing principle is a guide that, while far from perfect, aims the economy in this desirable direction. Marginal Costs and the Energy Industry Marginal cost is the change in costs when one additional unit of output is produced; alternatively, it is the cost that would not be incurred if the marginal unit is not produced. Since marginal cost is completely independent of fixed or sunk costs, the only cost relevant in deciding how much to produce in an existing plant is the variable cost of operating that plant. The longer the time perspective of the costing process, the greater the proportion of costs that become variable. If a 20 E.J. Mishan, "Second Thoughts on Second Best," in Welfare Economics , page 141. 21 Ibid., page 156.

PAGE 37

-27long enough time perspective is observed, all costs are variable and thus all costs are reflected in the marginal. The short run marginal cost reflects the social opportunity cost of providing the additional unit that buyers are, at any given time, trying to decide whether to buy or not to buy. The application of the marginal cost pricing principle requires that price be related 22 to short run marginal cost. Pricing at short run marginal cost does not necessarily mean price will be less than average cost. If the average total cost curve rises (or is constant) at any point along the output scale, it is because the marginal cost exceeds (or is at least equal to) the average total cost. If price is equated to marginal cost, which is greater than (equal to) average total cost, price must be remunerative to the firm. Most resources are fixed and most costs are not represented in the short run marginal cost of electric and gas utilities. Significant differences are thus predicted between the levels of short run and the long run marginal costs up to the capacity of the system. At capacity, infinite changes in costs would not enable the utility, in the short run, to increase output; at capacity, short run marginal cost is predicted to become vertical. Figure 1 represents a graphical illustration of the predicted relationship between short run and long run marginal cost. SRMC is the short run marginal cost and it is vertical at capacity (output OA) . LRMC is the long run marginal cost indicating constant returns to scale over the relevant output range. In this paper, constant 22 A. E. Kahn, The Economics of Regulation , pages 70-73.

PAGE 38

-28-

PAGE 39

-29returns is assumed as representative of the costs of the energy industry. Pricing in the Energy Industry Figure 2A indicates an energy system with capacity of OQ^. If price is equated with SRMC (at OP3) , price would exceed LRAC (=LRMC) indicating economic profits. The profits indicate that a larger capacity is necessary. In Figure 2B the system capacity is OQ2 . If price is equated with SRMC (at OP^ , it would be less than LRAC and thus force economic losses. The losses indicate that a smaller capacity is necessary. In Figure 2C, the system capacity of OC^. If price is equated with SRMC (at 0P 2 >, it is also equated with LRMC and LRAC. Resources are allocated in an optimal manner. If the same capacity serves more than one class of customer with more than one time period considered, capacity costs should be charged only to those users that use the entire capacity. For example, Figure 3A presents a vertical summation 2 Aof peak and off-peak demands. D 1 represents the peak use and Constant returns does not imply that two firms could supply the same output as one firm with the same level of average costs, for the duplication of resources that are required in the distribution of the energy would greatly increase the per-customer cost. -Constant returns implies that one firm can increase its system capacity and have the same per-unit costs. For a discussion of this assumption please see A. E. Kahn, pages 152-165. 24 For a discussion of the vertical summation technique see: P. 0. Steiner, "Peak Loads and Efficient Pricing," Quarterly Journal of Economics (November, 1953, pages 585-610; 0. E. Williamson, "Peak-Load Pricing and Optimal Capacity Under Indivisibility Constraint," American Economic Review (September, 1966), pages 810-827; and A. E. Kahn, The Economics of Regulation , pages 89-103.

PAGE 40

-30Pricing Under Capacity Constraints: The Case of a Single Peak Consumer

PAGE 41

-31Pricing Under Capacity Constraints: The Case of Both Peak and Off-Peak Consumers

PAGE 42

-32D 2 represents the off-peak use. Existing capacity is 0Q 2 . Optimum pricing requires the peak user to pay price OP^ (=SRMC) which will ration the peak quantity to that level of capacity (0Q 2 ) . The offpeak user will pay price (=SRMC) and excess capacity (Q 2 Q 1 ) exists during the off-peak period. Only the peak user pays any part of the fixed (capacity) costs. Since OP^ exceeds LRMC , the system capacity is not optimal. In the long run, a capacity of 0Q 3 should be built. The peak price should then be 0P 2 (=SRMC=LRMC) , the peak quantity OQ^ , and the off-peak price OP^ (=SRMC) , the offpeak quantity OQ^. The marginal cost pricing principle thus forces the optimal scale of plant and allocation of resources. In Figure 3B there is no excess capacity in the off-peak period. Existing capacity is 00^. Peak price is OP 5 ; off-peak price is OP^. Since OP 3 + OP^ is greater than LRAC, system capacity is less than optimal. If the capacity is expanded to OQ^ (where Dq is equated with LRMC), peak price should be OP^, off-peak price should be OP^ and both would use the entire capacity (0Q 2 ). Each class contributes to capacity costs, the contribution from the peak user (P^ SRMC) is greater than the contribution from the off-peak user (P 2 SRMC). Summary Policies are based on generalizations which do not always hold. Frequently, these generalizations must be quite broad. Since we cannot develop perfect policy rules, we must do the best we can and deal with exceptions in an ad hoc manner. Rules must be tested against the available alternatives and not rejected when they do not conform to the purely theoretical criterion. A

PAGE 43

-33set of simple, specific policy rules is essential and must be selected from the best of those alternatives that are available. The marginal cost rule appears to be superior to its only apparent rival, the average cost rule. The marginal cost rule can be stated: Prices should be set at short run marginal costs when possible, and when not possible, marginal costs should be recognized as both a target toward which prices should be 25 directed and a floor above which prices should always exceed. The optimal system capacity equilibrates the combined demands (D c ) with the long run marginal cost (LRMC) . Each class of customer must pay a price at least equal to the short run marginal cost (SRMC) , plus a contribution (in proportion to its usage at the peak period) to capacity costs. Under conditions of constant costs (LRMC=LRAC) , the equilibrating of price and SRMC results in a profit or a loss only if the system capacity is other than optimal (resources are not allocated in an optimal manner). The energy industry presently utilizes peak-responsibility 26 pricing to some degree. Electric and gas utilities employ a twopart tariff for sales to wholesale and to industrial customers in which an "energy" charge (SRMC) is added to a "demand" or "capacity" charge. The proper measure of the demand charge is the proportionate share of the peak demand, placed by each customer, on the system peak. The different rates charged to different classes of customers 25 For an excellent summary of the application of the marginal cost rule see M. J. Farrell, "In Defense of Public Utility Price Theory," in Public Enterprise , R. Turvey, ed. (Baltimore: Penguin Books, 1968), page 47. 26 A. E. Kahn, The Economics of Regulation , pages 95-100.

PAGE 44

-34is not discriminatory if the difference in rates reflects a difference in cost. Discrimination consists of price differences not corresponding to cost differences. The off-peak user should not be charged any capacity cost if capacity would be there whether or not the off-peak user made demands on it. Average prices are compared to both short run and to long run marginal costs in Chapters IV and V. In those comparisons, prices are shown as a percentage of marginal costs (P = X7» MC) . The percentage mark-up figure for the industrial customers is subtracted from the percentage mark-up figure for the residential customers. The difference is a rough estimate of discrimination.

PAGE 45

CHAPTER III EVALUATION OF EARLIER WORK The competitiveness of single service and combination utilities has been studied by several authors. In this chapter these studies are reviewed and evaluated. The policies suggested by these studies are in conflict. Indeed, it is this conflict that establishes the need for the work pursued in this thesis. The Study by Bruce M. Owen Bruce M. Owen was the first to analyze the economic impact of the common ownership of the sources of energy.'' This research is a valuable contribution for it stimulated interest and further writings. Owen bases his study on theoretical reasoning that suggests that combination utilities should have greater economic power than single service companies. If regulatory commissions have successfully prevented combination firms from exploiting this market power, there should be no significant difference between prices and outputs due to the market organization. Thus, the effectiveness of regulation may be measured by observing the relationship of prices and outputs between the types of utilities. '"B. M. Owen, "Monopoly Pricing in Combined Gas and Electric Utilities," The Antitrust Bulletin , Volume XV (Winter, 1970), pages 713-726. -35-

PAGE 46

-36Owen classifies any utility that has both electric and gas revenues as a combination firm and includes only utilities with revenues of $1 mill-ion or more in 1967. The data are derived from the Federal Power Commission's Statistics of Privately Cwned Electric Utilities and Brown's Directory of North American Gas Companies . Owen defines price as average revenue per kilowatthour or m-therm sold to final customers. All figures are studied on a national basis. An econometric model employing simultaneous equations is utilized. Owen finds that prices are significantly higher and output significantly lower for combination gas/electric utilities than for single service electric utilities. He concludes, therefore, that the results establish the ineffectiveness of regulation. More vigorous enforcement of antitrust policy is proposed. The results of the study can be questioned on several grounds. First, the lumping together of all utilities that have both gas and electric revenues with no regard to the relative significance of the components creates a group of firms that are not comparable. For example, Southern California Edison Company had electric revenues of over $720 million and gas revenues of less than $150,000 in 1970. Pacific Gas and Electric Company had electric revenues of over $704 million and gas revenues of over $474 3 million in 1970. These two companies are both classified as 2~ Ibid., page 722. 3 Federal Power Commission, Statistics of Privately Owned Electric Companies in the United States , Washington: U. S. Government Printing Office (1970), page 103.

PAGE 47

-37combination utilities and not distinguished in any manner even though one has gas revenues that comprise only .02 percent of its electric revenues and the other has gas revenues that comprise 67 percent of its electric revenues. Such inconsistencies within the groups of firms examined could lead to erroneous conclusions. A second criticism involves geographical differences and their effects upon energy consumption. Given that the marginal rates decline as use by a given customer increases, average revenue may not be an appropriate measure of price. Instead, the average revenue received by a particular utility from any given customer class will be determined jointly by the actual design of the rate structure as well as the quantity consumption per customer. If there are significant differences in quantity of consumption per customer among geographical areas of the country, and if certain types of utilities are found concentrated in specific geographical localities, the results of a study based on aggregated national data must be questioned. The Study by Patrick C. Mann Patrick C. Mann compares the economic performance of single service electric utilities with the electric operations of 4 combination gas/electric utilities. Only firms with revenues greater than $20 million in 1967 are considered. Mann uses data from the Federal Power Commission's Statistics of Privately Owned Electric Utilities . The method utilized is regression analysis. 4~ P. C. Mann, "The Impact of Competition in the Supply of Electricity," Quarterly Review of Economics and Business , Volume 10, Number 4 (Winter, 1970), pages 37-49.

PAGE 48

-38Combination utilities are defined as firms with revenues of any magnitude from both gas and electric sales. Mann recognizes the problem this classification creates and includes a variable in his analysis (the ratio of gas to electric revenues) to take the problem into account. He correlates this gas/electric revenue ratio with the level of average electric revenue per residential kilowatt-hour sold. Other variables are the rate of return earned on net electric plant, distribution expense per kilowatt-hour, distribution expense per customer, administrative-general expense per kilowatt-hour, sales expense per customer, cost per kilowatthour of steam generated electricity, and cost per kilowatt-hour consumption per residential customer. Mann finds no significant relationship between the ratio of gas to electric revenues and the price of residential electricity. Mann concludes that joint operations of gas and electric operations do not affect residential markets. " (R)esidential price differentials between combination and straight electric firms are caused primarily by factors independent of the dual service nature of the combination utility.""' Additionally, Mann concludes that there is evidence that the higher commercial and industrial prices associated with combination firms are "partly a function of competition between gas and electricity suppliers."^ Mann does make a significant improvement to Owen's statistical methodology by including the ratio of gas to electric ~*Ibid., page 48. 6 Ibid, page 49.

PAGE 49

-39revenues. However, his conclusions are questionable for at least two reasons. First, Mann attempts to test the differences in efficiencies, as measured by costs, between the types of utilities. In a regulated monopoly with zero economic profits, total revenue equals expenses (production, transmission, distribution and customer accounts) plus taxes and the rate of return. Mann analyzes variations in average revenues after accounting for all cost differences except differences in hydroelectric production expenses, transmission expenses and taxes. These factors are largely determined by the geographical location of the individual firm and not by the ownership of the utility. A firm will utilize hydroelectric facilities only if there exists a sufficient movement and fall of water in the service area. Transmission expenses are largely a function of the density of the population. Taxes are related to the areas the plant and equipment are located. Mann has included most of the potential sources of cost variations in his dependent variables. The remaining sources of cost variation are a function of geographical location. It is not surprising that he finds few differences in the level of costs. A second deficiency in Mann's study results from the lumping together of commercial and industrial customers. According to Federal Power Commission statistics for all Class A and B investor-owned utilities, the average revenue received from all commercial customers was 2.0812 cents per kilowatt-hour in 1970; the average revenue received from all industrial customers

PAGE 50

-40was 1.0175 cents per kilowatt-hour ? The much lower rate prevailing for industrial customers reflects the fact that such customers tend to take more power at higher load factors and higher voltages than commercial customers. In addition, demand elasticities are thought to be much lower in the case of commercial customers than in the case of industrial customers. When commercial and industrial customers are grouped, the average revenue of the combined commercial and industrial sales is affected by the number (and the usage) of commercial customers and the number (and the usage) of industrial customers. A utility with a high number of commercial customers (relative to the number of industrial customers) is expected to have a higher average revenue than a company with a low number of commercial customers (relative to the number of industrial customers). The ratio of commercial to industrial customers and the level of use of each is highly correlated to geographical factors. This geographical distribution is not investigated, by Mann. The Study by William H. Collins g Further work has been pursued by William H. Collins. He sought to identify and empirically test the social desirability of combination gas/electric utilities. The analysis is based on 1967 data from the Federal Power Commission's Statistics of Privately ^Federal Power Commission, Statistics of Privately Owned Electric Utilities in the United States (Washington: U. S. Government Printing Office, 1970), Table 4, page XVII. g W. H. Collins, Jr., "Combination Gas-Electric Utilities," Ph.D. dissertation, Southern Illinois University (197Q. See also W. H. Collins 1 testimony in Combination Utility Companies (Washington" U. S. Government Printing Office, (69-612), 1971), pages 441-448.

PAGE 51

"41" Owned Electric Utilities in the United States . The sample includes 52 combination utilities, 89 straight electric utilities and 62 straight gas utilities. Collins employs "z" tests, Student "t" tests, Mann-Whitney U tests, and Wilcoxon Matched Pairs Signed Ranks tests in his computations of the data. Collins finds that there is no significant difference between single service gas utilities and the gas operations of combination utilities. On the other hand, the results of the electric computations indicate that the performance of the electric operations of combination utilities is significantly below that of single service electric utilities. This conclusion is based primarily on the fact that when the performance of the single service electric firms is compared with the performance of the combination firms, the combination group has significantly higher levels of average revenue per kilowatt-hour for each customer class and significantly lower levels of kilowatt-hour consumption per residential customer. In addition, Collins finds that average salaries and wages, average distribution expenses, average operating revenue (less operation and maintenance expense) and average taxes are significantly higher on a per kilowatt-hour basis for combination utilities than for single service utilities. In contrast, sales and advertising expenses are found to be higher for single service utilities . Collins' study is a valuable contribution to the development of the literature concerning the competitiveness of public utilities in that he establishes a statistical methodology that

PAGE 52

-42identifies differences between variables indicating differences in consumption and efficiencies. Collins found that combination companies charge higher prices and have lower quantity sales than single service electric firms. Economic theory predicts that the greater the monopoly power, the greater the ability to raise prices and (thus) restrict sales. The conclusions formulated by Collins appear to support the hypothesis that a more efficient allocation and utilization of resources results when competition between the sources of electricity and gas is increased. As a policy matter, the Collins study thus supports the idea that antitrust, if applied to combination gas/electric firms, will benefit the consumer. The findings of the Collins study can be questioned principally because of too strong an assumption concerning the homogeneity of the groups of utilities. Collins states that he statistically tested the groups of data and found no significant difference with regard to such factors as utility size, relative importance of different types of customers, type of generation, percent purchased power, degree of urbanization, population density, geographic location and distance from fuel sources. However, his methods are questionable. Although he states the groups are not significantly different, he does not describe how he tested the homogenety in geographic location. If the combination utilities are concentrated in geographical localities that are characterized by higher fuel costs, wage rates, state and local tax burdens, generating unit costs, and other such cost items, a priori reasoning suggests that such firms will charge higher prices

PAGE 53

-43(and thus sell smaller outputs) regardless of the market organization. Conversely, if single service electric utilities are concentrated in geographical localities where the climate is warm, the high use of air conditioning will increase the monthly consumption per customer. This increase in consumption reduces average revenue since marginal rates are inversely related to consumption. The lower average revenues (average prices) may thus be due to the higher consumption; the higher consumption may not be due to lower average revenues. Collins uses a logical methodological procedure. However, he leaves his conclusions open to criticism by not analyzing the geographical location of the utilities and by neglecting the impact that a highly uneven distribution of the types of firms might have on the statistical findings. The Study by the N.E.R.A. The National Economic Research Associates, Inc. (NERA), analyzed various financial and operating data pertaining to electric and gas distribution companies in a report prepared for the Long Island Lighting Company. Data were gathered from a variety of sources including: Moody's Public Utilities Manual . Standard and Poor's Compustat tape, the Federal Power Commission's Statistics of Privat ely Owned Electric Utilities , and Brown's Directory of North Am erican Gas Companies . Single service utilities are defined 5~ . , National Economic Research Associates, Inc., "Combination Companies: A Comparative Study," as printed in Combination Utili ty Companies , pages 275-363. ~

PAGE 54

-44as any company with more than 90 percent of its revenue derived from the sale of one source of energy. The methodology utilized by the NERA involves the computation of the arithmetic means of the variables studied. These means are compared as indicators of the relative similarities and differences of the utilities. For example, the capital structure of the firms was studied. As an example of the methodology, it was stated that equity comprised 37.1 percent of the straight electric utilities' total capital, 36.7 percent of the combination utilities' total capital, and 42.2 percent of the straight gas utilities' total capital. From these measurements the conclusion was reached that "... straight gas companies have relatively more equity in their capital structure . . . than either straight electric or combination companies. The 'financial" data section examines the growth rate in earnings per share, the rate of return on invested capital and the rate of return on common equity (as defined by Moody' s Public Utilities Manual ). The study concludes that differences in the financial measures (stated above) between the groups of utilities are very small. Specifically the study states ". . . While there is a close similarity in the averages for the straight electric and combination companies . . . the straight electric would appear to have a slight overall edge over the combination companies. Similarly, except for return on invested capital, the combination companies have out-performed the straight gas 10~ Ibid. , page 282

PAGE 55

-45group."^' 1 " The observation is made that straight gas companies have relatively more equity in their capital structure (as stated above) , than either the straight electric or the combination companies. In the "expense and plant" data section the study investigates (1) customer accounts, (2) administrative, general, and sales expenses, (3) operating and maintenance expenses and (4) gross plant. Combination companies are found to have lower average customer expense than the straight firms although when these expenses are related to quantity sales, the relationship was reversed. Similar findings are observed with respect to the administrative, general and sales expense figures. The single service utilities have higher expense figures per customer although no difference is observed on a per-unit comparison. Operation and maintenance expense per customer is higher for both single service electric and single service gas utilities than for combination utilities. However, operation and maintenance expense per unit output is lower for single service electric utilities than combination firms, but higher for single service gas utilities than combination firms. With respect to the "consumption and customer" variables, the study finds that single service electrics have higher growth rates and percentage increases in customers than the electric side of combination companies. The gas side of combination companies has higher growth rates and percentage increase in customers than the single service gas companies. The "miscellaneous" data investigated include employees per dollar of revenue created, load factors, BTU heat rate, average iL Ibid. , page 298.

PAGE 56

-46revenue from residential electric sales and average revenue from residential gas sales. The study found that all three groups have similar employees per dollar revenue and that load factors are not significantly different between the firms. The average revenue from residential electric sales is lower for the single service electric than for the combination companies, while the average revenue from residential gas sales is lower for the combination utilities than for the straight companies. As an overall conclusion regarding the study, Jules Joskow, Vice President, National Economic Research Associates, 12 Inc. , stated : The results of our study show that while for certain variables one group of companies may have, on average, outperformed one or both of the other groups, the data, when viewed on an overall basis, do not suggest that any one group has a statistical advantage over either of the other two. Thus, we further conclude that the data do not indicate that a conclusion can be reached either in favor of, or against, combination companies based on their performance as herin measured. The findings of fact of the above study d o not differ significantly from the studies of Owen, Mann and Collins. The policy implications do differ from the earlier studies. Although similar data are studied, and similar findings discovered, the authors of the NERA study state that a conclusion could not be reached regarding overall performance. A question of data interpretation is thus raised. Ibid., page 275.

PAGE 57

-47The NERA study analyzes 47 straight electric, 40 combination and 13 straight gas utilities. The findings are based on a simple comparison of the arithmetic means of each group of firms and no generally accepted statistical technique is utilized to measure the significance of the findings. The analysis of only a portion of the firms with no explanation as to the basis for inclusion or exclusion of firms and the inadequate statistical method raises questions concerning the validity of the analysis. Differences between arithmetic means may be caused by actual differences in the variables compared; alternatively, differences between arithmetic means may be caused by sampling. This study does not have a statistically sound base and is theoretically weak. Thus, the findings are, in and of themselves, not analytically important. The data and the statistical findings of the NERA analysis are similar to those of previous studies. The fact that the conclusions are different is important. Summary of Owen, Mann, Collins and the NERA Differing samples of utilities, various definitions of the degree of competition, and a variety of cost-influencing variables have been used in the studies reviewed above. It is thus not surprising that somewhat different conclusions have been drawn by the various authors. However, despite the differences and the conflicts, several points stand out. The combined control of gas and electric operations is associated with lower levels of kilowatthour consumption and higher average revenues per kilowatt-hour. The studies also indicate that the costs of electrical operation are

PAGE 58

-48higher for combination gas/electric utilities than for single service utilities. No differences between combination and single service gas utility performance are apparent. The policy recommendations each author proposes are likewise similar although each has differences in his application. Since each of the studies found higher prices and lower outputs associated with combination gas/electric utilities, each concluded the separation of ownership of the sources of energy would increase the overall welfare of the consumers. The one exception is the NERA study which states that no conclusion can be reached . In conducting these studies, the authors have consistently failed to consider the geographical distribution of the utilities and the resulting variation in the demographic factors that would thus appear. Dissimilar demographic characteristics are expected to influence the level of each utility's rates and thus affect the quantities consumed by customers in the service area. Differences between combination gas/electric utilities and single service electric utilities that are isolated by the authors reviewed above may thus be caused by the geographical (and therefore demographical) differences between the groups of firms compared. The Study by Joe D. Pace Joe D. Pace studied the differences between types of utilities that have been isolated by the earlier studies. 13 13 iJ J. D. Pace, Senior Economist, National Economic Research Associates, Inc. The study referred to in this thesis is found in Combination Utility Companies , pages 518-553.

PAGE 59

-49His analysis is based on the fact that less electricity is consumed, in areas served by combination companies. He states that previous authors failed to probe deeply enough into possible relationships between this lower consumption and various demographic factors which may differentially characterize the service areas of single service electric utilities as opposed to those of combination companies. He utilizes multiple regression analysis to conduct an exhaustive statistical study of national data. Pace limits his study to urban areas in the United States that contained 25,000 or more housing units in 1970 and were served by investorowned utilities. Because of a "lack of gas industry data comparable in detail and uniformity to those of the electric industry"^ the study includes only the electric component. Data were obtained from the Annual Reports of Class A and B electric utilities to the Federal Power Commission. The study includes 194 urban areas of which straight electric utilities serve 108 (or 55.7 percent) and combination companies serve 86 (or 44.3 percent). Pace establishes the fact that various types of utilities are not evenly distributed throughout the United States. Combination utilities are found concentrated in urban areas which are larger, more densely populated and characterized by higher fuel costs and higher state and local taxes. Larger numbers of apartments exist in combination areas. On the other hand, because average temperatures are higher in the areas served by single service electric utilities, the potential demand for air conditioning is greater in areas served i4~_ , Ibid. , page 538.

