WELFARE ECONOMICS AND THE THEORY OF OPTIMUM PUBLIC UTILITY
PRICING AND THEIR PRACTICAL APPLICATION WITH SPECIAL
REFERENCE TO FEDERAL TRANSPORTATION POLICY
DISSERTATION
Presented In Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State Universi ty
By Thomas Anton Martinsek, A. B., M. A
The Ohio State University 1956
Approved by:
Adviser Department of Economics TABLE OP CONTENTS
Page
LIST OP ILLUSTRATIONS v.
Part One
WELFARE ECONOMICS AND EQUILIBRIUM CONDITIONS Chapter
I. INTRODUCTION . <...... 2
II. ECONOMIC WELFARE CRITERIA ...... >■ 8 The Nature of Welfare Economics: The Social Welfare Function Efficiency of Production and Consumption: Marginal Conditions for an Optimum Reorganizations: The Compensation Principle Welfare Measures To Be Used in the Study
III . LAWS OF COSTS AND R E T U R N S ...... 44 Firm Cost Patterns Industry Supply Patterns Summary and Conclusi ons
IV. EQUILIBRIUM CONDITIONS AND MAXI MUM WELFARE... 84 Equilibrium of the Firm Equilibrium of the Industry Summary and Conclusions
Part Two
THE THEORY OF PUBLIC UTILITY PRICING
V. THE CRITERIA FOR PUBLIC UTILITY CLASSIFICATION ...... 122 The Legal Criteria for Public Utility Regulation An Economic Criterion for Public Utility Classification Existing Public Utility Control In the Light of the Postulated Economic Definition
VI, THE DETERMINATION OF LONG-RUN PUBLIC UTILITY PRICES ...... 147 The Possible Choices Payments Necessary with Marginal-Cost Pricing Conclusions
ii Chapter - Page \ VTI. THE SHORT-RUN AND OTHER PROBLEMS. S OP PUBLIC UTILITY PRICING . .'...... $. 178 Short-Run Public Utility; Pricing New and Expanding Industries Dyinp; and Contracting Industries Conclusions
Part Three THEORY AMD PRACTICE IN PUBLIC UTILITY REGULATION VIII. PRA0TICAL PUBLIC UTILITY REGULATION: FEDERAL TRANSPORTATION REGULATION POLICY .... 206 The Goals of Federal Transportation Regulati on Federal Public Utility Price Setting Poli cy Conclusions
IX. USE OF THE THEORETICAL CONCEPTS I N 'ACTUAL PRICE SETTING...... '...... 241 Measurement of Cost and Demand Functions Guaranteeing the Efficiency of Operations Collection of the Tax and Payment of the Subsidy Market Imperfections and Marginal-Cost Pricing Standards for Investment Summary on Practical Application of the Theoretical Tools
X. SUMMARY, CONCLUSIONS AND RECOMMENDATIONS ___ 260
APPENDIX A ...... 269 BIBLIOGRAPHY...... 273
111 LIST OP ILLUSTRATIONS
Figure Page
1. Utility Surfaces ...... 22 2. The Production Function ...... 47 3. Increasing Costs to Scale ...... 55 4. Constant Costs to Scale ...... 56 5. Decreasing Costs to Scale ...... 57 6. Industry Cost Patterns ...... 74 7. Increasing Costs to Scale ...... 88
8. Decreasing Costs to Scale ...... 92 9. Constant Costs to Scale ...... ‘...... 99
10. Increasing Industry Costs ...... 103 11. Constant Industry Costs ...... 108 12. Decreasing Industry Costs ...... 113
13. • Marginal-Cost Pricing ...... 150
14. Average-Cost Pricing ...... 157 15. Tax on Public Utility Consumer ...... 163
16. Increase in Consumers' Surplus ...... 167 17. Short-Run Marginal-Cost Pricing ...... 180
13. Short-Run Average-Cost Pricing ...... 187 19. Average-Cost Criterion for Increased Investment ...... 197
iv PART ONE
WELFARE ECONOMICS AND EQUILIBRIUM