PAGE 60

.50by single service electric utilities. Pace concludes that it is not surprising to find that average residential electric revenues are lower in areas -served by combination companies since consumption is greater and step-rates are charged. "Whether the geographic and demographic conditions explain all or only a small part of the observed usage and rate differences. . ."is the question that Pace studies. ^ The variables that Pace studies are divided into six areas. Measures of the "housing" variables are (a) the number of apartment units as a percent of total housing units and (b) the number of seasonal and migratory units per 1,000 housing units. "Urbanization" is measured by the number of housing units in each community, population density and the ratio of residential electric customers in each urban area to the number of housing units in each community. To study "income" variables, Pace segments the percent of all housing units without all sound plumbing facilities, the percent of total housing units occupied by more than 1.51 persons per room, apartment units renting for less than $100.00 per month as a percent of total housing units, per capita income, effective buying income estimates (per household) and the percent of households with cash incomes of less than $3,000 per year. To investigate "weather" conditions Pace introduces as independent variables both the number of cooling degree days for June, July and August, and an income-adjusted cooling degree day measure. 15"" Ibid. , page 519.

PAGE 61

-51"Locational" factors include geographic indicators based on the section of the country the utility is located. The final subdivision relates to"cost-influencing" factors. Included as variables are state and local taxes per dollar revenue, fuel costs, hydroelectric generation as a proportion of total generation, purchased power as a proportion of total sales, fuel used in generation, system size, density and statewide net electric operating revenue as a percent of net electric plant. Pace found no significant difference between the average revenues of combination utilities and the average revenues of single service firms when the demographic variables are included in the analysis. "In plain language, this means that while demographic, geographic and underlying cost conditions very strongly affect the average residential electric revenue required, under current circumatances , it matters little, on average, whether a combination company or a single service electric utility renders service in a given community." 1 ^ Pace thus concludes that there is ". . .clear and substantial support to the finding that combination companies have no adverse effect, as a general matter, on residential electric rates. "^ The conclusions and the policy implications of the Pace study thus contradict those of previous authors and represent a significant contribution to the literature. As with the other studies, however, questions can be raised. nr Ibid. , page 533. 1 Ibid. , page 537.

PAGE 62

-52First, Pace examines only the electric operations of utilities. Further, Pace limits his study by examining only operations in large, urban areas. This restriction in the scope of the study may obscure significant differences that do exist. Additionally, the large number of items that Pace includes in his regressions introduce questions as to the interdependence between variables. Do the variables Pace examines explain the variations or are there more general, or basic, explanations? Summary of the Earlier Work Previous studies have resulted in conflicting policy recommendations. Owen, Mann and Collins found that combination utilities charge higher prices, sell smaller outputs and operate less efficiently than single service utilities. Their conclusions are based on national studies that largely ignore demographic differences between geographic areas. Pace concludes that there is no significant difference in the performance of utilities based on the type of ownership. High prices are determined by demographi characteristics which restrict consumption. It is this conflict that establishes the need for the study contained in Chapters IV and V.

PAGE 63

CHAPTER IV AGGREGATIVE ANALYSIS The objective of this thesis is to test the hypothesis that separate ownership of gas and electric supplies will promote competition and that this separate ownership will result in improved efficiency and lower rates. Data are gathered from reports submitted directly by each utility. These data are more detailed and comprehensive than those available in existing publications. The data are subjected to statistical tests that are widely known and commonly accepted as a basis for evaluating differences in variables. The data are analyzed on a national, or aggregative, basis in this chapter. The Data Earlier studies obtained data from Federal Power Commission publications Statistics of Privately Owned Electric Utilities and Statistics of Privately Owned Gas Utilities . 1 These publications provide basic information for each combination and single service utility operating in the United States. The data this thesis analyzes are gathered from publications entitled "Uniform Statistical Report" which are prepared by each utility and are filed with the Federal Power Commission, All Class A and B single service electric, single service gas, and combination Washington: U. S. Government Printing Office. -53-

PAGE 64

-54gas/electric utilities were sent letters requesting a copy of this 2 report. The results of the survey of the utilities are summarized in Table 4.1. The differences between the number of requests and the number in the sample reflect several factors. First, some of the firms are generation companies only and thus have no retail customers. In order to isolate differences associated with competition between utilities, retail sales must be examined; these generation utilities are therefore eliminated from the study. Second, many of the companies are part of holding companies and filed consolidated statements. As far as possible, each company is treated as an individual and independent entity, but where it appears that a holding company exerts policy influence at the subsidiary level, they are grouped together and treated as one company. Thus, individual utilities in the "Southern" holding system are studied independently while the "New England Electric System" and the "New England Gas and Electric Association" are studied as single f irms . For our purposes, a "straight" company is defined as one that derives not less than 95 percent of its revenue from the sale of one type of energy. The Federal Power Commission defines a combination utility as a firm with any revenue from the sale of both gas and electricity. As a third adjustment to the statistical reports, each firm's revenues are examined and combination firms 2 The requests are based on a list of all the utilities in the Federal Power Commission publications mentioned above.

PAGE 65

-55CO W to B3 Pk 3 < b H W ft, *4 O 1 CO m o to C7l 0) •s u OJ u 4-1 o to 0) o 3 cn 01 ft in (J O a r-l H DO H «3 !-i AJ CO M u o 0) 1-) w n 03 O C o CO c CO CO O CO •H M 4J CO

PAGE 66

-56are reclassified as single service firms if one type of energy constitutes less than 5 percent of the total dollar sales. A final factor involving the difference between those firms from which information was requested and those firms in the final sample involves the comprehensiveness of the data received. All firms must fill out the report and make it available to the public when it is requested; however, there are instances where portions of the data are not complete. Only those firms that supplied complete information are included in the final analysis. Because of the adjustments outlined above, the utilization of information from the individual firm's reports in place of information from the Federal Power Commission represents a trade-off: the data are more detailed but they are not as complete. In the opinion of this writer, the additional detail is worth the loss in completeness, for this detail allows comparisons that are impossible with published data. Methodology A major intent of this thesis is to reconsile the contradictory conclusions of other studies; the methodology thus closely parallels that established in these studies. A cross-sectional approach (which freezes time and searches the data for indications of differences) is utilized as it is in all studies reviewed in Chapter III. Detailed information is examined to ascertain the parameters that differ significantly due to the effects of competition. The study involves two similar but separate aspects of production: that of electricity and that of gas. Within each aspect of production,

PAGE 67

-57there are two types of utilities that are to be investigated: the electric figures of combination firms must be compared with the corresponding figures of straight electric companies and the gas figures of combination firms must be compared with the corresponding figures of straight gas companies. Due to the lack of a common measuring device, few relevant comparisions can be made between electric and gas figures. The test thus involves the question: Do statistics observed in two samples differ? This test must be posed independently with respect both to electric and to gas figures. To accomplish the statistical test, means and standard deviations are computed. The question then asked is: Can differences between two sample means be attributed to chance errors or to differences in the populations from which the samples were drawn? A statistical approach commonly used to compare means involves the computation of a "z" statistic (for large samples) or a 3 "Student's t" statistic (for small samples). The null hypothesis in both cases asserts that the means of the population parameters being tested are equal. This assumes that if a large number of samples were taken from one population (or two identical populations) , the differences between the sample means would give a normal distribution centered around zero. If samples were taken from different populations, the differences between the sample means would not average zero but would, instead, equal the differences in the means of the population parameters. For example, the J" See for example M. Hamburg, Statistical Analysis for Decision Making (New York: Harcourt, Brace and World, 1972), pages 330-348; Ya-lun Chou, Statistical Analysis (New York: 'Holt, Rinehart and Winston, Inc., 1969), pages 387-395; Smith and Williams, Statisti cal Analysis for Business: A Conceptual Approach (Belmont, California: Wadsworth Publishing Company, Inc., 1971), pages 393-429.

PAGE 68

-58assumption is made that there is no difference between the total population served by the straight electric companies and the electric operations of the combination utilities. The average population size of the sample of straight electric companies is then compared to the average population size of the sample of the combination utilities. The test is whether the divergence indicates differences in the populations served by the firms. Specifically, a statistic showing the difference between the sample means (x-^ is established and if this statistic differs significantly from zero, it is concluded that the population parameters are indeed different. To arrive at the significance factor, the "standard error 4 of the difference between two means" is estimated as: i = V s i /n i + s 2 /n : (x. X ) s 1 2 The test is two-tailed because the hypothesis of equal populations would be rejected if (x^ x£) differed significantly (either above or below) from zero. For large samples the test can be viewed as a comparison of the specific difference of sample means to a frequency distribution obtained by drawing repeated pairs of samples and observing the differences in these samples. If the distribution is normal, 69 percent of all observations should lie within plus-or-minus one standard deviation, 95 percent within plus-or-minus two standard deviations and 99 percent within plus-or-minus three standard deviations. ^M. Hamburg, Statistical Analysis for Decision Making , page 332.

PAGE 69

-59It is then assumed that if x-^ X2 is greater than three standard errors, the population means differ at the 99 percent confidence level. The computations of these comparisons is simplified by computing the "z" ratio which is defined as: ~* X l " X 2 (x x x 2 ) The "z" statistic measures the number of standard errors between the means. For example, a "z" of 3 indicates that x 2 (the difference between the means) is three times the value of the standard erros (s_ _ ). By selecting a level of confidence, a X l " X 2 a value of"z"is determined which, if exceeded, indicates that the means differed significantly from each other. For small samples, the "Student ' s t"is utilized. The theory of this test is similar to that of the"z"test. The estimate of the standard error is:^ (n x l)-(sf) + (n 2 l)-(sf) (X 1 " X 2 } n x + n 2 and the Vstatistic is defined as: 7 x l " x 2 s (x\ " x 2 ) 5 Ibid. , page 333. 6 Ibid. , page 347. 7 Ibid. , page 347.

PAGE 70

-60The critical values of "t" are based on the size of the sample. Tables are included in the statistical appendices presenting the values of both •"z" and of "t" for various levels of confidence. Table Notations In order to simplify the presentation of the many computations upon which the variables are to be compared, confidence levels of 95 percent and 99 percent are selected as critical. The data are analyzed and the corresponding "z" or "t" statistic is computed and presented in the statistical appendices. Those variables that are significantly different are presented in the body of Chapters IV and V. For example, the average "Total Gas Population Served" by the straight gas utilities is compared to the average "Total Gas Population Served" by the gas operations of the combination gas/electric utilities. The values are 1,901,721 for the straight gas companies and 984,968 for the combination gas/ electric utilities. The "z" statistic computed for this variable is -2.26 which indicates that the straight gas companies serve a significantly larger population (at the 95 percent confidence level) than the gas operations of the combination firms. The negative sign indicates that the straight utility figure is larger than the combination utility figure. The average population figures and the "z" statistics are found in Appendix C of Chapter IV. In Table 4.2, a "G" under the gas column of "Total Population Served" indicates that the straight gas company is larger (at the 95 percent confidence level) than the gas operations of the combination utilities. If the "z" is less than -2.8, the degree of confidence would be 99 per-

PAGE 71

-61cent and the character in Table 4.2 would be "G*". A dash ("-") indicates that there is no significant difference between the parameters examined. Aggregate Characteristics of the Firms Variables are selected that measure aggregate characteristics of each utility. These figures are presented in Tables 4.2 through 4.4. Table 4.2 presents demographic and sales characteristics. No significant differences are observed in the electric computations; thus, it is concluded that the electric operations of the combination utilities and the straight electric utilities are of similar size. However, differences are observed in the gas computations. Single service gas companies serve a larger population (although not with a higher density), sell larger quantities to industrial customers and receive larger revenues from industrial customers. Single service gas companies are larger than the gas operations of combination gas/electric utilities. Table 4.3 compares cost figures, plant size figures, interest charges, taxes, and efficiency and fuel cost measures (for electric firms). Few differences are observed in the electric computations. The differences that are observed indicate that straight electric companies spend more on sales efforts and have greater dollar investments in general plant than combination gas/ electric companies. Significant differences are evident with regard to the gas expense figures. The indication again is that the straight gas

PAGE 72

-620) r-l t • • co 0 to CO V o 1-4 T-l CO u oo CO •r-l r4 t (1) cC J-) o o CO T-l J~i (0 CU ra 60 cu O •U e CO M O a> u 60 CO 3 o 4J 03 CO CO 4-> & 5-) o 0) 1-4 CO e o •r-4 J-l to 3 § T3 0) C 0 T-4 CO r-4 3 a o Pm o H -a 0) 0) 00 CO 01 r-4 •r-l CU r-4 CO 3 CT CO CN 01 to 3 cr CO c 0 •r-4 J-> a r-i 3 Cu c I I I CO 3 u co CO co 0) u 5-4 cu co CO § 0) CU 3 0) 0) S E 3 3 3 CO o o i> 3 CJ CU co co CO 4J JJ a > ai cu r-4 Ol 0) 3 CO CO > > CO r-4 r-4 a 3 3 > Oi w d CO C_> O r-' Crf CO 00 00 r-4 a) r-i CD r-4 a) r-l r-i o CO r-i r-i rH CO r-l r-4 •H CO a 01 H tO CO cO •r-l CO CO U •H •r-4 4-J r-i •-4 w 4J ~i •H c O J-4 3 O M 3 o M CU S-4 iJ CO 01 u 4J J-4 -u 0) CO 01 ~ 9. co 4-J t) 0) CO •H 3 3 •r-4 3 •r-4 H 3 CO -a c D] a U CO a CU 0 e a o> 5 3 3 CU 5 3 erf r-4 Rev a! u H Qua erf o H

PAGE 73

-63CO a o •H 4J CO cfl Cfl 4J O 3 ft I CJ * l o o a i o CO cu i-l CO •H 0) •U 1-4 °-s a» -h 4J CO > <4-l o x as to FH o CO CO CO C CJ O >H i-l H 4J 4J CO O 4J -< cfl ft u CO « .c a CJ CO C a>
PAGE 74

-64companies are larger than the gas operations of the combination gas/electric firms. The higher production plant for single service gas companies reflects the fact that combination firms buy from gas transmission companies whereas single service gas companies tend to be vertically integrated. The higher interest charges for the combination utilities is partially due to the nature of the data: the data do not differentiate between electric or gas operations of combination firms. The aggregate "Interest Charged" thus reflects total interest (for both electric and gas operations) for the combination utility and reflects only the interest charges of the single operations of the straight utility. The significant point is that there are differences in the gas computations and there are not significant differences in the electric computations. This fact indicates that single service electric companies have interest charges as large as the combined operations of gas/ electric companies and, thus, are of somewhat equal size. Single service gas companies are smaller than combination companies and, thus, smaller than single service electric companies. State and local taxes do not differ in the gas computations, but single service gas companies pay higher income taxes relative to the payments of the gas operations of combination companies. These figures will be discussed in detail at a point later in the thesis; however, it should be noted that while straight gas companies have larger expenses than the gas operations of combination firms, they pay more in income taxes but do not pay more in state and local taxes.

PAGE 75

-65The efficiency of a utility's production depends to a large extent upon the state of technology; the organization of the industry is not expected to affect this efficiency. It is thus not surprising that differences in efficiency and fuel costs in the electric computations are not observed. Similar comparisons could not be made for gas operations as the reports do not contain figures of a corresponding nature for gas operations. Table 4.4 indicates that single service gas companies employ more workers and pay larger total wages than do the gas operations of combination firms. These facts are expected since the single service gas companies are larger than the gas operations of the combination firms. The fact that the data are not divided between fringe benefits for gas employees and fringe benefits for electric employees creates the (false) appearance that combination utilities pay larger dollar fringe benefits. This computation is discussed later in the study. The financial figures do not compare equal items since both the "Source of Funds" and the "Application of Funds" for combination firms represent the combined gas/electric operations. As with the interest charges, the interesting point to notice is that there are no differences between the combined gas and electric financial figures and the straight electric figures, yet there are significant (at the 99 percent level) differences in the gas computations. The single service electric companies are about the same size as the electric operations of the combination companies. The gas operations of the combination firms are small in relation to the whole company.

PAGE 76

-66CO O cfl •u 3 (X % O o o u * * * * u o c_> u 00 B CO o 0 1-4 CI) 0) T-l r-< «-l a •U w i E=> CO 0 •H u 0) >4 .c « 4-1 > i-l O 03 i-l CO o O C i-l co 4J C W iH •H En U o> -a 4-1 c U CO CO U 4J flj C •C 4) 0 ! 01 o 4J i-4 co a t>o e n 4? co 4) 0) o 4J c I o E U i-l O P co a. c < o •O r-l •W CO > 4J •r4 O Q H Cfl O -a

PAGE 77

-67In t his section, certain aggregate characteristics of the firms have been analyzed to ensure that the data are homogeneous and that the observations made from the data are correctly interpreted. Few differences are observed between single service electric companies and the electric operations of the combination gas/electric utilities. The gas operations of the combination utilities are small relative to the operations of the single service gas companies. Relevant comparisons may therefore be expected with respect to electric computations, but caution must be maintained when attempting comparisons with gas figures. Differences that at first appear to be caused by competition, may be caused by size differences. Findings of the Aggregative Analysis In the national study, certain variables are studied as measures of "Economies and Diseconomies" and of "Marginal Cost Pricing." Both are discussed in turn. Economies and Diseconomies Competitively organized industries, relative to monopolisticly organized industries, are expected to benefit the consumer by forcing lower prices and more optimal outputs. Table 4.5 illustrates comparisons of average revenues, average consumption and the average 8 yearly dollar bill per customer. The electric operations of combination companies charge higher prices and have smaller average consumption than single service electric companies. These price and 8~ In Table 4.5, average revenue is defined as total dollar revenue divided by quantity sales; average consumption is defined as quantity sales divided by the total number of customers; and the average yearly bill is defined as the total dollar revenue divided by the number of customers.

PAGE 78

-68C o 4-1 CO Cfl 3 co 3 o o 54 4J o CD w 4J 4-1 I o u * * * u u u 1-4 .n «3 H 4) U •r4 U cd 6C id i-i 5 gg T-l ca u S 3 03 3 CD CD 3 o e CD OS O M OS vy i— I -H CO Ct) Cfl 4J i-l i-l CO C CJ CD 54 >>T3 CD 4-1 -H £ 1-4 CO £ 4-1 CD CO H 5-1 4-1 ca 3 . 3 C « U H cfl 3 oCN co C cj CD CD 54 3 13 CD 3 cfl cfl •H t4 cd co E TJ > CD O C HI « O H OS 00 4-1 4-1 CJ 00 a 0 H 4-> ct) 4-> O

PAGE 79

-69consumption differences are not significant in the gas computations. This observation suggests that competition is a significant factor at least with regard to the electric aspects of the industry. Since the "competitive" indicators are not observed in the gas computations, it is concluded that there is a lesser degree of competition in the supply of gas. These findings are anticipated due to the size differential observed in the previous section. In the electric computations, the single service and the combination companies are more nearly equal in size and the effects of competition are thus more obvious. Competition is expected to sharpen the decision-making process and therefore expected to lower the average costs. Table 4.6 illustrates average cost computations for the utilities. Average total costs, average depreciation costs, average distribution costs, and average customer costs are significantly higher for the electric operations of combination utilities than for the straight electric companies. This observation suggests that competition forces single service firms to be more efficient than (monopoly) combination gas/ electric firms. Similar findings are not observed in the gas computations. These findings are not expected since competition is weaker in the supply of gas than in the supply of electricity. Table 4.7 presents figures that compare average wages. No significant differences are observed in the electric computations. Single service electric utilities are about equal in size to the electric operations of the combination utilities and each pays comparable wages. Pension benefits are slightly larger as a per. centage of the total wages for the straight electric companies,

PAGE 80

-70C O to 3 a g o o 3 o o 1-1 1-1 h 4J o 0) 1-1 w as u 3 a. * I CJ o w vO * .-I s H ID •U CO O U CD M d I CO 0) I-l CO CO •u 1-1 •u c 5 3 O" co ai to CO 4J CO >N O 4J c_> ^ 4J ^ c at a u 3 o cr H » CO >-> -u U -H «H 4J u u P i-l 3 to U to 3 C 3 OtO C 3 O" cu co > 4J CO >*» i-t CO 4J 4-1 •H 3 4J CO 4J 3 3 *h maw 3 C C^to 3 \ D a CO 0) to cu 3 CU 3 a) o •a 3 4 3 cu o CO T-1 3 4J CU CO ex X 4J W CO •t-l co 3 CU 'H -" s to -a co < \o roo cti o

PAGE 81

-71cn c O CtJ w 4J ct) 3 e> a 6 o u cn C o o tH -f-l (-4 4J •M 03 CJ 4J a) 3 •-i a w 6 o o * -It u • CD r-l s H n cd M cd 0) to ca n CD o cn C CD co to CD > CTJ CD CD (1) » >> pS O C V to ca > a cn jj If CD •H 0) o w 4J 4) H 01 I>> cn H) u O CD cn O CO c 01 H 60 O ee •H & diu cc5 a lo (3 IT al tl 1 03 CD dm 6' ot O as nu Cd i-i H H cn ve 03 M j-J CD o 4J O H o CO cn H &4 u u CD to •H fi c 3! CH CD •H CD CD C2 •U cn cn cn s a cd CD CD ct3 CD o e M to cq pq 0 CD fd Ot) cn •H a !3 cd C B cn o to o o f5 H iH CD •H «H CD :5 3 cn 00 Ph 03 U u C c ^— ' O o 01 CD o H H Ph Pn H CI

PAGE 82

-72although on a per-worker basis, there is no significant difference between the types of utilities. In the gas computations, combination companies pay higher wages per employee and have higher pension benefits. The gas operation of combination utilities is smaller than the electric operations of these firms; this size difference may cause the wage differential. Table 4.7 further indicates that the average productivity is not significantly different in either the electric or the gas 9 computations. This productivity constitutes a rough measure of efficiency; higher average productivity causes greater outputs and increased revenues. This study indicates that competition does not increase the productivity of the employees. Combination gas/electric utilities use their office buildings and other items of general plant in both gas and electric operations. Table 4.8 presents computations of the use and the importance of plant items. These figures indicate that combination companies have greater revenues per dollar general plant, and incur higher general expenses per dollar general plant. Single service utilities spend more as a percentage of total plant on general plant. These facts support the idea that combination companies are able to use their general plant in a joint manner. The relative importance of this economy is not great however, as the differences are not significant with respect to total plant computations. The average productivity is measured as the revenue generated per employee, or the total revenue divided by the number of workers employed. This productivity figure is presented as a rough estimate of the value of the marginal product.

PAGE 83

-73m s o •H 4J to cd cd c a I o a I u I i o a o •H M •u o a> H W to s o -H 4-1 C3 4-1 I U I W O * * 4-) § <0 4J c 4-1 «J 3 r-4 tO pu. 1-1 Pm 3 o •• C -H 4-> o C -H p CO 4J .O -4 CJ i-t l< 3 h T3 4J -s O 03 u 4-1 PM i-l Pm tfl U r-4 0> to 3 4J OJ O Pm Q O H *«. 3 C C a) Pi e •H 60 4-> C tO i-l J-J 4J O o O « 4-1 r-l O tO H 4J O H HOD 3 3 3 C C 3 a> a) tu > > > -j o a c' cpS 60 60 60 3 3 3 •i-l 1-4 -H 4-1 4J 4J tO tO to M M h a tO 3 .-I tO PM 4J c t3 r-4 PM C O 4-1 PL, 3 U T3 O S-4 tu a u 'Si tO a> to 3 4J aj o c3 H to tc tn to V J) J) J) 0) i— 4 i— 4 t— t i— 4 to tO to to tO , CO JJ 4J 4J 4J •H i-l i-l i-l >, 4J 4J 4J 4J 4J C 3 3 3 H to to to to U 3 3 3 3 3 C O 1 C C CO 3 C 4J 4J 3 3 cO CO CO 4J JJ 3 O CO H i-i Pm 3 4J O 3 i-l CO 4J >-4 CJ 3 T3 ^ O U £ PM P-i PM cd tO tO .. 4J 4-1 4-1 4-1 O O O S H H H cd 4J 3 tO .-4 PM 4J 3 3 O CO i-l PM CO U 4J 0) 3 3 CO 0) r-i O Pm 3 O 0) 4J 3 ••3d) 4J * p. 3 i-l X tfl M W i-l 4-1 PM tO i— I c0 H tu c
PAGE 84

-74Single service electric utilities have higher sales per customer and therefore greater sales per dollar of (fixed) distribu10 tion plant. The greater consumption per customer indicates that the distribution plant is used more efficiently by the single service electric utilities than by the electric operations of the combination firms. Table 4.8 indicates that the straight electric utilities do in fact use their distribution plant more efficiently since the quantity sales per dollar distribution plant is greater for the straight electric companies than for the electric operations of the combination firms. In sum, combination utilities use general plant items more optimally than the single service utilities, yet straight utilities use their distribution plant more efficiently than do the combination firms. The net advantage of this trade-off is small but weighted in favor of the single service firms since straight utilities have greater quantity sales per dollar total plant. Table 4.9 compares the aggregate utilization of funds between the two types of utilities. The data from the combination firm's "Uniform Statistical Reports" are not differentiated between electric or gas operations. This table thus compares the combined gas/electric operations of combination utilities with the single operations of straight companies. The significance of this comparison becomes clear when the gas data are investigated. The combined gas/ electric utilization cf funds is significantly larger than the single service gas utilization of funds although straight gas firms are 10 See Tables 4.5 and 4.8.

PAGE 85

-75* u * u * u u u y w n c o •H 3 \ O u * W H O c d 6 > > > > a) Erf pei erf erf nu & 0) > • • erf H H H \ co P CO •H cu H U M *H oj )-< no O CO oo u 4J e o 4-1 O ctJ T3 O H i-i 3J ca Cw M u H u e C O o CU 0 4-1 IH *S OJ 1 CO ca C u •H O 4-1 H d a 0) 03 4J o o erf •H co cn cn -3 •a "3 o i 0 C IM cu CD •a co n M-l 0 > > a O •H •H 3 a P o fa CD CO 3o

PAGE 86

-76signif icantly larger than the gas operations of combination companies. The data in Table 4.9 indicate that the utilization of funds of the gas operations of the combination utilities is significantly larger than the utilization of funds of the single service gas companies; this comparison probably indicates relative size differences and not the effect of competition. Competitive industries are expected to experience greater risk than monopolistic industries. Firms that experience greater risk are expected to have lower debt/equity ratios than firms that experience smaller risks. The data in Table 4.9 indicate that single service electric utilities pay significantly larger dividends per dollar revenue to common stock owners than the combined gas/ electric operations of the combination firms. This observation indicates that single service electric utilities have lower debt/ equity ratios than combination gas/electric firms and may indicate that single service utilities experience greater risk. Marginal Cost Pricing Marginal cost is the change in total cost when output is increased one unit; marginal cost is the first derivative of the total cost function. A long run cost function reflects changes in costs when all resources are varied. A short run cost function reflects changes in costs when capacity is held constant. Long run marginal cost is the extra operating cost plus the extra capacity cost as output is increased; short run marginal cost is the extra operating cost as output is increased. IT" See Table 4.3.

PAGE 87

-77In the energy industry, "Total Operating Revenue" equals total costs since profits are regulated. For this study, total operating revenues -are defined as long run costs. "Operating Expenses" includes fuel, purchased energy, storage (where applicable), transmission, distribution, customer accounts, sales, and administrative and general expenses. "Operating expenses" does not include maintenance, depreciation, depletion, taxes, interest or dividend expenses. For this study, operating expenses are defined as short run costs. Each of these costs is related to quantity (in KWH and M. Therms) output as approximations of the long run and the short run cost functions. Various regressions were computed in order to investigate the specific characteristics of the cost functions. For instance, both logarithmic and semi-logarithmic forms of the data were investigated. Second degree equations are reduced to linear form if the logs of the data are regressed. In each case, the fit was no better using the logarithmic form than using the original data. This observation indicates that a linear relationship adequately explains the cost functions. A detailed study of the nature of the costs in the energy industry is a thesis in itself and is beyond the scope of this study. The definitions of costs that are used are intended only as rough, relative measures and not as precise, absolute measures of marginal cost. Numerous preliminary relationships were investigated by the author before selecting the definitions as stated. For example, a stepwise linear regression (program "BMD-02R" of the Biomedical Computer Program package of the University of California) was utilized to sort out possible independent/dependent variable relationships. Independent variables measuring quantity output, fuel costs, efficiency, average wages, pension benefits, quantity consumption, taxes, size of plant, and a dummy indicating organization were regressed with each of the dependent variables. In each case, the quantity variable was entered first. No noticeable increase in R was observed as other independent variables were added. Although crude, the assumption was made that a single independent variable (quantity produced) determines

PAGE 88

-78Further verification of the linear relationship involves 2 a polynomial regression which computes Y = a + bjx + + . . . bj^k + e.^ This program ran six steps and the error sum of squares compared to the error sum of squares of the linear regression. There was no noticeable reduction in the error term. In fact, for the single service electric firms there was a significant reduction in the fit of the curvilinear function. Since both the short run and the long run cost functions are linear, both the short run and the long run marginal costs (the first derivatives of the total cost functions) are constant. Table 4.10 presents the results of the regressions; Figure 4 illustrates graphically the general form of the cost functions. The constant long run marginal cost is constant to plant (or system) capacity and then it is undefined. The short run marginal cost is lower than the long run marginal cost in each case. Both the short run marginal cost and the long run marginal cost are higher in the combination computations than in the single service computations. The cost curves of the single service utilities are thus lower than the cost 14 curves of the combination utilities. The single service utilities utilize resources more efficiently than the combination utilities. the value of the dependent variable (cost). This relationship is summarized as: Cost = f (quantity). 13 BMD-05R of the Biomedical Computer Program. 14 It was previously verified that the costs were also higher for the combination companies. See especially Table 4.6.

PAGE 89

-79« Cu 0 03 0) CO cu o u u r-l X! Ctf ,£> 5 + 60 U ctf ctf co 0) 00 4J 0) pi m e 0) fj 4J M ctf 01 c i-5 4-1 O CD U r-l 3 CO ctf Cu 6 ^ •H 4J CO w W P H cn CO o w Q H 10 U i — i pel H a w c o •H go cs rC to — i ctf r-l CO C O •U c | o u i-> u CM o " N vO t-i ON
PAGE 90

-80Q Of CO c •H u a c CO c o •H U K3 C •rJ ^3 u | C CO o C-J CO o
PAGE 91

-81Profit maximization requires lower prices in markets with higher price elasticities and higher prices in markets with greater price inelasticities. A vertical summation of peak and off-peak demands indicates that prices for off-peak customers may be below the long run marginal cost and still be optimal as long as they are above short run marginal cost. A utility may discriminate between classes 15 of customers in an attempt to maximize profits. For example, industrial customers have a greater potential to utilize alternative sources of energy than residential customers. The industrial demand is thus expected to be more price elastic than the residential demand. Additionally, however, industrial customers consume significantly larger quantities of energy than residential customers and thus have lower fixed costs per unit output and lower fixed costs per customer. In an attempt to attract industrial customers, a utility may lower the industrial rate and raise the residential rate. Lower rates to industrial customers which have lower costs are not discriminatory per se, since they may reflect the lower costs. But if the rate differential is greater than that suggested by the cost difference, the utility has discriminated with respect to rates. This discrimination is caused by the desire of any utility (whether a combination or a single service utility) to maximize profits and not by the competition between the suppliers of energy. Table 4.11 suggests a measure of this discrimination. The industrial average price is divided by the residential average Both the vertical summation of the demands and the concept of discrimination were discussed in Chapter II.

PAGE 92

-82. own om r L J'V'I "I'Vl cTV *pui 0RH1 DWOT 4J •H c u CJ u w 01 H 3 w W T3 CJ C 1 — i S3 H CO

PAGE 93

-83price (I .A.P./R.A.P.) to determine the industrial rate as a percentage of the residential rate. In both the electric and in the gas computations, similar differentials exist between industrial and residential rates relative to the class of customer. The industrial rate is (approximately) the same percentage of the residential rate for both the single service utilities and for the combination utilities. Further investigation of Table 4.11 indicates that the difference in rates may be greater than the differences in costs. The long run marginal cost is compared to the average price for each class of customer. A figure is computed that indicates the ratio of average price to marginal cost (A.P./LRMC). In each instance the industrial ratio is less than the residential ratio simply indicating that the industrial rate is less than the residential rate. A second measure of the rate "spread" is computed by subtracting the industrial A.P./LRMC ratio from the residential A.P./LRMC ratio. In both the electric computations and in the gas computations, the "spread" between the residential rates and the industrial rates is greater for single service utilities than for combination utilities. This difference in "spread" is attributable to differences in (marginal) costs between single service firms and combination firms. This difference in "spread" between industrial and residential prices indicates single service utilities have greater differences in rates relative to (marginal) costs (though not relative to classes of customers) than combination gas/electric utilities. This observation may suggest that a single service utility builds a more discriminatory rate structure than a combination utility in

PAGE 94

-84an attempt to take (industrial) customers from their competitors. Alternatively, this observation may suggest that a single service utility builds a more optimal rate structure as it is forced, by competition, to charge prices that more nearly reflect marginal costs. This discrepancy indicates that additional research is necessary on this subject. Summary In this chapter, certain variables have been studied as indications of the degree to which combination utilities compete with single service utilities. Where straight electric companies compete with straight gas companies, rates are lower and sales are larger than where combination companies control the supply of both sources of energy. Sales expenses are greater for separate firms indicating greater non-price competition. Single service companies have lower average costs which may be the result of competitive pressures forcing greater efficiencies. This greater efficiency is further indicated by higher KWH sales per dollar plant and a larger output per dollar input for the single service utilities even though these firms are unable to use their general plant in a joint manner. Single service electric companies spend a larger proportion of revenues on dividends to common stock owners, thus indicating that they experience greater risk. Comparisons relating to the gas operations are not as conclusive for several reasons. First, the gas operations of combination utilities are small relative to the total operations. Second, some data are presented in a combined manner from the combination

PAGE 95

-85firras and are not divided between the gas or the electric operations. Since the combined gas/electric operations of combination companies are larger than the" single gas operations of straight gas companies, the lack of detailed information makes relevant comparisons difficult, if not impossible. Taking these limitations into consideration, some comparisons can be made. Price competition does not appear as significant a factor in the gas computations as in the electric computations. No significant differences are observed in average costs even though wages are higher and pension benefits are greater for the combination companies. The inability to utilize general plant for both electric and for gas operations creates greater average expenditures for general plant for single service utilities relative to combination utilities. This study further indicates that not only are average costs lower, but the entire set of cost curves is lower, for single service utilities relative to combination utilities. The differences in costs create a potential discriminatory situation. Absolute rate differentials are not significantly different between single service and combination utilities. Relative rate differentials are significantly different however: single service utilities charge their industrial customers a rate that constitutes a smaller percentage of marginal cost than combination utilities charge their industrial customers. These findings support the theory that combined gas and electric utilities possess greater market power than separate ownership of the two sources of energy allows. Competition, especially

PAGE 96

-86with regard to the supply of electricity, appears to lower costs and may cause discriminatory pricing. These findings are in general agreement with the. studies conducted by Owen, Mann and Collins.

PAGE 97

CHAPTER V THE REGIONAL ANALYSIS Certain differences were identified in the previous chapter that appear to be associated with competition between single service and combination utilities. The analysis to this point indicates that competition between single service and combination utilities results in lower prices and higher outputs. Straight utilities seem to enjoy greater efficiency, as measured in terms of both average and marginal cost. These utilities appear to assume greater risk as measured by payments to common stockholders. Furthermore, it is observed that in the geographic areas where the ownership of the sources of energy are separated, sales expense is larger. There also appears to be more widespread rate discrimination between classes of customers in those areas where ownership is separate. It was not proven in Chapter IV that these differences are caused by the separation of ownership. It is possible that they are attributable to geographic or demographic factors which may differentially characterize the service areas of single service utilities as opposed to those of combination companies. As a crosscheck on the results of Chapter IV, this chapter classifies the firms into geographic regions and again tests the hypotheses stated in that chapter. If the differences found in Chapter IV are significant in the regional analyses, the hypotheses are supported. If, however, the differences found in Chapter IV are not significant in the regional analyses, the hypotheses are rejected. -87-

PAGE 98

-88Statistical techniques similar to those employed in the aggregative analysis are utilized in this chapter. Since the samples are small, "a "Student t" text is employed to compare sample means . The Regional Distribution The data are divided into the Federal Power Commission's "National Power Survey Regions"; these regions are presented in Figure #5. Most firms are located entirely within one region although some have sales in overlapping areas. Every attempt is made to place these companies in the region in which most of their sales are made. To expedite the statistical procedure, each geographical region is assigned a number. Thus, the Northeast is called Area #1, the East Central is called Area #2, etc. Table 5.1 shows both the absolute and the relative composition of firms within each geographical region. It can be seen that 29 percent of all combination gas/electric companies, 28 percent of all single service electric companies and 29 percent of all straight gas companies are located in Area #1 (the Northeast) where population and industry are concentrated. The uniform distribution between the classes of utilities that prevails in the Northeast does not hold in other areas. For example, single service electric companies outnumber combination gas/electric companies fourteen to five in Area #2, yet combination utilities outnumber single service electric companies fourteen to three in Area #3. Furthermore, two-thirds of all the combination utilities (667,) are located in the northeastern part of the United States (Areas #1, #2 and #3); only one-third of

PAGE 100

-90og u u e u u cn 0) 12 < OQ id a u 3 O OT (M o e o •r-l u 3 •H M U Od U u C o U 4J 2 t-H CM C o 2 CM 00 CO CD < CO cc

PAGE 101

-91all the combination utilities (34%) are located in the southern and western parts of the United States (Areas #4, #5 and #6). Aggregate Characteristics of the Firms As in the aggregative analysis, certain measures are selected as descriptive of the overall characteristics of the utilities. Table 5.2 presents data which show that the electric operations of combination utilities are significantly larger than the single service electric companies in Area #1; the gas operations of combination utilities are not significantly larger than the single service gas companies in Area #1. This observation lends support to the findings of Chapter IV which indicate that the gas operations of combination utilities are small relative to the total operations of the firm. Table 5.2 further indicates that this relative size differential exists only in Area #1. In Areas #2, #4, #5 and #6, no significant differences exist either in the electric or in the gas computations. In Area #3, where single service utilities far outnumber combination utilities (see Table 5.1), the single service utilities are larger in both the gas and the electric operations. Similar findings are presented in Table 5.3. The electric operations of combination utilities are larger than the single service electric utilities in Area #1 and the single service utilities are larger than the combination utilities in Area #3. In Table 4.3, sales expenses and general plant expenses are found on a national basis to be significantly larger for single service companies than for combination companies. Table 5.3 indicates that these differences virtually disappear when the data are analyzed on a regional basis.

PAGE 102

-92o i i i i i i i CO I I I CO CU eg i * * * o o o I I I * * * o o o I I I I I I I I I * * * o o o I I I i-l CO •U 0) E> -< a) ca 4-1 H Of <*-( > o CO co cD U i-l •H CO 4J 00 CO H M C 0) « 4J u o CO -r4 a re M to 0 s .3 4J CO Q V 3 co O (0 u 3 | o u o •H u u o 0) —I W in CO CM i i i I I I I I I * I I • I I I I I I * I I I I I I I I a> u CO CO CO rH CU o> CO 00 t-l OJ o> 3 a o a; ea e 3 3 3 DQ on U u c a) 3 3 01 CO co > aj 1) X 4-1 xj TJ > 01 01 .— 1 o G c u u 1 3 co CO 0) 0) > > 3 T-l --i o 3 U 3 3 > pi on re re 1-1 oo 3 U O •I-l CO 03 CO u cr ia i— i 01 CD i-j C3 CO 00 3 o re .—i 1-4 1 — 1 re 1-4 1-1 .-J 0) O re re »H CO re 3 •H re re 3 r-i c 4J •H •H Cd 4J •H —i w -U t-l rl CX •H c C u 3 u 3 o O 1-1 o 01 w CO a ^ u o> u 4-1 4J -a a C£ CJ re u o cr c 0> 3 H CO 55 • • • • • • CN CO

PAGE 103

-93co (1 •rl CD 4J i-l CD »H > ° g « H O 4J 3 to I I I i I i i i i WWW • I I U CJ I w u I I O O I o o I I I I I I I I w 0) co U & 3 to CO a> a H 4J ca DO C a X — i CD =3 60 3 4J O 4J c ai O 4J Ph a 3 C td cd •H 3 * cd a •H 3 0) 3 H >> 1-1 p H •u .n ai W o 4J ,fi rd o U 4-i a u O t-l B • O i-l cd c a 3 eg td jj a 3 P o CO a H 3 U M
PAGE 104

-94The higher expense figures in the national analysis are therefore attributable to differences in the relative size of the utilities and not attributable to competition between the sources of energy. Table 5.4 further demonstrates the similarity between the types of firms. The significantly larger size of the combination firms in Area #1 and the larger size of the single service firms in Area #3 is obvious. Lack of complete data for the companies in Area #3 does not allow a comparison of financial data. Findings of the Regional Analysis As in the aggregative analysis, certain variables are studied to measure the degree to which the suppliers of gas and electricity compete. Here these variables are studied on a regional basis as additional measures of the "Economies and Diseconomies" and of the "Marginal Cost Pricing" subjects discussed in Chapter II. Economies and Diseconomies Table 4.5 indicates that the electric operations of combination companies charge higher average prices than the single service electric companies charge. This table further demonstrates that quantity sales for residential and commercial customers are significantly larger for the single service electric companies. It was concluded on the basis of these data that average prices are lower and quantity sales higher with competiton than with single service utilities. Table 5.5 lists the same data but here the computations are on a regional basis. The significant differences that appeared in Table 4.5 disappear. Only scattered differences are observed and these differences do not appear to be associated with competition.

PAGE 105

-95m 0) r-t s co cu •H 4J •H 0) i-l 0) 1-1 !-< 4-1 XI co •H 4> W 4J > O CO i-t W U O C •H Ctf 4-1 C CO -H 0) o 4-) .-I &§ M 4? to C o kO in i i i i i i * * * o o o I I I o o n e o •u CtJ (J r o o •H W 4J O o t-l w m m CN I I I I I I I I I .-I I O I 4J c 0 £ CO CU OJ >> 03 O 4J i-i •H u-i §• a w to c SI 0) m M) « o C3 u on cu i— i •i-i co 1 jj C o 3 -o CO *w 3 3 • O O i — I 1-1 CO Q) S g -H CJ O O u u u 3 En Pu o o o 4-1 CO T3 cy u u a 4) ai •a C in c CO > -u i-l i-l «H o p,
PAGE 106

-96Whereas in Table 4.5 there were no differences in the gas computations, it appears that in certain instances single service gas companies charge higher average prices; in one instance, the single service gas companies have higher quantity sales. Only scattered and unrelated differences appear in the average utility bill ($ Revenues / # Customers). These indications are not consistent with expectations if competition is assumed to be a significant force. The "clear" differences in rates and in sales that were isolated in previous studies and in the national computations of this study, do not appear when the data are regionally divided. Thus, the differences in rates and in sales that have been observed are not wholly attributable to the competitive organization of the energy industry, but are instead associated with other forces. These other forces appear to be related to the geographical location of the utilities; these forces are more significant as an explanation of the observed rate and output differences than competition. In Table 4.6 the electric operations of combination utilities were found to have higher average costs although single service companies have higher average sales expenses. No significant differences were found in the gas computations. The data indicate that combination companies have higher costs than single service electric companies suggesting that they are less efficient. Table 5.6 presents the same data computed on a regional basis; again the differences disappear. Only in scattered instances are differences observed. One exception involves sales expenses. Four of the seven significant

PAGE 107

-97vO III I I I I I I cn C o H u a •• ii < n d o CN O l O I I I I I I I I I I I O I I O I I l I o I o I I I I I I cn C O •H iJ C3 •U 3 o o •H U 01 u i i i i i i i i CO a C o 5 C 11 CD 5 c a D PS o H Si OS o H C • • CN CO

PAGE 108

-98differences of Table 5.6 involve sales expense, and all four indicate a higher average sales figure for the single service utilities. The data indicate plausibly that separate ownership creates greater advertising and other selling expenses. Another measure of efficiency was presented in Table 4.7 by testing certain characteristics of the labor market. It was found that the gas operations of combination companies pay significantly higher average wages and pension benefits, although it was indicated that single service electric companies pay higher pension benefits as a percentage of total wages. When the data are investigated on a regional basis, no significant differences are found in the electric computations. In the gas computations, the few differences that are found all involve pension benefits and each indicates higher payments by combination firms. There is no reason to suspect that the differences are caused by competition for they are not uniform throughout all areas. No explanation is advanced in this study for these differences. Additional research is needed but it is beyond the scope of this thesis. In the national analysis it was observed that combination companies have larger operating revenue per dollar of general plant, larger quantity sales in relation to general plant, and a smaller general plant as a proportion of administrative and general expenses. Single service companies however, have larger dollar investments in general plant as a proportion of total plant. These factors all support the view that combination companies operate more efficiently with regard to plant by using common facilities in a joint manner. Separate ownership causes duplication of resources and a less

PAGE 109

-99ii iii ii iii ii iii ii iii ii iii ii iii i i i i i i i i i i • i i i i cj l I o l l l l O I I I I I I I I I O I I I I I I I I I I I I I I * O I I i i w i i CO o 03 u o H co 01 i-i « 00 4J iH 4-1 c cd & 0! CO C 0) &. 0) g c oi > 0) CtS 00 fx] 0) 60 t-l 60 C « C fl 00 -H 4J 4-1 co >> « u u u CD i-l CD « «-i 3 i-i (SCO) U 4J O O H H 03 co CD CD i-i CO 1-4 CO" CU 05 00 t-l 00 « >* oo S*. u 4-1 -H >> i-l 4-1 4J *J C ft C 9 W « 3 3 3 C ctj O"-^ 3 CD C/ 01 cn C O i-l 60 3 4-1 C c« ctj 3 U 4-1 c 1-1 CO o) 3 CO 0) a o u c3 M 0) <§•: o 0) PL. d) Q CO a) CO i-4 01 CO t— 4 00 CO CO CD 00 >M-< 4J (0 >itIM 4-1 4-1 •H 3 >> 4J tt 4J 3 3 i-i co cy 4-i 3 3 O* 0) cd v. B 3 01 3 C CO CD a. oi x co a, w s X Q> u c o. OX c^w O 4J •H 3 « OJ O i-l 3 H •a c a) M 1-4 3 fn O O 01 CO c Oi a x CO w cu i-l i-l cO CO oo u 0) >» 3 4-1 01 •H O 4-1 3 T3 co 3 3 co co C 3 O) o CO -H C -u 0) co Ou U U X 4-1 i-l " W CO 4J i-l 3 co 3 cO OJ -h 3 C o 4-1 01 4J O co co i-i P CO •< CM
PAGE 110

-Inefficient operation. In Table 5.8 the data are computed on a regional basis as they were computed in the national analysis. Most of the differences cited above appear only in the electric computations of Area #1. In both the national and in the regional computations, single service utilities have a larger percentage of their total plant absorbed by general plant. However, the importance of this observation is not large for the differences in Total Revenue/ Total Plant and in Quantity Sales/Total Plant are not significant. The joint utilization of resources generates efficiencies that are unavailable to competing firms. These efficiencies are relatively minor in relation to the total operations of the combination utilities. Table 4.9 presented data that indicate that single service utilities, relative to combination utilities, devote greater proportions of their capitalization to common stock. This fact suggested that single service utilities experience greater risk than the combination utilities experience. The same data are presented in Table 5.9 utilizing regional computations. Here the difference not only disappears, but in one instance, is reversed. The single service firms do not apportion larger amounts to common stock than the combination firms in any of the comparisons. In fact, in the electric computations of Area #4, the single service firms apportion smaller amounts to common stock than the combination firms. The data for the combination utilities are not divided between their electric and their gas uses. Thus, this observation may simply indicate that the data for the combination utilities combine both the gas and the electric utilization of funds. However, these

PAGE 111

-101g •H m ft in co o CN * 0) H s co cu 60 « 0) to C3 !-i | u o 01 w CO CO 0) a) 60 ai CO ca 3 0 4J 03 C i-l H 0) OJ i-l 4-4 r-4 d ai > cci 1OJ US >> a) 4J H c 4J n s o oi O oj o S) C 1-4 H a ?a H a 3 CO > « 4-4 C c 4J cu w 4J o 0) o •H co erf o > co H * — CO CM a) =8= H 5-3 4J ta OJ 4J OJ 60 pi t-l c H a cc3 S a IW o >. cw OJ 3 •H ca 60 a) PL, 0 pg i-i 0) CO go CO C c r-t C CtJ OJ cu a> * to U cq 00 £ PC 0 4J OJ 0 on « CO « CT5 W •H 0 1—1 cu 3 M C co H <§ CO 0 OJ h o C 1I— 1 i-l « •H 0) 1-1 OJ w CO CC 3: co ft CO ft c td u 4J C C 4J o O cu 0) o H H ft PL| H • • • 1 i-i c-i CO «ct in I O v© CO 0) 60 ft 4-1 X cj 0) OJ CO PS o c« •u 0 2

PAGE 112

-10200 m d U 3 d d O •H BO 1 1 1 1 1 1 1 1 I . 1 I 1 cn 3 1 1 1 1 .... . I 1 1 1 o t-> -X * S) f < O 1 1 1 1 lilt 1 1 o •T-l 1-1 u CM tilt 1 u o CJ r-i M 1-H 1 1 CJ 1 l l O . w c 4J 4J « 4-1 e e .-< e d cS P-i d r-t i-H i-l PU CX< i-l eu cd • • V4! 1-H •u cj d CO C 4J O 3 d Ph H —i CJ 0 4J 4-1 4J 4J 4-1 Ph Q U H e 3 3 73 C d 4-i 4-1 d 4-1 O 4-1 d 0 d t-i 3 3 1-4 3 4-1 3 3 1-1 4J 3) 1-4 P-i d <5 Ph d 3 d d Ch Ph CJ CJ CJ i-i 1-1 i-l cd i-( r-4 n 3 3 3 3 /-\ P-i P-I i — 1 Ph i— i Ph Ph i— 1 1) S s 3 3 3 3 d Ph cd CD CJ CJ CJ G CJ • • U i-t r-4 i-H 4-1 3 J) > > > > 4-1 CJ d i-i d cd 0 S) 3 CJ CJ s 11 o 00 3 4-4 d 4J 4-1 H p. CJ pi p; erf ^ u — i CJ c 4J 0 o > Ph Q H O H H \ W & to 60 to co H fcj c 3 3 3 4J <+< -J n •H i-l CO 3 o 3 CJ u 4-1 4J 4-1 CJ CO n en CO 4J 4J d O 3 d d d cd i-l CJ CJ CJ CJ 4-1 3 3 i-H •H 3 H Jh Sh qj i—l i— i i—i r-H 3 re cd Ph •-4 4-1 CJ CJ CJ CO CJ co d cd d cd cd r-4 — H -3 3 ev Op <5 CO CO CO CO p i r I Ph Ph al og 3 ib erf 4-1 • • • • O 4J M • • •H 3 3* 3* 3 s irt 4-> 01 •-4 3 4-1 r-H 4-1 4-1 4-1 4-1 4-i d cd d d 0 02 3 4-1 d co d o o o O 3 3 3 3 3 U H CJ d i— i •-4 •u H H H H d oc c c Ph G O i-i Ph a o 3 CJ H c OS • • • • iH CM CO CJ 00 3 CJ a d CJ 3 CJ CJ 3 i-l d d >H • CJ C 3 tH CJ B o <3

PAGE 113

-103in C8 CU p3 to c co -a s O 4-1 > > P-i u a 4-1 H > H S-i •H H a Q Q

PAGE 114

-104observations do not, in any manner, support the suggestion of the national analysis that single service utilities experience greater risk than combination utilities experience. Marginal Cost Pricing The level of average cost was shown in the aggregative analysis to be higher for the combination firms than for the. single service firms. Table 4.9 presented the results of linear regressions in which the slope of a total cost function is used to estimate marginal cost. Marginal cost is higher for the combination firms than for the single service firms. Thus, the findings of the aggregative analysis support the suggestion that both the average costs and the cost curves are lower under competitive conditions than under monopoly conditions. Linear regressions also have been calculated and are made for the data on a regional basis. The results of these regressions are summarized in Table 5.10. Three factors are signficant. First, the marginal costs in all but one instance are higher for combination firms. In Area #5 of the gas computations, the combination utilities have lower marginal costs than the single service gas utilities. Due to the small number of combination firms (2) in Area #6, a regression is not possible and thus no comparison is presented. The second factor relates to the differences in costs between the classes of utilities. For instance, long run marginal cost (LRMC) in the electric computations of Area #3 is .19225 and in Area #5 it is .10620. The marginal cost in Area #3 is almost double that of Area #5. These cost differences are explained by

PAGE 115

-105u 0) §r H CO V ft U i-l (0 0) o J2 U o) to C C O -H •H 60 co U m ct5 0) 2 H to o a) .c ps 4J M CD 4-1 | I •J f-i 4J H-l CO o w 3 CO 2 n e o JJ co a) cd 4J O 3 &> CO C 0 O 4-1 4-1 aj U 4J O) 1-1 W | O CO cr. ^ — . \D /— n 00 S * ON ON <* ^-^ CN CN i-l 1-1 o CO CN NO o CN cn O » CM ON i NO rm fx ON o cn o co o ON o co o ON o On • a • • • • • N-/ c o •H 4J ed C t-) i-CM rO Pi 6 (0 O
PAGE 116

-106o 00 oo vD IT) ON CO <) Os O On vO CM co on o on on r~co vo r>en i-4 On ON oo oo r»» on m oo o on c o •w u (0 c •HCM .O OS < CO o on CO CM co oo O ON On i—l ON CO
PAGE 117

-107exogenous factors such as the cost of fuel that accrues to the firms due to geographical location. The third observation involves the 2 R factor which indicates the amount of the variation that is explained by the regression. In the national computations the fit was better for the electric operations than for the gas operations. In Table 5.10 the fit for the gas computations improves in all areas except Area #1. While there are differences between areas, the firms within areas are similar and thus the regression coefficients represent good estimates of the costs. The measures of discrimination that were developed in the aggregative analysis are applied to the data on a regional basis. Table 5.11 (Parts 1 and 2) presents the results of these discrimination computations. The Industrial A. P . /Residential A. P. ratios are similar within each area although these ratios vary considerably between areas. For example, the I .A .P . /R .A.P . ratio for the combination utilities of Area #2 indicates that the industrial gas rate is 60 percent of the residential gas rate; the I .A.P . /R.A.P . ratio for the combination utilities of Area #6 indicates that the industrial gas rate is 20 percent of the residential rate. Since rate levels are established by each state's regulatory commission, the uniformity within regions is expected. The differences between areas reflect differences outside the control of a single state regulatory agency. There is no way to assertain, from the data at hand, the causes of these inter-area differences. A detailed study of individual characteristics of each customer is necessary. Table 5.11 presents measurements of the average price/ marginal cost relationship introduced in the previous chapter. The

PAGE 118

-108Table 5.11 Average Price as a Percent of Long Run Marginal Cost Area #1 COMBINATION UTILITIES: Residential Industrial ELECTRIC OPERATIONS CAS OPERATIONS 2.6957 1.4456 54X 1181 631 58* 15.389 7.046 461 1181 54X 64* Area tl SINGLE SERVICE UTILITIES: Residential Industrial 2.4218 1.2789 531 135X 7 IX 64* 16.540 9.753 57X 2121 1261 861 Area #2 COMBINATION UTILITIES : Residential Industrial 2.2403 1.1485 511 120X 621 58X 8.753 5.277 60X 991 607. 391 Area #2 SINGLE SERVICE UTILITIES: Residential Industrial 2.1441 1.0763 501 134X 62% 67X 9.385 4.763 5 IX 131X 66X 65X Area #3 COMBINATION UTILITIES: Residential Industrial 2.3391 1.2115 52X 117X 6 IX 56X 10.172 4.606 45X 133X 60X 73X Area #3 SINGLE SERVICE UTILITIES: Residential Industrial 2.4791 1.1820 48X 129X 6 IX 68X 11.832 4.668 39X 320X 126X 194X

PAGE 119

-109Table 5.11 continued ELECTRIC OPERATIONS GAS OPERATIONS u o G -< > u Area #4 COMBINATION UTILITIES: Residential Industrial 2.1250 1.0149 481 1221 581 641 11.090 4.559 411 2231 921 1311 Area #4 SINGLE SERVICE UTILITIES: Residential Industrial 1.8931 .8719 461 1121 511 611 12.674 3.762 301 3841 1141 2701 Area #5 COMBINATION UTILITIES: Residential Indus trial 2.5328 .9849 391 2311 901 1411 8.069 3.156 391 2481 971 1511 Area #5 SINGLE SERVICE UTILITIES: Residential 2.2130 Industrial .9838 441 2081 931 1151 8.312 3.035 361 1761 641 1121 Area #6 COMBINATION UTILITIES: Residential Industrial 2.0092 1.2206 611 20.674 4.112 2VZ Area #6 SLNGLE SERVICE UTILITIES: Residential Industrial 1.8192 .9240 517. 1231 631 601 13.941 3.571 261 2451 631 1821

PAGE 120

-110residential rates in all computations are greater than the LRMC. The industrial rates in most computations are less than the LRMC. However, the industrial rates for single service gas utilities in Areas #1, #3 and #4 are greater than the LRMC . The R. A. P. /LRMC I. A. P. /LRMC "spread" does not establish any consistent trend. In three of the five electric computations and in four of the five gas computations, the single service "spread" exceeds the combination "spread". However, in two of the five electric and one of the five gas computations, the combination "spread" exceeds the single service "spread". Because there exist variations in the empirical results, it is not possible to attribute the differences in "spreads" to competitive forces. Summary of the Regional Analysis In Chapter IV, several factors were isolated that indicate significant differences between single service utilities and combination utilities. Most of these differences are not significant when examined on a regional basis. For example, in Chapter IV we observed that single service firms had lower rates and larger outputs, lower average costs, greater sales per dollar of plant and assumed a larger amount of risk. These differences were not significant when studied regionally and thus must be associated with geographical or demographical factors within each area and not by competition between the sources of energy. A few factors do remain significant when studied regionally. First, single service companies incur larger sales expenses than the respective operations of combination utilities. These sales

PAGE 121

-Illexpenses are not large enough to raise the average total costs of the single service utilities above those of the combination utilitie Economic theory predicts that competiton between firms with differentiated products results in large expenditures on advertising. Our findings support this prediction. Since there is no measurable increase in average costs, this difference does not impose a serious economic loss. Second, joint utilization of general plant allows combination utilities to utilize resources more efficiently than single service utilities. The importance of this economy is again minimal for the efficiency in the utilization of the total plant (including general plant) is not significantly different between firms. Thus, while economic theory predicts an efficiency gain where combination firms exist, this economy is not large enough to be significant. The final factor observed in both the national and the regional studies centers on the large variation in rates between residential and industrial customers that are charged by the utilities. However, this difference cannot be attributed to competition since the direction of the spread is not uniform.

PAGE 122

CHAPTER VI POLICY IMPLICATIONS The organization of the energy industry is presently characterized by single service electric, single service gas and combination gas /electric utilities. This thesis examines the hypothesis that organizational separation of the supplies of gas and electricity increases competition between utilities and thereby reduces costs. Though perfect competition is the organizational ideal upon which resource allocation is based, it is not attainable under either "straight" or "combination" organization. The question then is whether "workable competition" can and/or should be obtained. Combination firms are expected to have lower common costs due to joint utilization of assets and lower fuel costs due to higher load factors. These factors are expected to lower both the cost curves and encourage fuller utilization of plant for the combination companies relative to single service utilities. On the other hand, joint utilization of resources encompasses a relatively small portion of total operating expenses and single service firms have both the ability and the incentive to increase their load factors. Additionally, regulation is expected to encourage efficiency in the overall operation of all utilities, whether single service or combination in organization. To the extent that these factors are significant, single service utilities are expected to have lower -112-

PAGE 123

-113costs and greater efficiencies than combination gas/electric utilities. However, it has been suggested that regulatory agencies have not recognized technological developments in the energy industry and thus have not forced the realization of potential efficiencies. It seems there does not exist one attainable organization that can be expected to be superior to others on an a priori basis. It appears that empirical analysis is necessary. Data and Methodology The empirical method by which the hypothesis is tested entails a comprehensive cross-sectional analysis of variables that are expected to be influenced by competitive forces. Data were requested from 133 single service electric utilities, 77 combination gas/electric utilities, and 73 single service gas utilities. Of the initial request, 75 single service electric utilities, 51 combination gas/electric utilities, and 36 single service gas utilities provided the detailed information requested. The data were subjected to a statistical test that compared the mean values of each variable utilizing a "z" or a "t" statistic. This analysis is conducted on both a national and on a regional basis. The Aggregative Analysis On a national basis it was discovered that the level of rates is lower and the volume of sales is higher for single service firms than for combination gas/electric utilities. Sales expenses are greater for single service firms, supporting the prediction that competition results in a greater degree of non-price competition than monopoly. Single service utilities also have lower average and

PAGE 124

-114marginal costs, indicating that competitive pressures (or, in any case, separation) may force greater efficiencies. Further evidence of increased efficiency is indicated by higher kilowatt-hour sales per dollar plant investment and greater kilowatt -hour output per dollar resource input in spite of the fact that these firms are unable to use general plant in a dual manner. Straight electric companies distribute a larger proportion of gross profit to common stock leading to an inference that they incur greater risk. Single service utilities discriminate between types of customers to a greater extent than combination firms suggesting the possible "taxation" of customers with low price elasticities to subsidize customers with high price elasticities. The differences between single service utilities and combination utilities that are discovered in the national analysis characterized both electric and gas operations although the computed differences are more marked with respect to the electric than to the gas side of the operations. The findings conformed to those reached in earlier studies which incorporated similar data and methodology. The conclusions support the view that combined gas/electric utilities posses greater market power than single service utilities. Competition, especially with regard to the supply of electricity, appears to be a force that reduces costs and lowers prices. On the basis of these general findings, the Metcalf proposed divestiture is supported. The Regional Analysis The same statistical methodology was applied to the data subdivided into geographical regions. Significant differences are

PAGE 125

-115observed between the results of the aggregative analysis and of the regional analysis. Most of the variables that are significant in the national study are not significant on a regional basis. The present analysis, as well as the analyses of studies preceding this thesis, utilizes average revenue as a rough index of the rate level of the utilities. In these studies, average revenue is significantly lower for straight utilities than for combination utilities. Since average revenues are not found to be significantly different between the utilities on a regional basis, an alternative explanation for the differences in the national and the regional findings must be formulated. It was discovered that the distribution of the two types of utilities is not geographically uniform. Specifically, single service utilities are concentrated in the south and in the west where climate causes high per-customer consumption regardless of the organizational structure. Since the price charged per unit is inversely related to the quantity consumed, the higher consumption caused by the demographic characteristics reduces the average revenue figure. The causal relation is thus verified: Larger consumption causes lower average revenue (average price) . Differences observed as significant on a national basis and not significant on a regional basis are due to geographic or demographic factors within each area and not due to competition between the sources of energy. A few factors do remain significant when studied regionally, however. First, straight companies incur larger sales expenses than do the respective operations of combination gas/electric companies, although these sales expenses are not large enough to raise average

PAGE 126

-116total costs of the single service utilities above those of the combination utilities. Economic theory predicts that competition between duopolies with differentiated products would result in large expenditures on advertising. The findings support this prediction. Since there is no measurable difference in average total costs, this difference is not considered significant. Second, combination companies devote a smaller proportion of total plant to general plant than single service companies. This factor indicates that combination companies use their general plant more efficiently by utilizing general plant items in a dual or joint manner. Total Revenue/Total Plant and Quantity Sales/ Total Plant are not significantly different between the combination gas/electric firms and the single service firms, however . Thus, while economic theory predicts an efficiency gain where combination firms exist, this gain does not appear significant. The final factor observed in both the national and the regional studies involves the large variation in rates between residential and industrial customers. In some regions the discrimination is greater for single service firms while in other regions the discrimination is greater in the case of combination gas/electric firms. Thus, this difference cannot be attributed to competitive factors. The differences suggest that some utilities may base rates on demand elasticities to a greater extent than other ''"Since this thesis has been prepared, Professor John H. Landon has conducted a study that investigates the performance of electric and gas utilities. ("Electric and Gas Combination and Economic Performance," Journal of Economics and Business , Fall, 1972, pages 1-13.) Professor Landon uses data from the Federal

PAGE 127

-117Comparison of Expected Benefits and Costs The analysis developed in this study indicates that at the present time and given the present institutional constraints, competition between gas and electricity is not a strong force. Single service firms undertake larger advertising compaigns than combination gas/electric firms. This greater emphasis on advertising leads to greater average sales costs for single service utilities relative to combination utilities. On the other hand, managers of separate utilities may be encouraged, by competition, to operate their utility more efficiently than they would under combined ownership. If the competitive stimulus is stronger than the diseconomies discussed above, average total costs are expected to be lower for the single service utilities than for the combination utilities. If the diseconomies are stronger than the competitive stimulus, average total costs are expected to Power Commission's Statistics of Privately Owned Electric Utilities in the United States . This thesis uses data received directly from each utility. Professor Landon attempts to overcome geographical differences among the utilities in a national study by including measures of exogenous factors that tend to vary by region. These exogenous factors are: taxation, temperature, type of generating equipment, and fuel costs. This thesis attempts to overcome geographical differences by studying the data in parallel but independent regional studies. Professor Landon uses ordinary and two-state least squares multiple regression analysis. This thesis uses a statistical comparison of the mean values of critical variables. However, even though these two studies use different sources of data and different metholologies , the conclusions are similar. For example, the Landon study finds prices are not higher for either type of utility; there is no significant relation between combination and kilbwatt-hour sales; and sales costs are lower for combination companies. Professor Landon concludes that combination utilities are not successful in exerting any monopoly power they may possess.

PAGE 128

-118be lower for the combination utilities than for the single service utilities. The regional analysis indicates that the average total costs are not in fact significantly different between the single service and the combination utilities. Thus, the separation of existing combination gas/electric firms into independent single service firms is not expected, ceteris paribus, to lead to gains in efficiency. The benefits that are expected to result due to the proposed divestiture are not large. The costs of the proposed divestiture may be significant. It is reasonable to assume that the smaller acquiring (single service gas) company will have higher interest costs since its financial rating will not be as high as that of the old, combination utility. The increased cost is expected to be significant since the industry is capital intensive and traditionally places heavy emphasis on department financing. Allan C. Mustard, Senior Vice President, South Carolina Electric and Gas Company, estimates that ". . .the money costs could increase some 20 to 30 percent, consequently 2 the rates could increase substantially." Frederich T. Searls, Vice President and General Counsel, Pacific Gas and Electric, estimates the divestiture would cause ". . .an increase in cost of 10 percent to gas customers and 18 percent to electric customers."' Robert H. Willis, President, Connecticut Natural Gas Corporation, questions the above estimates as being exaggerated 4 in magnitude. Mr. Willis suggests that all utilities, whether Combination Utility Companies , page 29. 3 Ibid. , page 65. 4 Ibid. , page 383.

PAGE 129

-119old or new, face higher interest rates in the current money market as they refinance maturing bonds. Most of the utility bonds issued at low interest rates are now maturing so these bonds have to be paid off. And as they are paid off with new money at sometimes three times the existing rates, this obviously raises the weighted average debt cost for the utility. In addition, most utilities are roughly doubling their size every ten years and they have to issue new securities in order to meet expanding capital requirements. The increased interest cost for new issues plus the increased interest cost for refinancing is therefore expected to raise the cost of capital for all utilities. The benefits of divestiture have been described as relatively insignificant. The costs have been estimated as positive and possibly substantial. Since the costs are expected to outweigh the benefits, the hypothesis examined in this thesis must be rejected. Categorical divestiture employs a meat-ax approach in dealing with an important segment of the economy. It sets no standard for good and bad, but seems to assume that all combination utilities are bad. Alternative Policy Suggestions The rejection of the primary hypothesis that requires divestiture does not justify the acceptance of an alternative hypothesis that would permit the formation of new combination utilities. Competition in the energy industry may significantly increase as innovations such as fuel cells, gassif ication of coal and breeder-reactors are introduced. Combination utilities are

PAGE 130

-120not expected to introduce technological advances as rapidly as single service competitive firms if the effect of the technological change simply reallocates existing customers within the same organization. The advancement of technology may increase the benefits of separate ownership and therefore make divestiture both realistic and feasible as a policy matter. The energy industry must be carefully monitored to observe and respond to changing competitive conditions . The decision to reject the proposal for divestiture is based on conditions of the energy industry as they presently exist. The analysis suggests that the performance of a regulated duopoly does not exceed the performance of a regulated monopoly b y a margin sufficient to offset the divestiture costs. Regulatory commissions have failed to encourage competition and, in fact, have attempted to eliminate competitive pressures in many instances. It is possible that a change in regulatory attitudes and methods could, in and of itself, make the proposal for divestiture feasible. Alternative policy approaches that consider each case on its own merit are more realistic. For example, with no change in present policy or legal status, any merger proposed today that would form a new combination utility, is subject to Section 7 of the Clayton Act and, perhaps, to Section 2 of the Sherman Act. In order to support divestiture measures in addition to these, the traditional antitrust standards such as the observance of price discrimination, tying contracts and other related predatory practices should be followed. Such case by case analysis will direct divestiture

PAGE 131

-121only where benefits are realistically expected to exceed costs. At the present time, stringent implementation of present policy appears to be sufficient-

PAGE 132

APPENDIX A: Results of the Statistical Computations: The Aggregative Analysis

PAGE 133

AREA: TOTAL UNITED STATES PACE 1 sTlectric caggTMiais CAS COMPUTATIONS COMBINATION 31 ST. ELECTRIC »-75 1. TOTAL POPULATION SERVED 2. SQUARE MILES SERVED 3. POPULATION SQUARE MILE 4. NUMBER OP CUSTOMERS Residential _MEAN_ 13,400,435 130,048 3 23 385,502 STANDARD DEVIATION 17,656,752 197,954 2,433 515,813 K£A» 13,343,313 152,091 358 355,534 STANDARD DEVIATION 1 13,991,391 .02 175,604 .64 655 +1.13 COMBINATION • -51 STANDARD MEAN DEVIATION 984,968 1,438,000 7,456 16,199 1,051 1,828 . CAS -36 400,854 + .35 221,943 357,900 MEAN 1,901,721 11,171 1,269 280,497 STANDARD DE VIATION 2,112,993 27,702 3,554 301,547 2.26 .72 .34 .8 Commercial 49,232 75,018 43,129 44,922 .52 19,485 29,461 22,855 22.471 .60 Industrial 1.693 2.158 2,313 2,097 -1.14 525 1,299 1,560 2,145 -1.83 REVENUES RECEIVED Residential 54,151 66,202 54,691 38,421 .05 28,927 40,320 43,869 53,845 -1.41 Commercial 41,308 75.429 38,688 49,724 .22 9.717 13,208 12,274 12,401 .92 Industrlsl 30.961 17,115 14,181 35,612 .49 10,526 19.905 23,753 24,551 -2.68 QUANTITY SALES Residential 2.221 2,523 2,630 2,654 .70 255,772 382,758 397,864 475.312 -1.49 Commercial 1,836 5,207 1,981 2,480 .27 109,136 147,062 152,482 178,701 -1.20 Industrial 2,702 3.303 1,572 4,028 -1.13 226,181 480,617 563,921 660.902 -2.62 AREA: TOTAL UNITED STATES PACE 2 ILRCTRIC CCtmiTATIqiS COHBDIATTON ST. tXECTllC CAS COMPUTATIONS »-31 •-71 e-51 tTAirDARS STANCAP.S STANDARD MEAN DEVIATION MEAN DEVIATION MEAN DEVIATION 1. EXPENSES: Operating 371,005 784,204 584,006 611,320 .10 14.701 55.695 Maintenance 103,997 156,444 98,193 111,017 + .23 2,190 3.468 Depreciation 138,713 203,197 159,744 175,342 .01 1,159 5,164 Production 413,043 553,166 424,064 471,289 .12 26,255 44,125 Distribution 99,786 138,924 88.021 104,570 .51 4,678 6,952 Customer 40,685 68,484 14,758 36,327 + .58 1,877 3,053 Sales 16,229 18,619 26,767 22,350 -2.86 764 1,048 Administrative and General 86,195 116,984 86,069 100,753 + .01 2.847 4,047 TOTAL OPERATING 1.098.232 1.438.994 1.089. 058 1.192.816 * .04 46.821 71.443 2. PLANT: Production 2.U0.458 3,199.151 2,356,821 2,504,057 .04 6,424 12,801 Distribution 2,432,098 3,688,483 2,214,353 2,303,692 + .41 93,593 131,772 General 103,257 119,016 216.374 202,978 -1.93 2,221 4,025 3. TOTAL INTEREST 171,896 230,822 148.133 158,166 .64 17,190 23,082 4. TAXES: State and Local 177,518 329,708 111,997 177.938 + .90 4,206 6,593 Income 76,569 108, 130 98,836 118,995 -1.09 2,086 3.237 5. AC/rTD 2,984 1,597 1,187 1,644 -1.11 6. BTU/KUH 11,500 10,189 10,129 3,232 + .30 ST. CAS 0-36 66,778 2,676 6,438 50.784 5,705 2,961 1,723 5,297 STANDARD DEVIATION 62,496 2,392 7,779 50,244 5,345 2,842 1,509 5,183 87.801 18,062 121,879 11,792 6,031 4,692 30,725 126,097 10,499 4,064 6.278 -2.47 .77 -2.21 -2.36 .78 -1.70 -3.29 -2.37 81.015 -2.44 -2.14 -1.01 -5.21 7,830 »3.02 -1.33 -2.29 -123-

PAGE 134

-124AREA: TOTAL UNITED STATES FACE 3 ELECTRIC COMPUTATIONS CAS COMPUTATIONS COKE ISA! ION ST. ELECTRIC COMBINATION ST. GAS •51 •-75 •-51 036 MEAN STANDARD DEVIATION MEAN STANDARD DEVIATION t MEAN STANDARD DEVIATION KEAN STANDARD DEVIATION t I. EMPLOYMENT Humbar of Emploveae 3,668 3,182 2,932 3,149 + .36 913 1,544 1,576 1.308 -2.16 Total Haste 29, 440 44,420 30,483 36,187 .13 8,809 14,502 14.514 12,374 -1.97 Peneloa Beneflta 4.549 8,650 3,532 5,156 + .70 4,581 8,410 1,675 1,743 + 1.40 FINANCIAL SOURCE OF FUNDS: From Outelde 42,313 54,882 41,496 44,780 .08 39,888 54,106 63,026 90,690 67,220 87,911 • .24 60,914 88.839 16,277 20.270 <-2.»5 20,530 35,260 <-2.94 APPLICATION OF FUNDS Croat Flua to Plant 74,615 DlrloWa on Preferred 2,892 Mrldanoa on Coaeaan 14,696 104,211 5,780 19,144 63,862 7,703 13,773 111,263 40,896 15,348 .61 .86 + .17 71,707 2,775 16,046 102,090 3,633 18,788 16,928 476 3,620 19,874 a-J.73 192 +2.86 6,216 43.05 TOTAL APPLICATION Of FUNDS 106,622 .18 102,349 33,108 +3.02 AREA! TOTAL UNITED STATES PACE 4 tUCTlUC COMPUTATIONS CKSLRATIO* ST. ELECTRIC •41 »-73 JQUB_ STANDARD DEVIATION 8A3 cctavTAnow COMBINATION ST. CAS ••SI a-34 ME.'N STANDARD DEVIA TION sollar TunENUE/guAimn ReiWentlel .24064 Ceaaaetelal .23220 Industrial .12132 .03324 .05938 .03623 .11698 .10837 .10671 .01983 .04018 .02534 1.70 +1.49 .11894 .09021 .03191 .04214 .03237 .02236 .1283} .10832 .03632 .04497 .08215 .0311] .98 .75 .49 1. ODANT1TT/NO. CUSTOMERS Eealaaaf lal talal Industrial .06438 ,34233 13.13439 .01421 .11373 34.06103 .07313 ,41138 30.31262 .02391 .19998 88.93203 •3.16 1.27674 •1.74 6.46778 •1.19 945.44385 .45496 2.66651 1.30749 4.69658 2900.95752 1565.82251 .44691 2.88188 3162.22144 .32 .03 .93 I, DOLLAR RET. /NO. CUSTOMERS Evidential ,01302 Compare lal Industrial .07324 4.04950 .00174 .01074 9.99335 ,01332 .08332 4.37389 .00251 .03114 6.58780 •1.03 .14491 •1.73 .54398 •1.33 37.72607 .04132 .21619 108.28311 .15312 .58841 65.01216 .03767 .19890 117.59773 .96 .95 I. 10

PAGE 135

-125TOTAL UNITED STATtS electric cch futatiq ns combination «t. uremic m*!l »-7J STANDARD STANDARD HEAW DEVIATION MEAN DEVIATION CAS COH PUIATIOKS COMBINATION ST. GAS •-51 o-JO STANDARD MEAN DEVIATION STANDARD Dr.;»TioN_ 1. COCTS/qUANTlTY Total Operating Ravanue/ Quantity .20514 >. .52 .18099 .04102 +3.22 .0S612 .03270 .08777 .05837 • .11 Total Operating Expenae/ Quantity .15999 .03623 .13*99 .03744 3.13 .07490 .02765 .07553 .05046 .08 Operating. Expenaa/Qusntlty .06552 .02767 .07640 .02(67 1.79 .05515 .0186t . 05655 .03859 .20 Maintenance,' Quantity .01291 .00446 .01154 .00426 1.35 .00354 .00348 .00246 .00179 +1.89 Decree lac Ian/Quant 1 ty .02255 .00600 .019)5 .00515 2.62 .00503 .00252 .00538 .00448 .42 Preduc t loo/ Quant 1 ty .0*221 .02252 .05462 .02290 1.(5 .04109 .01102 .04153 .03196 .08 Matrlbut loo/ Quantity .01115 .0040* .01124 .00467 2.43 .00785 .00612 .00588 .00463 1.70 CuatoaMr/Quaatlt* .00552 .00206 .004 SO .00172 2.06 .00J01 .002)0 .00310 .00241 .17 f alaa/Quantlty .00 256 .00136 .00356 .00114 -4.32 .00147 .00173 .00196 .00173 -1.29 Adailn. * Canaial/Quantlty .0120} .00399 .0111] .00395 1.26 .00485 .00311 .00484 .00328 .01 AREA: TOTAL UNITED STATES PAGE 6 lucnqc COHrVrATIONS CmnUTIOH (T. ELECTRIC •-31 «-75 STANDARD DEVIATION STANDARD DEVIATION CAS COMPUTATIONS CCKSINATION • 41 ST. CAS —36 STANDARD DEVIATION STANDARD DEVIATION 1. TOTAL WAGES/NO. BtTLOTKES 9.41974 2.28896 9.83874 1.27335 -1.1J 11.44239 8.17420 9.00073 1.44660 +2.09 2. TOTAL WAGES/TOTAL OPERATING REVENUE .1(769 .03294 .19332 .03837 .61 .16637 .06702 .15485 .04570 .97 3. KKSIOIf MNEPITS/1IO. EMPLOYEES .88992 . 51744 1.06251 0.4*212 -1.77 5.17881 4.92001 1.01703 0.53734 +5.99 4. TORS ION BENEFITS/ TOTAL WAGES .08719 .04717 .10801 .04373 -2.31 .32937 .46542 .11286 .05288 ««.33 3. TOTAL OPERATING REVENUE/NO. MPLOYEES 31.531(6 9.41531 52.33979 9.83355 .44 69.30168 28.94148 64.69197 30.91081 .63

PAGE 136

-126area: total united states pack 7 ELECTRIC COHPUTATIONS GAS COHPUTATE KS COMBINATION ST. ELECTRIC CC+1BINATI0H ST. GAS n-51 • -75 n-51 n-36 MEAN STANDARD DEVIATION STANDARD DEV IATION MEAN STANDARD DEVLATION STANDARD DEVIATION r t. TOTAL OPERATING REVENUE/ ( ) PLANT Total Rev. /Prod. Plant .85491 .80423 .74041 .47121 + .92 53.98720 127.20238 Total Oper. Rev. /Diat. Plant .63718 .11463 .65532 . 12964 .83 .60878 .19196 .85007 .81033 -1.75 Total Oper. Rev. /Gen. Plant 27.50652 48.41737 7.06309 2.77855 +3.01 39.96916 36.11139 9.84339 8.01672 +3.78 Total Oper. Rev. /Total Pit. .23229 .05801 .22282 .04739 +1.04 .46235 .11824 .46040 . 18385 + .06 2. QUANTITY SALES/ ( ) PLANT Quantity Salea/Prod. Plant 4.20131 3.78136 4.12228 2.29743 .13 311.63525 678.46265 754.80444 1,606.96704 -1.56 Quantity Sales/Disc. Plant 3.22577 .86830 3.79918 1.21336 -3.09 8.59011 6.53160 14. 17492 21.37105 -1.69 Quantity Salee/Gen. Plant 147.36656 289.20776 40. 3418 16.47037 +2.64 517.32056 629.24146 152. 28873 214.92397 +3.84 Quantity Salaa/Total Pit. 1.16U5 .28609 1.26739 .29017 -2.03 6.11046 3.01789 6.83590 5.52520 .72 3. ( ) PLANT/ TOTAL PLANT Production/Total Plant .34059 .09166 .34050 .06405 + .01 .04592 .05650 .07395 .09613 -1.57 Dletrlbutlon/Total Plant .37306 .09771 .34885 .08303 +1.45 .79489 .17295 .63446 .27863 +3.06 Ganer at Ion/Total Plant .02173 .01644 .03548 .01342 -4.95 .02207 .01894 .05835 .02731 -6.89 4. DISTRIBUTION EXPENSES/ DISTRIBUTION PLANT .04020 .01006 .03940 .01183 + .41 .04722 .01524 .05011 .02144 .69 ADMINISTRATIVE 6 CENERAL EXPENSE/GEN. PLANT 1.39610 3.15008 .42996 .20317 3.21 1.90668 2.68784 .47368 .19300 3.79 A«A: TOTAL UNITED STATES FACE 8 ELECTRIC COH PUT AT IOWS ccraniATioii sr. electric -51 b-75 STANDARD DEVIATION MEAN STANDAW) DEVIATION CAS COMPUTATIOHS COMBINATION ST. GAS n-51 o-16 STANDARD DEVIATION HEM 1. use or funds as percent TOTAL OPERATING REVENUE Grose + Plant /Tot el Operating Revenue Dividend* on Preferred/ Total Operating Revenue .01667 Dividends on Coerion/ Total Operating Revenue .10517 .18632 +1.70 1.96970 2.8*474 .16068 .05551 .82 .06223 .10411 .00523 .06101 +4.48 .00579 +3.91 .02820 + .54 .37064 .54013 .05144 .02274 +4.22 .11963 +1.13 .40678 .68303 .09735 .10902 +3.18

PAGE 137

APPENDIX B: Results of the Statistical Computations The Regional Analysis

PAGE 138

AREA: I FACE 1 1. TOTAL FOPULATION SERVED 2. SQUARE MILES SERVED 3. FOPULATION/SQUARE MILES 4. HUMIEI OT CUSTOMERS: Realdanelal Cowerclal Indua trial 5. REVENUES RECEIVED: (000) Resident ill CoaaMrclal Industrial i. QUANTITY SALES: R. aidant lal Coanerclal Industrial ELECTRIC COM PUTATION S CMIKATIOIT ST. ELECTRIC •-15 n-2l 483.053 83.389 2,35* 95,087 74,033 51,820 3.5U 1,927 4,176 441,831 99,80} 2,451 •5,451 111,390 48,177 2.726 1,929 4 .400 225,170 27,438 1.334 33,429 23,733 20,505 1,423 1,084 1,757 178,911 +2.93 22,954 +2.40 2,036 +1.27 26,739 +3.01 27,099 +1.93 18,316 +2.70 1,208 +3.03 1,301 +1.94 1.694 +2.23 CA S COMPU TATIONS COMEINATIOH ST CAS • -15 STANDARD HKAN DEV 1AT ION 1,499,216 1,519,915 1,727 2,418 299,651 361,893 26,609 17,843 1,109 1,494 40,781 12.963 8,879 258.660 94,324 127,653 41.119 15,710 10,150 262,228 111,058 147,283 MEAN 1,860,833 2.904 STANDARD DEV IATION 267,019 326,519 15,548 16,370 1,280 1,986 38.276 8,660 13,564 301,437 77,930 178,570 37,399 6,297 12,243 373,103 78,208 195, 0u8 + .15 .79 -1.00 .12 + .39 .71 AREA: II FACE 1 Ei*CTJtIj:._cml^ATJOHS COfOUATIOH »T. ELECTRIC CAS COMPUTATIONS COHaiRATlOK s-1 S TARTARS DEVIATION STANDARD PEV 1ATIOM STANDARD STANDARD KEAM DEVI ATION WAN DEVIATION 1. TOTAL FOPULATION SERVED 2. SQUARE MILES SERVED 5 J. FOPULATION/ SQUARE MILES 4. UMBER OF CUSTOMERS: , kasidautlal CooDBrclal Industrial 1. REVENUES RECEIVED: OUT lasldsjntlal ' CoMMerclal Induatrlal ' 6. QUANTITY SALES: Residential Coaaanrcial Industrial 11,201.200 77.502 307 399,!.ii 43.114 2,676 58,160 15,020 42,440 2.590 1.667 1,595 11,118,091 110,054 441 117.187 38. 277 2,(24 44,157 12,452 15.777 2,060 1,496 2,715 15,606,344 79,185 427.726 44,612 2,116 40,279 19,129 54,544 2,743 1,904 3,510 11,741,888 45,846 435 339.287 28.097 1,744 .66 1,199,910 1,539,361 1,260,730 1,654,509 47,456 31,988 43,181 2,026 1.337 4,(82 .54 .07 . 23 4,117 627 316,382 25,963 1,761 33,598 18,578 19,366 .11 533,6(2 .31 234,046 .79 352,116 3,394 652 275.243 19,497 2.449 59,494 21,163 20,189 513,453 240,541 145,263 1,3(2 130 310,644 30,624 1,018 33,684 17,507 32,149 589,315 224,249 636,680 1,718 47 398,309 36,066 1,477 71,001 20,937 44,517 803,385 262,761 831,319 .05 + .13 +1.14 -02 ;2 00 .07 -128-

PAGE 139

-129AREA: III PACE 1 «ltCTUC_C»«miTK«S COWllNATIO* »t. eiectric ••44 »-3 KyAM DEVIATION team STANDARD DrVIATK* CAS. COM IVT AT I OK5 COMBINATION mean STAN1WRD DEVIATION^ ST. CAS »--S MEAN STANDARD DEV IAT IO!f I. TOTAL POPULATION SERVED 1. SQUARE KILE* SERVED 3. POPULATION/SQUARE MILES 4. NUMBER Of CUSTOMERS: Eaaldantlal C(MH«rct«l Indujtrlal 5. REVENUES RECEIVED: (000) Ecaldrntlal Coa«ercUl Industrial ft. QUA AT I TY SALES: ftaaldaatlal CoM^rc lit Inftuitrial 6,762,523 104,888 157 180,51) 24.207 715 26.579 15,379 17,350 7,002,(40 94.202 301 192,358 19.821 97* 1,155 *29 1,4*0 30,545 13.944 14,331 1,244 483 1,7»3 32,340,220 39,654,384 •7,050 43,055 3)55 301 945,274 95,017 387 139,900 124,733 67,053 5.544 5,695 5.937 1,207,144 111.806 I»7 177,220 144, Jul 79,221 4,444 7.J10 7,348 -2.11 + .32 .97 -2.16 -2.09 + .34 -2.11 -2.34 -1.84 -1.14 -2.35 -1.89 428,037 329,060 7,(81 11,028 338 «92 96,817 9,951 405 15,706 4,097 4,585 154,565 79,745 144,603 74.435 7,115 570 12,446 4,184 4,782 115,022 49,148 104,131 2,794,180 3,916 1,516 449,731 34,097 3,932 98,873 24,651 42,659 765,028 341,248 956,246 1,946,061 7,603 1,512 ;-ii,6na 26,816 2,827 100,834 20.918 32,199 641,967 322,009 731,032 -4.14 4 .70 -1.82 -3.35 -2.91 -4.18 -2.84 -3.01 -3.85 -3.22 -2.76 -3.81 1KA: If FACE 1 flJm ic jwnjTATj oh s iikatio* it. lucnic •-9 «-U ST AN DA R0 DKVLATIU* STANUARO DEVIATION CAS CCHIVTATIOWS COHllINATlON .-9 STANDARD H»a JEYiATipi^ MEAN STANDARD DEVIATION 1. TOTAL POPULATION SERVED 2. SQUARE MILES SERVED 1. POPULATION/SQUARE MILES 4. NUMBER OP CUSTOMERS : Eaatdantldl Coaaaarc lal Induatclal 3. REVENUES RECEIVED: (000) U. 374,442 341,9*4 101 415,798 tl,341 1,799 12,026,348 373,850 133 486,191 108.721 1,334 13,328,181 141,100 106 365,692 48,834 1,213 19,958,224 184.841 111 394,124 77.860 1,810 .10 1,462,022 2,386,619 .64 21,206 33,104 .09 488 «39 .17 .29 .59 369,1)1 30,173 1,044 656.055 45,335 1,282 2,443,902 ft, 916 671 106,089 12.916 341 3,798,159 9,043 71* •4,016 8,337 2*1 .»» * .9* .49) .9* * .8* 41.23 (daldaatlal 51,(33 79,*** 31,145 75 jo; + .02 36,045 60,361 16,665 13,233 .72 Cartlil 13.333 94,011 41,2*4 70 4 .31 11 .636 13,877 5,97* 4,087 * .80 Industrial 23,444 29,229 23.500 40 ,279 .11 10,291 42,547 13,955 8,834 .33 OJIAETITT SALES: Raaldcntlal 2,4*8 3,849 2,783 3 190 .19 404,288 712,878 162,402 175,720 .7* Coaaaarc lal 1,887 5,315 1,341 ), 774 + .24 173,235 241,927 58,032 31,740 1.08 Industrial 2.458 3,1*2 1,0*1 4 532 .32 525,765 1,077,191 400,648 292,301 » .2*

PAGE 140

-130AREA: T PACE 1 KLIXTRIC C'HIUTATJONS 1. TOTAL POPULATION SERVED 2. SQUA1C MILES SERVED 3. POIM1.ATIOM/SOIJARE MILES 4. NUMBER Or CUSTOMERS: I«al8cntUl Co«m«rclal LndufCrtal 3. REVENUES RECEIVED: (000) RaaldantUl CoaaMrclal Industrial 6. QUAimn SALES: RastdaaClal Caaaparcial UduatrUl COMBINATION o-6 MEAN 6,210,296 139,765 1,659 151,457 19,117 1,275 26,767 16,617 16,367 1,153 STANDARD DEV IATIOH 3,590,223 103,229 3,973 97,421 11,547 1.580 19,735 12.715 23,199 »39 . 7U J,27» ST. electric g-16 HKAM 10,320,826 259,988 246 263,396 35.336 3,717 46,425 31,438 28,269 2,212 1,700 3,274 STANDARD DEVIATION 4,728,287 167,001 718 1,206,900 162,462 20,839 24,323 16,954 16,795 1,336 1,176 3,066 -1.94 -2.14 -2.48 -1.69 -1.86 -1.16 -1.49 -1.59 .80 CAS COKI^rjATIUNS CCtWlNATIOH ST G* 5 n-* n-5 -1.84 245,737 -1.57 4,703 71,637 7,013 240 6,065 1,692 3,747 81,865 37,854 142,114 STANDARD DEVIATION 194,659 5,869 3,340 53,037 4,783 199 3.457 1.448 4,021 56,080 33,746 166,977 mi; an 380,119 36,655 1,915 42.180 10,788 28,699 513,256 157,614 1.033,481 STANDARD DEVIATION 1,686,780 2,018,183 34,476 64,308 43 10 407.582 31,513 2,875 44,396 8,907 29,044 546,247 124,170 ,148,157 -1.57 -1.02 .93 -1.66 -2.06 -1.29 -1.74 -2.05 -1.70 RJ.KCTlUC_COM PJUTATJONS CO 1» IN Aril* 8T. ELECTRIC 0-2 a-10 COMIUIWTIOH «T. CAS WEAK STANDARD DEVIATION MKAM STANDARD DEVIATION MEAN STANDARD DEVIATION HEAP) STANDARD DEVIATTOH .iTOTAL POPULATION Sent* 4,525,500 5,946,060 20,364,288 14,716,834 -1.33 562,300 429,497 967,942 1.098,504 .41 J. SQUARE MILES SERVES 61,640 82,647 225,110 182,682 -1.13 6,059 8,412 7,617 9,957 .19 J. rOtOLATirW/ SQUARE MILES J7 20 117 84 .47 1,212 1,611 33* 441 .93 4. mam* or ccsroxtisi aataWUI CaaMaarelal ladtl«tltal 119,680 19,000 49S 155,196 23,193 659 491,354 64,192 2,131 369,132 47,183 2,066 -1.25 -1.23 • 1.01 67,494 5,999 136 64,773 6,188 190 243,710 19,320 1,022 22V, 395 17.759 1,976 .87 • .85 .52 i. RETENUSS RECEIVED: (000) J-,t 1.1 Coaaerctal tn4u aerial 22,200 13,050 10,300 29,274 16,900 13,576 83,080 51,830 43,520 66,313 43,259 38,846 -1.16 -1.14 -1.09 7,456 3,396 4,944 7,453 2,602 5,745 29,319 11,659 24,131 30,233 9,746 23,114 .13 -1.07 . .93 6. QrUKTITY SALES: Residential Coaaaerolal Indnsertal 1.161 759 1,245 1,541 1,009 1,700 4,634 2,891 4,996 3,667 2,397 4,886 -1.20 -1.13 .98 51,196 39,094 109,946 61,301 35,356 125,989 277,345 179,535 648,520 307,601 133,663 643,623 .94 -1.19 .96

PAGE 141

-131AREA: I PACE 2 mSffiC COMPUTATIONS COMBINATION n-15 EXPENSES: Operating Maintenance Depreciation DltCrlbutlon CultOMIl Salt* Attain, and General Total Operating ST. ELECTRIC n-2l STANDARD STANDARD EAR DEVIATION MEAN DEVIATION 1,072,090 1,044,969 376,612 310,670 211,882 231,993 67,065 64,153 262,697 253,491 99,549 92,401 184, B98 178,654 62,027 56,647 79,682 104,245 22,756 18,484 24,871 21,547 15.626 14,437 166,418 163,700 52,975 47,580 1,969,974 1,996,214 653,186 551,704 +2.80 +2.64 +2.64 +2.87 2.38 + 1.50 +2.93 +2.79 GAS CCV.a nATIO^S COMBINATION 43 37,965 3,803 3,889 7,814 2,554 1,295 4,212 55,807 34,999 4,665 4,886 8,593 3,093 1,246 4,138 55.544 ST. GAS 0-10 55,004 2,545 5,071 5,705 2,874 1,532 5,054 73,866 STANDARD DE VIATION 52.358 2,519 5.560 5,704 2.920 1,003 4,129 66,819 .93 + .57 -2.13 + .65 .25 .48 .48 .73 2. PLANT : Production Distribution General 4,070,268 4,484,400 Wl.703 4,102,532 5,306,896 154,130 1,419,999 1,508,956 127,105 1,402,702 1,365,358 118,075 +2.67 +2.39 .31 11,928 132,148 2,413 19.541 137,033 3,273 23,518 95,579 9,151 32.961 86,168 7,340 .1.06 + .72 -3.34 3. TOTAL urmiST 296,302 2.73 29,630 5,026 4. TAXES: Stat, end Local 369,865 54,353 533,023 75.159 85,567 24,977 69,312 32.013 2.33 1.55 7.213 2,479 6.905 3.104 6,835 3,218 5,730 4.336 + .11 .48 5. AC/810 i. ITO/M 3.910 15,688 10,47) 4,897 +1.04 AREA : II FACE I EXPENSES: Operating Maintenance Depreciation Production Dlatrlbutlon Cuatoaere Salaa Administrative & General Total Operating ELECTRIC CCHPUTATIOHS CCKMRAncH •-5 628.786 99,264 161,202 461,672 97.956 39,704 22,538 90,338 ,162,624 STANDARD DEVIATION 560,926 85,941 122,322 383,369 82,371 39,462 23,044 103,313 948,693 . ELECTRIC •-14 728,416 138,351 190,303 557,085 108,883 37,928 36,587 98.366 1,325,440 STANDARD DEVIATION 638,358 116,668 125,137 483,006 116.396 30,529 20,745 99,007 1,069,449 CAS COMPU TATIONS COMBINATION -5 STANDARD STANDARD t KEAN DEVIATION KEAN DEVIATION t .29 63.948 61.604 73,503 100,050 .16 .65 3.417 3,229 2,715 3,254 .28 .42 5,532 7,561 5,904 8,553 .06 .38 49,428 43,523 56.152 76,016 .15 .18 7.293 7,828 7,300 9,562 .00 + .10 3,044 3,361 3,586 4,687 .18 -1.19 1.109 1,455 1,863 2,396 .51 .15 5,071 7,035 6.539 9,542 .23 .29 84,610 •7,357 92.447 123,331 .10 PLANT: Production Distribution General 2,297,842 2,209,401 145,854 1,434,455 1,836,921 149,895 3,061,713 2,195,416 307,396 2,038,573 1,501,107 259,051 .73 + .02 -1.23 7.006 153.181 6.576 8,149 172,969 10,953 3,957 149,814 12.672 6,172 203.169 16,482 .55 .03 .59 3. TOTAL TjmiZST 179,426 7,920 +1.15 4. TAXES: State and Local Lneoaa 113.870 133.122 92,536 86,223 150,683 97,265 122,532 88,850 .58 .74 4,931 5,146 4,890 7,976 4,407 3,873 5,790 3.006 + .13 + .27 ). AC/1TH 6. BTLVinm 10,944

PAGE 142

-132AREA: HI PAGE 2 ELECTA ^COM PUTATIONS COMBINATION n-14 STANDARD DEVIATION ST. ELECTRIC n-3 STANDARD DEVIATION CAS CmPTJTATIOWS STANDARD DEVIATION EXPENSES : Or.r.tln, 260,354 272,511 Kaint.n.nc. 45,061 47,471 748,433 Depreciation 62,763 78., 63 430,633 Fraction 169.219 196.104 1,097,610 Distribution 45.881 49.673 237,637 Customs 15.763 15,489 88,203 $.1.. 10,331 15.121 42.963 Ad.inl.tr.rW. 1 Cn.r.1 36,621 38,783 2 "'^ Total Op.r.tln, 541.189 556,634 2,991.900 1.735,904 279,713 559,501 1,273,486 261,784 117,165 47,994 271,079 3,713,150 -2.32 -2.40 -2.09 -2.38 -2.38 -2.10 -1.98 -2.24 -2.21 19,980 819 1,663 16,279 1,992 787 374 1,271 26,587 12,151 122,395 80.029 -4.39 665 4,987 2,720 -3.02 1,432 15,891 14,860 -3.35 9,627 90,462 59,998 -4.23 1,358 9,306 3,673 -5.99 617 6,558 2,766 -4.52 338 3,253 1,972 -4.97 932 10,513 7,097 -4.50 17,884 171,866 113,296 -4.39 plant: Production Distribution General TOTAL DrTESEST 1,070,505 1,102,319 6,277,696 7,705,757 1,054,327 1,056,533 5,513,323 6,694,553 81,697 71,347 117,212 74,651 413,853 370,490 472,898 439,632 -2.27 -2.22 -2.18 1,734 49,901 1,461 2,790 39,878 1,393 7,463 34,996 243,649 23,433 62,290 216,185 15,104 -3.03 -5.09 4. TAJOES: St.t. and Local Xacora 77,357 M.318 91,621 38,444 486,523 274,673 671,412 348,395 -2.05 -2.05 2.012 1,767 2,000 1,883 13,447 12,759 9,187 10,318 -1.76 -3.61 S. AC/ITU i. RU/CUB 1,155 AREA: 17 •ACS 2 EXPENSES: Operating Halncananca Depreciation Production Distribution CuetoMr. Sale. Admlni.tr. tlv. and Central Total Operating PLANT: Production Distribution Genera i gLBCTRlC CO MP UTATIONS COMBINATION HtAM 471,637 79,396 175,346 299,862 100,936 39,544 14,291 75,306 1,017,637 2,681,345 2,673,486 72,679 STANT-AJID DEVIATION 722,367 131,519 293,356 452,307 183,641 66,374 18,868 103,621 1,590,338 4,502,106 4,277.293 31,736 ST. ELECTRIC •-11 CAS CCmVTATIOKS COMBINATION 2.276,288 2,389,184 189,196 3.368,096 3,535,388 194,297 .22 + .15 -1.66 8,919 127,356 1,679 13,331 207,703 1,809 ST. GAS n-6 STANDARD STANDARD DEVIATION t MEAN DEVIATION: MEAN 485,569 765,967 . '34 58,586 113.145 52.784 80,900 157,950 .02 2,459 4,315 1,044 142,065 242,595 .36 4,935 8,695 3.516 302,132 483,836 .01 46,505 93,362 45,384 77,092 134,512 + .32 5,497 8,954 2,483 36,167 52,695 .12 3,010 5,295 1,354 21,916 25,999 .70 810 1,290 1,002 90,171 149,676 .24 3,726 5,437 2,378 927,629 1,493.422 + .12 73,042 135,575 62,539 11,402 84,672 7,450 STANDARD DEVIATION 59,855 .11 655 .74 1,579 .37 59,546 .02 2,074 .75 992 .70 780 .10 1,019 .56 59,876 * .17 18,912 .23 54,770 .46 1. TOTAL INTERJSST 308,669 213,838 + .35 19,491 4. TAXES: State and Local 159,609 110,129 239,681 211,529 151,827 79,529 218,485 115,132 .24 .04 5,259 1,461 8,684 1,435 3,594 1,279 1,545 1,040 .43 + .25 3. tcttm t. 87UTCWH 1,663 5,902 1,664 5,613 4,971

PAGE 143

-133AREA: V PACE 2 EXPENSES: Operating Maintenance Depreciation Production Distribution Cuatoatere Sales Adalnietrative and Ceneral Total Opereting PLANT: Production Diatrlbutlon Ceneral ELECTRIC COMPITATIONS COMBINATION ST. ELECTRIC STANDARD liyMTJ*>J CAS COMPUTATIONS COMBINATION ST. CAS n-5 254,430 38,177 8. 1 , 455 184,373 40,037 15,300 9,692 34,075 518,638 1,201,425 1,047.833 81,925 3. TOTAL INTEREST ». TAXES: Stat* and Local bcoaa 3. AC/BTl! «. 1TU/RVH 35.423 68,037 2,399 11.174 183.135 32,156 77,471 129,311 29,443 9,823 10,121 29,112 400,374 37,419 48,158 432,768 72,392 128,907 290,863 68,812 30,338 27.491 73,844 903,278 ,284,269 1.772.565 801,909 1,877,057 84,663 96,677 153,766 1.181 10,670 DEV IATION t MEAN DEVIATION 202,968 -1.80 8,857 5,633 46,129 -1.59 537 448 59,290 -1.43 863 580 140.149 -1.55 7,315 5,679 38,905 -1.57 1,026 751 15,315 -2.13 506 356 14,531 -2.63 114 135 43.205 -1.99 659 452 434,075 -1.80 11,795 7,517 985,064 -1.06 1,106 1,692 825.349 -2.02 19.513 13,010 127,862 -1.92 1,130 1,547 43,610 .40 7,934 1,466 48.335 -1.80 7H MO 97,838 • 1.96 671 580 242 1.67 59.379 2,937 4,928 44,580 5,986 3,520 1,555 4,417 80,022 20,744 103,605 9,560 3,439 6,669 STANDARD DEVIATION 53,096 2,528 4,556 40,323 5,941 3,671 1,589 4,215 71,667 22.348 107.519 9,111 5,771 7.995 -2.09 -2.07 -1.96 -2.02 -1.83 -1.81 -2.00 -1.96 -2.10 -1.94 -1.72 -2.02 -1.78 -1.67 AREA: fl PACJ 2 ELECTRIC COMPUTATION S COMBINATION a-2 MEAN STANDARD DEVIATION ST. ELECTRIC -10 STANTjARD DEVIATION CAS CO MPUTATIONS COMBINATION a-2 STANDARD DEVIATION STANDARD DEVIATION EXPENSES : Op.catin, 239,905 305.817 901,610 739,723 Maintenance 27,413 36,861 122,499 96,844 Depreciation 58.225 79.783 230,877 168,050 Production 203.380 262,026 690,983 591,268 Diatrlbutlon 17,440 21,482 111,276 119,316 Cuatoeurrs 11,500 14,185 45,011 32,575 S.l„ 5,250 6.237 35,736 28,903 Administrative and Canaral 26.300 33,927 113,915 101,149 Total Operatic* 400,550 522.410 1,377.466 1,201.449 -1.14 -1.25 -1.30 -1.43 -1.01 -1.30 -1.36 -1.11 -1.24 10.192 370 1,119 8,021 706 5 50 401 700 13,678 9,864 346 1,279 3,598 393 354 its 434 13,349 55,200 1,821 5,940 42,802 4,377 2,460 1,377 4,089 69,088 47,229 1,420 3,813 38,672 4,533 2,439 1.341 2,680 53,201 -1.09 -1.16 .95 -1.03 .93 .91 .84 -1.45 -1.18 PLANT: Produc t loo Diatrlbutlon Canaral 1,266,295 817,615 60,940 1,773.615 1,077,977 82,010 3,154.457 3,079,889 284,428 2,107,889 2,322,369 201,138 -1.10 -1.23 -1.41 1,243 31,611 930 741 33.795 381 4,627 87.205 7,774 5,358 78,740 5,070 .71 .7S 3. TOTAL INTEREST 110.599 169,290 8,219 4. TAX2S : Scata and Local Income 42,540 14,140 58,435 16,362 151,482 136,736 117,038 120,761 -1.17 -1.31 '58 733 8«4 2,424 2,337 1,273 1,569 -1.40 .96 3. •C/ITU 6. 4TU/XUH 3.634 12,757 2.215 3.356 3,2»7 10.439

PAGE 144

-134ARM: I PACE 3 gJ^RlC^COWTATIONS COMMKATION n-15 ST. ELECTRIC n-21 CAS CCHt UTATIONS COMBINATION ST. CAS •15 -!<> STANDARD STANDARD STANDARD STANDARD MEAN deviation MEAN DEVIATION MEAN DEVIATION MEAN DEVIATION t employment UMBE1 07 EMPLOYEES 6,525 4,595 1,820 1,867 .11 1,267 1,336 1,442 1,21* .32 TOTAL UACES 57,872 60,619 18,677 20,278 +3.30 13,823 14,503 14,376 12,980 .09 FENS ION BENEFITS 10,751 13.0?) 2,264 2,154 1.30 10,560 12,624 1,633 1,827 2.13 FINANCIAL SOURCE OF FUNDS: OUTSIDE •5,430 68,1)4 21,918 23,213 3.36 61,068 67,794 14,347 20,220 2.03 INSIDE 123.116 125,588 44.294 58,372 2.96 114,909 125,125 17,604 19,266 2.34 APPLICATION OF FUNCS CROSS '+ TO FLANT 139,169 141,405 43,409 33,867 3.01 126,937 140,854 16,339 21,465 2.41 DIVIDENDS OH PREFERRED 3,409 9,132 1,959 2,420 1.11 3,048 8,910 319 EM 1.61 DIVIDENDS ON COM40N 24,568 22,770 7,981 7,678 .28 22,949 22,842 4,776 4,383 2.38 TOTAL APPLICATION OF FUNDS 190.391 184.953 66,685 78,671 2.74 177.698 184,880 33,486 38,657 2.33 AREA . XX FACE J ELE CTRIC COHfUTATIOWS 9*5 OOMWTATIOWS COMBINATION ST. ELECTRIC COMBINATION ST. CAS .-14 -3 MEAN STANDARD DEVIATION MEAN STANDARD DEVIATION _L_ MEAN STANDARD DEVIATION MEAN STANDARD DEVIATION t EMPLOYMENT NUMBER OF EMPLOYEES 2.913 1,608 3,930 3,216 .2* 1,281 1,924 1,660 2.122 .23 TOTAL UACES 21,010 11,702 42,301 39,723 .83 3,802 3.217 13,190 20.974 .83 PENSION BENEFITS 2,801 1,730 4,466 1.980 .04 2.241 1,966 2.069 1.978 .09 FINANCIAL SOURCE OF FUNDS: OUTSIDE 36,411 18,599 30,333 39,421 • .10 29,128 22,904 14,312 17.2*1 .95 WIDE 48,293 43,683 100,078 100,204 .41 38,636 43,362 13,7*5 23,859 .91 APPLICATION OF FUNDS CROSS TO PLANT 37,290 39,514 101,173 101,229 .2* 43,832 42.749 14,292 18,602 1.21 DIVIDENDS ON PREFERRED 1,710 1,187 2.139 2,264 .00 1.366 1,281 44 94 1.62 DIVIDENDS OH COMMON 13,499 8,311 18.074 16,063 .02 10,759 9,380 3,406 7,107 .83 TOTAL APPLICATION OF FUNDS 84,703 60,923 150,430 137.059 .23 67,764 64,951 28.037 40,800 .94

PAGE 145

-135ASIA: HI tT.Lf.TK I C CCW PUTAT I ONS CCKblHATIOH ST. ELECTRIC n-U n-3 CAS COHIVI'ATIOWS COHBINATIQH ST. CAS n-5 MEAN STANDARD DEVIATION MEAN STANDARD DEVIATION WEAN STANDARD DEVIATION WAN STANDARD DEVIATION t emiloymlht HUKSf.R OF EMPLOYEES 1,728 1.601 6,234 7,124 -2.01 340 29 7 2.548 1,101 -6.31 TOTAL WAGES 13,181 14,816 56,984 98,700 -1.48 3.994 3.234 26,107 12,05* -5.41 PENSION IBDIB 1,107 692 8,628 14,945 -1.7» 1,017 692 3,13* 1,036 -3.20 FINANCIAL SOCLCE OF FUNDS: OUTSIDE 18,783 17.006 •5,879 148,746 -1.5* 18,783 17,006 39.433 3*. 599 -1.63 INSIDE 76,243 31,263 121,190 204.908 -1.51 26,263 31,265 52,988 85,96* .9* APPLICATION OF FUliDS CROSS + TO PLANT 35, 99* 41,549 166,059 287,622 • 1.32 35,994 41,549 37,131 35.478 .09 DIVIDENDS OK PREFERRED 1,313 1,978 3,477 6,023 -1.03 1,315 1,978 l.»3» 2.181 .11 DIVIDEND OH COMMON 7,478 9,603 30,987 53,670 • 1.44 7,478 9,603 12.71* 10.409 .97 TOTAL APPLICATIOH OF FUNDS 48,808 34,088 207,069 358,654 -1.47 48,808 54,088 93,137 117.5*1 -1.0* AREA: IV FACE 3 ELECTRIC C ONFUTATIONS COMBINATION • -9 ST. ELECTRIC ••11 C*5 COMPUTATIONS COMBINATION j-9 WEAK STANDARD DEVIATION MEAN STANDARD DEVIATION MEAN STANDARD DEVIATION MEAN STANDARD DEVIATION t EMPLOYMENT NUMBER OP EK7TOYEES ». 420 7,337 2,512 4,005 .70 1,33* 2,8*0 J 79 437 .51 TOTAL WAGES 31,620 52.971 28,093 48,662 .13 14,961 17.178 8,33* *,5*3 .53 PERSIC* BENEFITS 4,410 7.377 4,431 (.6*7 .00 4,430 7,377 • 22 Ml +1.11 FINANCIAL SOURCE OF FUNDS: OUTSIDE 57.9*5 •1.111 17,610 34,483 4-1.05 57,9*5 82,111 10,30* 9,239 1.32 INSIDE 6*. 189 96,723 42,241 66,926 » .93 6*. 189 96,723 8,600 • ,237 1.31 APPLICATION OF FUNDS CROSS + TO PLANT 7*, 047 123,639 77,273 1*8, 197 .02 76,0*7 123,639 9,8*3 3,39* 1.22 DIVIDENDS OH PREFERRED 3,336 5,795 39.01S 107,436 .9* 3,336 5,793 531 371 1.16 DIVIDENDS OH COMMON 17,260 27,6*5 6.317 3,375 41.22 17.260 27,6*5 2,682 1,621 1.20 TOTAL APPLICATIOH OF FUNDS 122, 145 176,660 85.788 145,600 4.48 122,1*3 176,6*0 19,22* 9.831 1.33

PAGE 146

-136AREA: V FACE 3 gun kic ^ccKPyrAii ons cohbi nation B-6 ST. ELECTRIC •-It GA5_ COKft'TATIOtft CO2WKATI0N G* 5 man STANDARD DEVIATION KEAH STANDARD DEVIATION t KLAN STANIiAKD DEVI AT I CM BEAN STANDARD DEVIATION MPLOYMEKT UMBER OF EKFLOTEES 1.(10 1.142 2,219 1,147 178 1,94 1,970 1 92 TOTAL MACKS 9.750 7,202 13,5*6 12,378 -1.00 3,011 2,742 13.985 13,382 -t 77 PENSION BFJIEFTTS 1,093 1.W1 1.472 1,202 .59 1,09) 1,461 1,626 1,691 31 FINANCIAL SCORCI Of FOTDS: 61 OUTSIDE 25,845 24.6M 2»,4«0 23,185 .22 25,845 24.6CA 14.514 15,614 *! nam 2t,t53 34,174 22.990 18.398 .31 26,653 34.174 18,053 16,813 .46 APPLICATION Or FUNDS cross to plant 35,903 35,460 37,912 32,355 .11 35.903 39,460 13,152 11.230 • .1.1 DIVIDENDS OM PREFERRED 1.395 1,344 1,159 Ml * .43 1,395 1,364 341 * 6DIVIDENDS OH COMMON 7,670 7, S0« 9,359 4,901 .47 7,670 7,806 3,488 6,795 .44 TOTAL APPLICATION Of rUHM 52.498 57,535 31.985 40,279 • .02 52,498 57.535 32,623 31,091 61 AREA: VI face > pjCraiC COMPUTAT IONS COMBINATION 14-1 ST. ELECTRIC •-10 CAS COH1V TAT IO NS COKE ISA 71 OM •-2 ST. CAS STANDARD STANDARD STANDARD STANDARD BAN DEVIATION MAN DEVIATION t WEAN DEVIATION MEAN DEVIATION I. tMPLOYMENT i. c>i>t UMBEI Of EKTLOYZES 1,471 1.684 3.9(1 3.181 .91 228 40 1.263 928 -1.29 In . » TOTAL KACeS. -*,m 12,924 39,082 32.091 -1.18 2.140 1,022 10,195 7,014 -1.32 nxsicm torerm 1,484 2,182 3,360 3,882 .70 1,684 2,182 1,143 691 " *.M 2. rauMcuu .0?*.*. . sSOt flCZ OF rUHQB: omxat 14,353 21,292 58,840 44.060 -1.21 16,553 21,292 8,354 6,428 • .58 2 ' IMS CDS ' 42.390 59,825 110,496 110,371 .7* 42,390 59,825 14,087 15,357 73 APPLICATION OF rUMDS C10SS + TO PLAHT 44,882 472.187 137.11) 123,770 .94 44,882 62,187 11,502 9,154 '.If DIVIDENDS OM FUEFtllKED 1.064 1,409 3,579 3,605 .89 1,064 1,409 360 267 » .79 DIV 1DEHDS OH COW-ON t.180 t,2tl 18,996 13,361 -1.19 6,180 8,261 2.469 1,404 . 71 TOTAL APPLICATION OF FUNDS 38,942 81,116 169.913 144,309 .96 58,942 81, lit 25,835 25,229 .61

PAGE 147

-137CM COHPITTATIOHS COMBINATION -15 STANDARD MEAN DEVIATIO N ST. ELECTS IC -21 STANDARD DPTIAHOM COMBINATION -15 STANDARD DEVKTI.IK ST. CAS -10 STANDARD DEVIATION 1. DOLLAR UTOBI/QUAHTlTr R*sid«icl*l Ct in rclal Industrial .26957 . 16313 .04564 .06449 .04228 .24218 .23688 .12789 .03021 .03871 .02328 +2.16 +1.47 +1.47 .15389 .12858 .07048 .0341* .02900 .03003 .14540 .17644 .09753 .04226 .13740 .08547 .72 -1.30 -1.08 1. quAimTT/no. cusTaoats tttlaaotlal -05579 .01016 .04220 Co-.~lal .30519 .09996 .33654 I»« u .trl.l 20.18433 14.11258 33.95886 .01172 .20826 41.6154] 1.66 .52 1.20 1.04326 4.34760 326.33594 .53640 1.961M 549.425/8 1.08446 5.03400 331.60367 .48233 .15 2.75068 .70 298.79150 .03 3. DOLLAR UTOnjI/lK). I taaldaatlal CiiMircUl Mm trial .01468 .07578 J.7111J .00148 .01926 1.70352 .01479 .07578 4.42491 .00165 .03914 5.43981 .16677 .53953 16.18246 .04722 .18594 18.51042 .16352 .66659 42.28601 .07940 .15628 83.27116 .19 -1.71 1.12 AREA: II net 4 EUCTtlC COHfUTATtOHS ccKtiatnon STANDARD HTYIATIOH NT. ELECTRIC bpU standard pptiation CAS C0MHJTATL0H3 COMBINATION ST. CAS STANDARD DEVIATION STANDARD DEVIATION I. dollar mDn rvocumrr MtatUI .««3 .01298 .21441 CMtltl .20974 .02488 . 20312 Uduatrlal ."**J • 10761 .02748 .02417 .01849 .71 + .49 + .77 .08733 .0734] .05277 .01374 .00808 .00371 .09385 .07530 .04743 .01213 .00813 .00334 .43 .30 +1.41 OTUITITT/MO. CUSTOMERS aaiaattlal tnicUl .06704 .38197 19.72337 .10480 10.14619 .08388 .39*19 29.90436 .01099 .14280 27.14107 + .22 1.69678 . .23 8.09206 • .77 341.00122 .1110] 2.03650 143.07242 1.75784 .l«5»l 7.36383 1.332)1 717.13698 120.29124 .S* + .39 2.41 3, DOLLAR RI7DTOS7H0. CUSTOMERS U.ldo.tl.1 .01301 .00U1 .01191 .00144 +1.40 taarel.l .07812 .015*8 .0796* .01574 .13 Ikdu.trUl 1.20812 1.03518 1.03798 2.811U .»! .1*94* .40419 18.1*381 .03198 .11999 10.91107 .14544 .57263 34.44924 .01923 . 12*17 4.9113* .48 .21 1.45

PAGE 148

-138ARFA: III PACK 4 . EUgCT tUC C OHrjTA TIONS COMBINATION ST. ELECTRIC n-U Q-3 CAS COMPUTATIONS DEVIATION beas STANDARD DEVIATION COMBINATION a-14 STANDARD DEVIATION 1. DOLLAR REVENUE/QUANTITY S.fldmtl.1 -23391 .06574 . 24791 .02766 .34 .10172 SSiElll -23748 .0704* .24155 .02760 .09 . 076*4 Indu.trl.l .12115 .03469 .11820 .01000 .14 .04606 .01169 .01330 .00986 .11832 .08154 .04668 .01961 -2.12 .02105 .59 .00943 .04 2. qUAKTlTY/XO. CUSTOMERS Ualdantlal .06277 . 00562 .05834 .00533 +1.17 1.59847 .11590 Co«, e rcUl .28450 .10616 .45656 .18616 -2.07 8.69)66 2.48064 Iodo.trl.l 40.29341 62.45462 106.78247 98.05753 -1.40 701.33081 583.8S013 1.6S262 9.19622 280.59033 .19501 2.59853 171.45763 -1.08 .36 +1.50 DOLLAR REVENUE/NO. COSTOKSIS •aalaentlal .01455 .00391 .01436 .00038 .08 .16170 .01281 r , , | 1 1 • .06609 .023aj .10753 . 03385 2.36 . 65443 . 20436 Industrial 4.82492 7.43565 12.23102 10.42024 -1.36 30.35159 25.42795 .19871 .71531 12.11397 .03934 .14706 5.60884 -2.92 .57 1.50 ARM: r» RAGS » EJECTEIC COMPU TATIONS COMBINATION IT. ELECTRIC ••11 CAS COMPUTATIONS COMBINATION ST. CAS n-6 ms. STAjmAJD DEVIATION STANDARD DEVIATION t mail STANDARD DEVIATION 1. DOLLAR REV EKUE/ QUASI ITT MltatUl .21250 .044*5 .U931 .05854 .92 .11090 .0277* .1267* .02893 .97 Ca—rclal 19613 .02932 .1*431 .0558* .5* .07676 .01795 .10513 .02666 -2.25 latutrl*! .10149 .02210 .0*719 .02*78 l.li .0*559 . 0153* .03762 . 00620 rt.U I. ODANTITT/HO. CUSTOMERS aaldantlal .07086 .02219 .09011 .04090 -1.20 1.12726 .38727 bmrclal .43386 .1093* .501*5 .19338 .75 6.3737* 1.58474 biatrial 75.64406 93.62437 136.0*321 1*1.06734 .86 691.84058 509.27783 1.33003 .40305 ». 90906 1.76335 1,233.03516 356.03271 .91 1.56 -2.10 1. DOLLAR REVENUE/HO. CUSTOMERS R..ld.ntlal .01432 .00232 .01504 .00251 .62 .11973 .03380 .15895 .00691 Cmerclal .0*549 . 0193* .0?601 .02240 .05 . 48693 .15956 . 51154 .16986 Jjuluatrlal 7.96100 10.02073 9.4*258 11.31130 -.30 27.25491 20.33*44 46.25195 15.04868

PAGE 149

-139AREA: 1 FACE 4 ELECTRIC COMPUTATIONS OAS COMPUTATIONS COMBINATION STANDARD DEVIATIOH ST. ELECTRIC »-16 COMBINATION STANDARD DEVIATION STANDARD DEVIATIOH STANDARD DEVIATIOH 1. DOLLAR REVENUE/QUANTITY k.ldantlal 25328 .05704 .22110 .03056 .08069 .01529 ^" 22372 .04542 .19903 .02648 +1.50 . 05866 . 01944 Indu.trlal -09849 .02111 .09838 .01726 + .01 .03156 .00835 .08312 .06527 .03035 .00643 .00834 .00300 .30 .64 .28 1. QUAjrrm/iio. customers Uslda-tinl .06944 .01495 .07973 .02247 .99 1.21027 .34014 1.38363 £_.„.., .38087 .17414 . 47349 . 26243 .76 4.92701 1.58916 4.82328 taiuatrUl 1J.04014 6.24070 1 9.82133 4 9.19347 .23 646.21313 877.81030 1,266.16553 . 29598 1.68735 .252.62207 .31 .87 ). DOLLAR REVENUE/NO. CUSTOMERS laaldaotlal -01713 .00290 .01714 .00290 Co«n«rclal .07899 .02312 .08916 .03400 Indu.trlal 1.50620 .79695 1.51424 3.10221 .09790 .27565 17.56253 .03346 .08221 20.98018 .11372 .30730 39.29956 .01745 .09264 39.68398 .86 .54 1.05 UFA: »I raof. 4 1U 8CTRIC COMPU TATIONS COM ISA TT OH • -2 STAnuARD D EVIATION ST. ELECTRIC -10 STANDARD DEVIA TION cas c omputati ons combination it. cas n-2 a-4 STANDARD DEVIATION nam STANDARD DEVIATION 1. DOLLAR REVENUE/ QCANTI TT Residential CsaMTclal Industrial .20714 .12206 .01463 .05274 .05761 .18192 .18731 .09240 .01613 .02516 .01524 +1.41 + .74 +1.28 .20674 .09605 .04112 .10199 .02031 .00513 .13941 .06443 .03571 .07581 .74 .01481 +1.76 .00629 * .69 J. QUANTITY/NO. CUSTOMERS bilaantlal .08898 .01353 .09455 .00868 Casaaerclal .35458 .10370 .44805 .08652 Industrial 19.91145 7.88864 53.81935 65.38527 .92 -1.23 .59788 7.74572 .67 10,798.57813 .33443 2.13131 14,228.57813 .92440 10.00518 8,631.97656 .42178 .78 1.72149 .84 6,215.84766 + .21 3. DOLLAR REVENUE/NO. CUSTOMERS Residential .01778 .00142 .01761 .00254 .08 .10655 .00616 .10448 .02557 .01 r | il .07071 .00277 . 08440 . 02127 .83 . 76565 . 34204 .45524 . 24991 .34 Industrial 2.20312 .18414 4.54674 3.06240 .40 407.49243 529.43330 304.75684 223.87077 .27

PAGE 150

-140FACE 5 BJEM S ! [ y M P J -I A Tl , PHS C AS i CCT1 WTAT10HS COMBINATION ST. ELECTRIC COMBINATION ST. CAS -03 o 15 n-10 gEAN 3TAN0ARD DEVIATION MEAN STANOARO DEVIATION t MEAN STA-'IDAUD D~VlMK>'1 MEAN STANDARD DEVIATION t COSTS/qUAJfTITY TOTAL OPERATING REVENUE .22831 .04401 .21174 .04771 1.03 .12263 .03140 .14843 .08287 -1.05 TOTAL OPEHATLNC EXPENSE .13069 .03847 .16817 .04352 .87 .10502 .02816 .12841 .07131 -1.10 OPERATING EXPENSE .10004 .02280 .10355 .03335 .34 .07382 .01904 .09459 .05663 -1.26 MAINTENANCE .0173* .00392 .01465 .00442 +1.85 .00694 .00489 .00464 .00168 +1.36 DEPRECIAT ION .02371 .00372 .02195 .00604 + .97 .00710 .00256 .00843 .00715 .63 nouxnicm .07 2 26 .01721 .07528 .02532 .39 .05046 .01033 .06673 .05234 -1.12 DISTRLKITIOS .015*1 .00408 .01526 .00491 .23 .01438 .00771 .01072 .00615 +1.21 CUSTOMERS .00679 .00244 .00601 .00190 +1.05 .00452 .00339 .00549 .00322 .69 UUI .00291 .00164 .00373 .00129 -1.63 .00266 .00227 .00377 .00243 • 1.12 AEKINISTRATIVe AND GENERAL .01339 .00373 .01362 .00455 +1.34 .00814 .00345 .00895 .00359 .34 ELECTRIC COMPUTATIONS combination •4 ST. EI. LO Hid .-14 OAs connjT ATions CU1BIHATIOU IT. CAS BEAM STANDARD DEVIATION JEAN STANDARD PLVIATION MEAN STANDARD DEVIATION MEAN STANDARD DEVIATION C COSTS /QUANTITY IDEAL OPERATING REVENUE .17470 .01187 .17067 .02356 + .32 .07216 .00610 .07075 .00791 .23 TOTAL OPERATING EXPENSE .13482 .01169 .1293* .02661 .42 .06436 .00604 .06223 .00348 + .48 ORUTIK EXPENSE .07053 .01140 .06626 .02177 + .11 .04978 .00575 .04855 .00275 .33 AnrrraANCE .01101 .00176 .01332 .00263 -1.17 .00283 .00061 .00204 .00078 +1.31 0EPRXCIATXON .01927 .00089 .02004 .00312 .11 .00365 .00143 .00351 .00061 + .16 PSOOUCTUN .05369 .00629 .05026 .01846 .39 .04020 .00506 .03730 .00231 + .91 DISTRIBUTION .01070 .00226 .01034 .00324 + .10 .00543 .00078 .00501 .00066 + .76 CUSTOMERS .00407 .00132 .00426 .00171 .21 .00218 .00053 .00249 .00011 -1.01 SALES .00226 .00144 .00393 .00101 -2.26 .00069 .00033 .00133 .00017 •3.11 BHUSISTKATIVE AND GENERAL .00921 .00333 .01008 .00393 .42 .00320 .00135 .00384 .00063 .82

PAGE 151

-141ARKA: HI tKCT. i PJCTTtlC COHfVTATIOtO CAS CCHIUTAT'OfIS COMBINATION ST. ELECTRIC COMBINATION ST. CAS n-14 n-3 »-5 STANDARD STANDARD STANDARD STANDARD WEAK DEVIATION WAN DEVIATION MEAN DEVIATION "FAN DEVIATION t cosTS/guAKim TOTAL OPERATING REVENUE .21476 .02421 .21241 .01621 * .IS .07234 .01096 .07293 .02024 .08 TOTAL OPERATING EXPENSE .17033 .02231 .17108 .01270 .03 .06428 .00831 .06124 .01802 6 .16 OPERATING EXPENSE .08653 .02300 .08907 .01264 .16 .04944 .00515 .04523 .01289 + .96 aintenance .01341 .00239 .01488 .00340 .64 .00192 .00057 .00211 .00117 .45 DEPRECIATION .02558 .00434 .02432 .00378 .33 .00389 .00156 .00497 .00140 -1.29 production .06332 .02177 .06449 .01126 .07 .04082 .00481 .03326 .00953 t-2.14 DISTRIBUTION .01407 .0037a .01388 .00438 .07 .00477 .00095 .00503 .00315 • .26 CUSTOMERS .00527 .00141 .00609 .00306 .62 .00184 .00058 .00233 .00143 -1.00 RALES .00266 .001)1 .00337 .00142 .79 .00084 .00047 .00153 .00078 -2.20 ADMINISTRATIVE AND GENERAL .01173 .00339 .01378 .00333 .89 .00295 .00074 .00383 .00110 -1.88 AREA: n RACE 1 ILRCTRIC COMPUTATIONS CAS COMPUTATIONS COMBINATION ST. ELECTRIC COMBINATION ST. CAS •-9 -9 • •6 EM STANDARD DEVIATION MEAN ' STANDARD DEVIATION t MEAN STANDARD DEVIATION MEAN STANDARD DEVIATION t COSTS/CUANTITT TOTAL OPERATINC REVENUE .17239 .02724 .1*977 .03334 +1.31 .07438 .02476 .06222 .01677 .98 TOTAL OPtUTINC EXPENSE .12788 .02314 .10798 .03001 1.30 .06336 .01970 .05247 .01238 1.14 OPERATINC EXPENSE .06327 .02430 .03606 .01603 .96 .04733 .01586 .04072 .00905 .89 Maintenance .00846 .00339 .00771 .00292 .30 .00210 .00105 .00107 .00052 2.07 DEPRECIATION .01938 .00707 .01614 .00345 1.16 .00314 .00300 .00414 .00174 .69 PRODUC. ICS .04370 .02046 .03607 .01463 .92 .03546 .01118 .03227 .00869 .33 DISTRIBUTION .01059 .00310 .00785 .00234 +2.14 .00541 .00260 .00314 .00168 1.76 CUSTOMERS .00486 .00132 .00416 .00087 1.22 .00297 .00135 .00184 .00104 1.46 SALES .00199 .00096 .00285 .00098 •1.87 .00097 .00073 .00134 .00079 .83 ADMINISTRATIVE AND CENERAL .01011 .00244 .01020 .00374 • .06 .00433 .00175 .00269 .00109 1.93

PAGE 152

-142AREA: V PACE 5 CAS COM P UTATIOMS COMBINATION T, ELECTS IC COMBINATION ST. CAS 0-6 STANT1ARD STANDARD IXMBtM D STAN HARD MEAN DEVIATION MEAN DEVIATION MEAN DEVIATION MEAN DEVIATION t COSTS/QUANTITY + .69 TOTAL OPERATlfC REVENUE .20153 .06132 .17824 .03022 +1.23 .05440 .01619 .00750 TOTAL OPERATING EXPENSE .15446 .04438 .13462 .02544 1.81 .04804 .01471 .04100 .00692 .89 OPERATING EXPENSE .08 204 .03381 .06592 .01708 1.40 .03629 .01300 .02927 .00875 .93 MAINTENANCE .0104} .00236 .01042 .00332 + .01 .00201 .00096 .00145 .00028 1.16 DEPRECIATION .022.46 .01000 .01914 .00321 1.12 .00346 .00113 .00339 .00213 .06 PRODUCTION .06105 .02883 .04497 .01482 1.63 .02758 .00942 .02107 .00937 1.03 DISTRIBUTION .OHM .00337 .01017 .00313 1.12 .00494 .00272 .00327 .00136 1.13 CUSTOMERS .00442 .00139 .00455 .00111 .59 .00230 .00113 .00180 .00070 .77 SALES .00234 .00079 .00399 .00099 3.34 .00049 .00034 .00094 .00050 • 1.60 ADMINISTRATIVE AND CEHERAL .00934 .00233 .01050 .00214 1.03 .00292 .00139 .00250 .00071 .35 AREA : VI PACE 3 1. costs/ouantitt total operating revenue total operating expense OPt RATING EXPENSE -MAINTENANCE DEPRECIATION noOUCTION DISTRIBUTION CUSTOMERS SALES ADMINISTRATIVE AND GENERAL ILECTRJC. COMPUTATIONS CAS COMPUTATIONS COMBINATION ST. ELECTS IC COMBINATION ST. CAS •2 ••10 ••2 6-4 MEM STANDARD DEVIATION MEAN STANDARD DEVIATION t MEAN STANDARD DEVIATIOIt MEAN STANDARD DEVIATION c .19192 .04533 .16017 .02174 1.38 .09171 .02336 .06106 .00691 1.84 . 15637 .04912 .12259 .01400 1.47 .07881 .02530 .05258 .00642 1.59 .10861 .03168 .06611 .00994 2.18 .03932 .01992 .04142 .00834 1.23 .00837 .00012 .00948 .00301 .47 .00224 .00087 .00139 .00006 1.39 .01161 .00493 .01898 .00313 1.47 .00311 .00032 .00433 .00357 .23 .08622 .03510 .03020 .00995 2.30 .04073 .00304 .03167 .00743 1.27 .00948 .00612 .00841 .00429 .27 .00618 .00507 .00344 .00171 .80 .00621 .00397 .00369 .00089 1.64 .00421 .00305 .00206 .00109 1.03 .00336 .00236 .00287 .00070 .43 .00420 .00412 .00108 .00041 1.23 .01106 .00440 .00852 .00230 1.03 .00567 .00421 .00323 .00082 .91

PAGE 153

143AHfA: I FACE 6 ElTCTRtC COMPUTATION S CAS COMPUTATIONS £OKBINATIC» n-15 ST. ELECrRIC a-21 COMBLNATION n-15 STANDARD STANDARD HFAN DEVIATION MEAN DEVIATION STANDARD PEVIA1 ECU STANDARD DEVIATION I. TOTAL WAGES/NO. EMPLOYEES 10.19955 2.82876 9.71567 1.1476* « .57 10.64049 1.50608 9.65448 .61445 +1.88 2. TOTAL WAGES/TOTAL OPERATING REVENUE 3. PENSION BENEFITS/ NO. EMPLOYEES 4. FENS ION BENEFITS/ TOTAL MAGES 5. TOTAL OPERATING REVENUE/ NO. EMPLOYEES 1.33651 .6160* 48.94257 6.553*7 .20966 .02733 .19 .20196 .04411 .17961 .05805 +1.05 1.14616 .39891 * .93 7.76714 5.16543 1.30047 .68576 +3.78 .11688 .0355* .32 .71786 .48418 .13561 .06785 *J.62 *6. 90097 7.01*90 + .77 54.18408 9.81351 61.77661 31.29982 .8* AREA: II PAGE 6 ELECTRIC COHftlTATTaB COMBINATION B*5 ST. ELECTRIC n-1* CAS COMPUTATIONS COMBINATION •4 STANDARD DEVIATION STANDARD MEAN DEVIATION STANDARD OEVUTION STANDARD DEVIATION 1. TOTAL UACES/NO. EMPLOYEES 9.01084 1.32029 10.3*337 1.13878 .37 1.22417 .48696 8.273*1 1.0*7** »1.*3 2. TOTAL WAGES/TOTAL OPERATING REVENUE .22205 .0*469 .01 .10669 .06376 .13992 .01522 .90 3. PENSION BENEFITS/ NO. EMPLOYEES 1.11276 .28861 .05 4.08550 .93755 1.06723 .50538 *5.30 *. PENSION BENEFITS/ TOTAL WAGES .10891 .031*7 • .00 .38343 .10123 .12740 .02471 «4.36 J. TOTAL OPERATING REVENUE/ NO. EMPLOYEES 4*. 85611 8.09420 47.83263 7.87150 .38 57.18275 30.09064 59.49763 8.24592

PAGE 154

-144AR£A: IH PACE 6 ELECTRIC CagUTATIOig CAS COMPUTATIONS combination n-14 ST. ELECTRIC o-3 COMBINATION n-14 ST. CAS n-5 STANDARD or; iatioh STANDARD DEVIATION STANDARD DEVIATION STANDARD DEVIATION 1. TOTAL WAGES /NO. EMPLOYEES 9.66956 1.24433 4.09576 7.09407 +2.58 10.38656 2.38802 10.20929 1.31264 .15 2. TOTAL WAGES/ TOTAL OPERATING REVENUE .02064 .06424 .11127 +3.59 .12256 .02923 .14833 04235 -1.14 3. PENSION BENEFITS/ NO. EMPLOYEES .69134 .34046 .60913 1.05504 + .22 4.01048 4.52262 1.13123 .65884 +1.54 4. PENSION BENEFITS/ TOTAL WAGES .10039 .08318 .35193 .28498 .11453 .06792 +1.74 5. TOTAL OPERATING REVENUE/ NO. EMPLOYEES 52.32683 8.22342 49.48773 15.28148 + .43 90.53871 36.30591 .17375 .04781 .93 AREA: TV PACE t E LECTR IC COMPUTATIONS CAS COMPUTATIONS COMBINATION n-9 STANDARD DEVIATION ST. ELECTRIC o-U STANDARD MEAN DEVIATION COMBINATION ST. CAS • -6 STANDARD DEVIATION STANDARD DEVIATION 1. TOTAL WAGES/NO. EMPLOYEES 10.33055 1.12149 10.41108 1.37796 -.14 10.02476 .83994 9.71572 1.74602 +.43 2. TOTAL W ACES/TOTAL OPERATING REVENUE 3. PENSION BENEFITS/ NO. EMPLOYEES ». PENSION BENEFITS/ TOTAL WAGES 3. TOTAL OPERATING REVENUE/ NO. EMPLOYEES .18952 .04290 + .JJ .17409 .03685 .14111 .05627 +1.28 1.18166 .67191 .87 4.24690 3.25487 . 79792 . 50901 +2.40 .09245 .02974 .11567 .05969 53.79475 13.21075 57.01434 13.26037 -1.01 .41095 .27257 .08088 .04816 +2.73 .51 60.16858 14.80727 84.52820 56.91396 -1.14

PAGE 155

-145AREA: V PACE 6 ELECTRIC CHPUTATinNS C AS CCMPUTATIONS COMBINATION n-6 STANDARD MEAN XVJXTim ST. ELECTRIC 0-16 COMBINATION n-6 ST. GAS n-5 STANDARD DEVLATLON STANDARD DEVIATION STANDARD DEV IATION 1. TOTAL WAGES/NO. EMPLOYEES 8.61135 1.03739 7.33819 3.81094 .77 15.3917: 16.97408 7.08951 1.01717 .99 2. TOTAL WAGES /TOTAL OPERATING REVENUE .1*333 .01535 .13522 .07127 + .26 .20879 .11265 .01847 +1.10 3. PENSION BENEFITS/ NO. EMPLOYEES 4. PENSION BENEFITS/ TOTAL WAGES 5. TOTAL OPERATING REVENUE/ SO. EMPLOYEES 60.47656 8.10856 .77711 .54904 -1.25 4.63847 6.42869 .72133 .29279 +1.23 .11423 .04965 -2.J1 .56684 .80251 .10115 .03798 +1.17 55.15903 7.59138 +1.37 64.24600 28.61462 49.00854 9.61111 +1.03 AREA: VI PACE 6 ELECTRIC CCttPUTATIOMS CCMINAT10B MEAN STANDARD DEVIATION ST. ELECTRIC n-10 STANDARD DEVIATION GAS C'JMPVTATIONS CCKBINATIC* ST. GAS STANDARD DEVIATION STANDARD DEVIATION I. TOTAL WAGES.' NO. EMPLOYEES 8.24111 .94969 -2.18 9.13014 2.87021 8.07497 .69535 .58 2. TOTAL WAGES /TOTAL OPERATING REVENUE .18264 .03536 .32 .19857 .13624 .13879 .02787 3. PENSION BENEFITS/ »0. EMPLOYEES .47072 .48 6.37813 9.30287 .93841 .31682 + .99 4. PENSION BENEFITS/ TOTAL WAGES .08594 .03984 .29 .38840 .83212 .11449 .03071 .93 3. TOTAL OPERA TINC REVENUE/ NO. EMPLOYEES 47.51613 1.64595 55.88029 10.23442 -1.05 66.75060 60.25238 60.31683 15.22887 + .16

PAGE 156

146AR7A: 1 PACE 7 BJ6CTWC X^i-U™!1 ™CAS C OM PUTATIONS COMBINATION 0-15 STANDARD DFVIATlOy ST. electric •-21 STANDARD DEVIATION COMBINATION -15 STANDARD DEVIATION CAS -10 STANDARD DEVIATION TOTAL 0PEPAT1NC REVENUE/ ( ) PLANT /Production /Distribution /General /Total Plant .66442 .63573 24.87929 .23496 .14113 .12281 34.72356 .04296 1.01589 .61755 7.84009 .24614 .78878 .11519 3.40778 .06885 -1.66 * .44 2.17 .54 7.76709 7.69830 .52010 .07363 47.15280 86.41510 .42643 .07749 9.32238 .60617 9.85087 .42747 10.28005 .14039 5.94613 .06400 .41 -1.92 1.31 .03 QUANTITY SALES/ ( /Production /Distribution /General /Total Plant 2.96034 2.87492 117.24857 1. 04498 .59819 .73621 168.97050 .17361 4.87847 3.04190 38.13348 1.19621 3.85072 .85883 15.87101 .35806 -1.86 .59 +2.07 -1.47 56.38251 60.37840 4.33029 2.41395 403.67725 756.32788 3.J2916 1.86419 76.09526 5.42826 73.39189 3.47751 85.02330 3.10214 36.39972 1.45709 .32 .0? < ) PLANT/TOTAL PLANT Production/ Dlatrlbutlon/ Caneral/ .35857 .37421 .01658 .05241 .05647 .01143 .30463 .40133 .03629 .10090 .09753 .01399 +1.84 .94 -3.97 .06598 .82202 .02086 .04813 .10947 .01968 .10146 .65360 .05486 .10224 .28264 .02578 -1.12 2.00 -3.38 DISTRIBUTION EXPENSE/ DISTRIBUTION PLANT ADMINISTRATIVE 6 GENERAL EXPENSE/ CENERAL PLANT 2.53900 .51872 AREA: XI PACT 7 fLFCTRl C_ COMPUTATIONS COMBINATION »I. ELECTRIC CAS COMPUTATIONS COMBIUATKW n-5 TOTAL OPERATIKC REVDRJE/ < ) PLANT /Production /Dlatrlbutlon /Central /Total Plant QUANTITY SALES/ ( ) PLANT /Production /Dlatrlbutlon /Casual /Total Plant .60433 .69053 71.52626 .22693 3.44795 3.98814 446.06299 1.30760 .09596 .06916 112.42873 .01184 .37919 .68715 723.47046 .16214 MEAN DEVIATION t MEAN DEVIATION KEAN .61937 26940 .12 24.78067 17.02681 46.38475 .75764 16286 .84 .66460 .05288 .74740 6.27467 2 33983 2.05 55.77084 62.76033 8.17960 .20726 04434 .93 .51638 .09783 .51394 3.60299 1 33093 .26 361.60643 273.28223 699.69165 4.57852 1 61500 .75 9.33081 1.48605 10.75668 37.56760 U 48656 2.00 827.87329 970.62446 116.15576 1.22311 24427 .66 7.27357 1.70073 7.36356 STANDARD DEVIATION 30.86942 .10371 1.90057 .06263 555.82275 2.47904 26.39456 1.45446 -1.19 -1.37 1.36 1.29 .07 ( ) PLANT/TOTAL PLANT Production/ Dlatrlbutlon/ Central/ .38470 .32984 .01823 .07476 .01579 .01905 .36378 .28166 .03339 .08731 .07048 .01664 .45 1.43 -1.95 .02887 .77472 .01963 .02310 .12237 .01325 .01369 .69133 .06670 .00537 .07186 .02405 1.05 1.06 -3.29 4. DISTRIBUTION EXPENSE/ DISTRIBUTION PLANT ADMTJIISTRATTVE 6 GENERAL EXPENSE/GENERAL PLANT

PAGE 157

-147AREA: III PAGE 7 uremic cowrorATiao COMBINATION n-14 total operating revenue/ { ) plant /Production /Distribution /Gcnsral /Total Plant MEAN .91274 .66674 26.10368 .22359 STANDARD DEVIATION .95271 .10282 48.99675 .04118 ST. ELECTRIC .65296 .65431 7.35643 .22149 STANDARD DEVIATION .12610 .02312 2.242S8 .01514 CAS COMPUTA TIONS STANDARD DEVIATION 23.04012 19.57838 .64837 .14197 32.55156 35.31924 .55678 .13836 ST. CAS n-5 151.32358 1.40685 9.40937 .46549 STANDARD D EVIATION 307.76929 1.79413 3.73921 .13752 -1.47 -1.48 1.38 +1.20 I. (jiAirrm sales/ ( ) plant /Production /Dlitrlbutlon /Canaral /Total Plant 4.10860 3.15138 132.69988 1.04395 3.70148 .66288 270.55640 .15723 3.06781 3.09839 35.28256 1.04516 .46587 .35514 13.24748 .07713 314.91650 273.38232 9.43007 3.66380 453.73145 459.62061 8.13020 3.3*7684 1,737.99121 3,476.57471 29.02657 47.91403 144.73911 86.78851 6.53347 1.55927 -1.44 -1.44 +1.41 + .96 ( ) PLANT/TOTAL PLANT Production/ Distribution/ Cnneral/ .31678 .34465 .02272 .09143 .09656 .01820 .34832 .33915 .03225 .07546 .03249 .01055 .03087 .96112 .02388 .03004 .11603 .01482 .04478 .68828 .05791 .04591 .35193 .03022 +1.52 -3.08 DISTRIBUTION EXPENSE/ DISTRIBUTION PLANT ADMINISTRATIVE A CENTRAL moastrauujj. plant ELECT R IC COMPUTATION S COMBINATION IT. ELECTRIC m-» »-U TOTAL OPERATING REVENUE/ < ) PLANT /Production /Distribution /Cnneral /Total Plant SEAN .85075 .33154 24. 37587 .21585 STANDARD DEVIATION .88044 .07688 27.71120 .04313 KEAN .58426 .39604 6.51267 .199U STANDARD DEVIATION .17115 .13826 2.21225 .03282 CAS COMPUTATIONS COMBINATION ST. GAS n-9 a-6 + .93 -1.19 2.02 .93 10.9t326 .57810 49.30745 .40146 STANDARD DEVIATION 13.53391 .23789 38.56288 .09338 KEAN 7.78129 .45855 13.47765 .49300 STANDARD DEVIATION 4.38625 .03639 17.96411 .38773 .31 1.14 +1.98 .63 qUABTITT SALES/ ( ) PLANT /Production 4.99438 5.29182 4.13891 1.82581 /Distribution 3.20144 .98076 4.07327 .99763 /Canaral LM.28247 158.04445 45.59909 14.63849 /Total Plant 1.27103 . 29260 1.37715 . 31698 +1.86 .72 148.36430 9.17829 749.29590 5.89732 178.77594 6 . 14872 615.62256 2.04228 109.17215 7.08268 302.13794 9.97952 55.86595 2.15719 510.77588 11.98903 + .48 + .75 1.37 .93 ( ) PLANT/TOTAL PLANT Production/ Dlatrloutlon/ Canaral/ .34215 .41522 .02314 .10376 .11219 .01902 .36063 .34673 .03303 .09059 .08156 .00928 +1.50 -1.44 .06179 .77215 .01413 .10571 .23240 .01186 .05918 .60648 .05110 .08556 .31205 .01664 + .03 +1.10 -A.67 DISTRIBUTION EXPENSE/ DISTRIBUTION PLANT 5. ADMINISTRATIVE 6 GENERAL DLPENSL7GENERAL PLANT

PAGE 158

-148AREA: V PACE 7 ELECTRIC CCM njTATIONS CAS COMPUTATIONS COMBINATION TOTAL OPERATING REVENUE/ ( ) PLANT /Production /Distribution /Comal /Total Plsnt WEAN 1.12899 .69165 11.0*931 .2*613 STANNARD DEVIATION 1.23271 .14333 7.70853 .03079 ST. ELECTRIC -16 .69991 .6*501 6.73897 .23278 STANDARD DEVIATION .11306 .08*06 1.75821 .01967 1.29 + .70 2. DO * .59 COMBINATION 107.17012 .75270 28.53551 .39375 STAN^AF.D DEVIATION 127.76736 .36503 23.26767 .05611 33.73570 1.1751* 9.70990 .45523 71.37306 .90141 4.12722 .19331 H.03 .94 + 1.62 QUANTITY SALES/ ( ) PLANT /Production 5.70607 5.90521 4.01710 .8 1107 /Distribution 3.59447 1.30175 3.76555 1.12311 /Cansral 52.64758 22.97974 38.6*269 12.20047 /Total Plant 1.26553 .41112 1.34399 . 28266 1.06 1.5*4.36328 1.713.49463 .29 16.55785 13.30679 1.73 525.31055 344.53149 .*« 7.73717 2.37627 660.16089 1,386.35938 27.28456 27.20569 203.67920 83.85915 9.49540 3.63411 .84 .77 +1.84 .87 ( > PLANT/TOTAL PLANT Production/ Distribution/ Canaral/ .32140 .36193 .03011 .13639 .07874 .01902 .33957 .36*63 .03630 .05330 .0*045 .00810 .43 .02052 .03207 .10 .64591 .29102 -1.02 .02823 .03025 .14520 .5254* .05163 .16715 .32834 .02838 -1.62 .58 -1.89 4. DISTRIBUTION EXPENSE/ DISTRIBUTION PLANT .12 .053*3 5. ADMINISTRATIVE & GENERAL EXPENSE/GENERAL PLANT 1.15 1.315*3 ARIA: TX PACE 7 ELECTRIC COM P UTAT IONS 645 COMPUTATIONS COMBINATION ST. ELECTRIC COMBINATION ST. CAS •-! a10 n-2 n-i STANDARD STANDARD STANDARD STANDARD BAN DEVIATION MEAN DEVIATION NEAN DEVIATION REAN DEVIATION t 1. TOTAL OPERATING REVENUE/ ( ) PLANT /Production 1 . 70197 1.(3918 .59420 .14818 +1 71 11.26306 6.47734 123.08973 129.30487 -1.00 /Distribution .64993 .01337 .67339 .11310 27 .36362 .08382 1.10926 .90295 .70 /Carvcra! 10.44219 2.73520 7.57143 3.84971 92 15.30473 11.34826 9.33162 2.63545 .81 /Total Plant .33643 .21333 .20621 .01964 ft 73 .41337 .13242 .44014 .16719 .39 2. QUANTITY SALES/1 ) PLANT /Production 7.95804 7.70261 3.72776 .11920 + 1 54 137.8*766 1C8. 74805 1,837.15039 1,909.30640 -1.02 /Distribution 3.47326 .75155 4.26103 .82761 -1 14 6.25861 .81673 17.55829 12.71658 -1.02 /Cansral 54.239*0 1.43518 48.32761 26.69290 + 29 191.30530 176.64290 151.69415 37.42838 .35 /Total Plant 1.67034 .71790 1.29946 .13304 1 38 4.50165 .19901 8.05928 3.15369 -1.30 3. ( ) PLANT/ TOTAL PLANT Production/ .31272 .21247 .36028 .06730 33 .04824 .03950 .02773 .03240 .55 Distribution/ .31502 .31795 .31271 .05137 1 73 .72754 .12674 .66563 .38134 .14 Canaral/ .03063 .01243 .03209 .01189 14 .04185 .03969 .05595 .02821 • .41 4. DISTRIBUTION EXPENSE/ DISTRIBUTION PLANT .03064 .01413 .03380 .01310 28 .03661 .02670 .04479 .00832 .44 5. ADMINISTRATIVE & CENTRAL EXPENSE/GENERAL PLANT .60293 .23470 .3*342 .18439 +1.23 .7115* .19554 .49713 .19065 +1.06

PAGE 159

-149AREA: I PAGE 8 it*?TKl c _ COMPUTATIONS COKalNATION ST. ELECTRIC CAS COMPUTATIONS OOKSINATION ST. CAS 1. ose or funds as percent Of TOTAL OPERATINC REVENUE STA-IDAHD wjjngi STANDARD DEVIATION STANDARD DEVIATION STANDARD DEVIATION CROSS AKD PLANT/TOTAL OPERATING REVENUE DIVIDEND ON PREFERRED/ DIVIDEND ON COMMON/ FUNDS RETAINED/ .60759 .01736 .10113 .35-40 .01240 .02039 .35492 .01117 .OS950 .18635 +2.25 2.59072 2.29694 .00765 .20 .06722 .07632 .02010 +1.46 .40932 .24456 .12635 .37 .51051 .76461 .17706 .00549 .05434 .11130 .07457 4-J.H .00647 +2.45 .01011 44.39 .12473 +1.37 AREA: II PACE I ELECTRIC COMPUTATIONS CCtUl f NATION ST. ELECTRIC o-5 >-U STANDARD DEVIATION STANDARD DEVIATION CAS C OMPUTAT IONS caaiNATicH it. cas STANDARD STANDARD MEAN DEVIATION PffiAN DEVIATION 1. USE OP HINDS AS PERCENT OP TOTAL OPERATING REVENUE CROSS AND PLANT/TOTAL OPERATING REVENUE DIVIDEND Oil PREFERRED/ DIVIDEND OH COMMON/ PONDS RETAINED/ .01620 .12665 .28188 .48399 .01296 .09436 . 10052 .17667 .01006 .04232 .13598 .01 .87716 .00 .02656 .00 .20752 4 .02 .40498 .50135 .01553 .11713 .00299 «.5» .00364 4-2.75 .01175 42.35 .04272 4 .89

PAGE 160

-150AREA: III MCI 8 COMBINATION ST. ILECTRIC •-14 «-3 standard wan deviati on MEAN CAS COH1U1AT10NS COfUUNMION ST. CAS • -14 n-5 STANDARD DEVIATIO N t STANDARD DEVIATION^ STANDARD DEVIATION I. OSI OF FUNDS AS PERCENT OF TOTAL OPERATING REVENUE CiOSS AND PLANT/TOTAL OPERATING REVENUE DIVIDEND ON PREFERRED/ DIVIDEND ON COMMON/ FUNDS RETAINED/ .49913 .24206 .U722 .12427 +1.78 I.1S507 .79107 .15709 .01491 .00976 . 00392 . 00679 4-1.75 .03829 .03882 . 00500 .09390 .04«16 .03493 .06051 +1.72 .21669 .15701 .05790 .05092 .10010 .00108 .00187 .81 .10027 .18912 .08116 .06649 +2.66 .00594 -H.81 .01052 »2.16 .09907 .20 IHCTRIC C OMPUTA TIONS COMBINATION ST. ELECTRIC -* —11 STANDARD DEVIATION WEAN STANDARD DEVIATION , . CAS COMPUTATIONS COMBINATION ST. CAS •-4 i-6 STANDARD DEVIATION STANDARD DEVIATION 1. USE OF FUNDS AS PERCENT OF TOTAL OPFJtATlNC REVENUE CROSS AND PLANT/TOTAL OPERATING REVENUE DIVIDEND ON PREFERRED/ DIVIDEND 01 COMMON/ FUNDS RETAINED/ .13898 .01036 .04349 .06615 .07375 .05957 •1.03 1.36562 1.15396 .93 .04403 2. 66 .30571 *2.C2 .32285 .05440 tl.it .00653 -2.67 .01490 +1.69 .04317 -1.69

PAGE 161

-151AREA: V PACE 8 ELE CTRIC COM HHATIONS COKUINATIOH ST. ELECTRIC n-6 o-lS standard deviation STANDARD DEVIATION CAS C OMPUTATIONS COMBINATION ST. CAS STANDARD DEVIATION STANDARD DEVIATION 1. OSS OF FUNDS AS PERCENT OF TOTAL OPERATING REVENUE CROSS AND PLANT/TOTAL OPERATING REVENUE DIVIDEND ON PREItRRCD/ DIVIDEND OS COIMON/ FUNDS RETAINED/ +1.20 4.25064 7.23001 2.45 .168)7 .26221 + .68 .90457 1.47060 .71 .80149 1.16472 .08399 +1.13 .00309 41.21 .04076 +1.15 .05021 +1.21 AREA: VI PACE S IX ECTRIC COM PUTATIONS COMBINATION •-1 JKAN. STANDARD DEVIATION ELECTRIC •-10 CAS COMPUTATIONS BIMAT10N o-2 STANDARD DEVIATION STANDARD DEVIATION 1. DSE OF FUNDS A3 PERCENT OF TOTAL OPERATING REVENUE SSOSS AND PLANT/TOTAL OPE1ATIHC REVENUE .» 1.68155 2.11053 .05609 +1.20 DIVIDEND ON PREFERRED/ .01935 + .17 .04374 .00465 .00143 +1.31 DIVIDEND ON COMMON/ FUNDS RETAINED/ .10369 .04410 + .6] .24967 .32 .07239 .03937 .11269 .01734 +1.33 .13(32 + .29

PAGE 162

BIBLIOGRAPHY Books Anderson, Ronald A. Government and Business. Cincinnati: South-Western, 1966. Anderson, Thomas J. Our Competitive System and Public Policy . Cincinnati: South-Western, 1958. Arrow, K. J. Social Choice and Individual Values . New York: Wiley, 1951. Bain, Joe S. Industrial Orgnization . New York: John Wiley and Sons, Inc., 1968. Baumol, William J. Business Behavior, Value and Growth . New York: Harcourt, Brace and World, 1967. Boulding, Kenneth E. Economic Analysis . New York: Harper and Row, 1966. Dimock, Marshall E. Business and Government . New York: Holt, Rinehart and Winston, 1961. Garfield, Paul J. and Lovejoy, Wallace, F. Public Utility Economics . Englewood Cliffs: Prentice-Hall, Inc., 1964. Graaff, J. Theoretical Welfare Economics . Cambridge: Cambridge University Press, 1963. Hamburg, M. Statistical Analysis for Decision Making . New York: Harcourt, Brace and World, 1972. Kahn, A. E. The Economics of Regulation: Principles and Institutions . Volumes I & II , New York: John Wiley and Sons, 1970. j .. . ' Mishan, E. J. Welfare Economics: Ten Introductory Essays . New York: Random House, 1969. Needham, Douglas. Economic Analysis and Industrial Structure . New York: Holt, Reinhart and Winston, 1969. Reder, M. A. Studies in the Theory of Welfare Economics . New York: Columbia University Press, 1947. Samuelson, Paul A. Foundations of Economic Analysis . Cambridge: Harvard University Press, 1955. -152-

PAGE 163

-153Shepherd, William G. Market Power and Economic Welfare . New York: Random House, 1970. Shepherd, William G. and Gies, Thomas G. Utility Regulation New Directions in Theory and Practice . New York: Random House, 1966. Singer, Eugene M. Antitrust Economics . Englewood Cliffs: PrenticeHall, 1968. Turvey, R. , Ed. Public Enterprise . Baltimore: Penguin, 1968. Wilcos, Clair. Public Policy Toward Business. Homewood : Richard D. Irwin, 1960. Williamson, 0. The Economics of Discretionary Behavior: Managerial Objectives in a Theory of the Firm . Englewood Cliffs: PrenticeHall, 1964. Articles Averch, Harvey and Johnson, Leyland L. "Behavior of the Firm Under Regulatory Constraint," American Economic Review , 52 (December, 1962), 1052-69. Bailey, E. and Malone, Jr. "Resource Allocation and the Regulated Firm," The Bell Journal of Economics and Management Science , 1 (Spring, 1970), 129-42. Baumol, W. and Klevorick, A. "Input Choices and Rate-of -Return Regulation: An Overview of the Discussion," The Bell Journal of Economics and Management Science , 1 (Autumn, 1970), 162-190. Bergson, Abram. "A Reformulation of Certain Aspects of Welfare Economics," Quarterly Journal of Economics , 52 (February, 1938), 310-34. Bonbright, J. C. "Fully Distributed Costs in Utility Rate Making," American Economic Review , 51 (May, 1961), 305-12. Bonbright, J. C. "Major Controversies as to the Criteria of Reasonable Public Utility Rates," American Economic Review , 30 (May, 1940) 379-89. Bonbright, J. C. "Two Partly Conflicting Standards of Reasonable Utility Rates," American Economic Review , 48 (May, 1957), 386-93. Brandon, Paul S. "The Electric Side of Combination Gas-Electric Utilities," The Bell Journal of Economics and Management Science , 2 (Autumn, 1971), 688-703.

PAGE 164

-154Buchanan, James M. "Peak Loads and Efficient Pricing: Comment," Quarterly Journal of Economics , 80 (August, 1966), 463-71. Coase, R. "The Theory of Public Utility Pricing and Its Application," The Bell Journal of Economics and Management Science , 1 (Spring, 1970), 113-128. Coase, R. H. "The Marginal Cost Controversy," Economica , 13 (November, 1946), 169-82. Hicks, John R. "The Foundations of Welfare Economics," Economic Journal , 49 (December, 1939), 696-712. Hotelling, Harold. "The General Welfare in Relation to Problems of Taxation and of Railway and Utility Rates," Econometrica , 6 (July, 1938) , 242-69. Kafoglis, Milton Z. "Output of the Restrained Firm," American Economic Review , 59 (September, 1969), 583-9. Kaldor, Nicolar. "Welfare Propositions in Economics and Interpersonal Comparisons of Utility," Economic Journal , 49 (September, 1939) , 549-52. Landon, J. H. "Electric and Gas Combination and Economic Performance," Journal of Economics and Business (Fall, 1972), 1-13. Little, I. M. D. "Review of T. Scitovsky, 'Welfare and Competition'," Econometrica (1952), 703. Lipsey, R. and Lancaster, K. "The General Theory of Second Best," Review of Economic Studies , 24 (1957). Mann, P. C. "The Impact of Competition in the Supply of Electricity," Quarterly Review of Economics and Business , 10 (Winter, 1970), 37-49. Monsen, R. J., Chiu, J. S., and Cooley, D. E. "The Effect of Separation of Ownership and Control on the Performance of the Large Firm," Quarterly Journal of Economics , 82 (August, 1968), 435-51. Nelson, J. R. "Practical Applications of Marginal Cost Pricing in the Public Utility Field," American Economic Review , 53 (May, 1963), 474-81. Owen, Bruce M. "Monopoly Pricing in Combined Gas and Electric Utilities," Antitrust Bulletin , 15 (Winter, 1970), 713-727. Ruggles, Nancy. "Recent Developments in the Theory of Marginal Cost Pricing," Review of Economic Studies , 17 (1949-50), 107-26.

PAGE 165

-155Ruggles, Nancy. "The Welfare Basis of the Marginal Cost Pricing Principle," Review of Economic Studies , 17 (1949-50), 31. Scitovsky, T. "The State of Welfare Economics," American Economic Review , 41 (1951) , -303. Shepherd, William G. "Marginal Cost Pricing in American Utilities," Southern Economic Journal , 23 (July, 1966), 58-70. Smith, D. B. "Inter -Energy Competition Between Gas and Electricity in the Industrial Market: A Case for Deregulation?" Business Studies , 12 (1972). Steiner, Peter 0. "Peak Loads and Efficient Pricing," Quarterly Journal of Economics , 71 (November, 1957), 585-610. Stigler, George J. "The Theory of Economic Regulation," The Bell Journal of Economics and Management Science , 2 (Spring, 1971), 3-21. G. J. Stigler and C. Friedland. "What Can Regulators Regulate? The Case of Electricity," The Journal of Law and Economics , Volume 5, 1962. Vickery, William S. "Some Implications of Marginal Cost Pricing for Public Utilities," American Economic Review , 45 (May, 1955), 605-620. Vickery, William S. "Some Objectives to Marginal Cost Pricing," Journal of Political Economy , 56 (June, 1948), 218-38. Williamson, Oliver E. "Peak-Load Pricing and Optimal Capacity Under Indivisibility Constraint," American Economic Review , 56 (September, 1966), 810-27. Government Publications Combination Utility Companies . Hearings before the Subcommittee on the Judiciary, United States Senate, 92nd Congress, First Session. Washington, D. C: Government Printing Office, 1971. Federal Power Commission. Statistics of Privately Owned Electric Companies in the United States . Washington, D. C. : Government Printing Office, 1970.

PAGE 166

BIOGRAPHICAL NOTES John Angus And erson was born March 27, 1943 in St. Petersburg, Florida. In June, 1961, he was graduated from St. Petersburg High School. In August, 1961, he received the degree of Bachelor of Arts from the University of South Florida, Tampa, Florida. In August, 1967, he received the degree of Master of Arts from the University of Florida, Gainesville, Florida. He is presently employed as an Assistant Professor of Economics, University of South Florida, St. Petersburg, Florida. John is married with two children. He is a member of Beta Gamma Sigma and Omicron Delta Epsilon honorary fraternities .

PAGE 167

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Milton Z. Kafoglis Professor of Economics I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Charles W. Fristoe Associate Professor of Economics I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Norman G. Keig Associate Professor of Economics I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. John H. James Associate Professor of Management

PAGE 168

This dissertation was submitted to the Department of Economics in the College of Business Administration and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. Dean, Graduate School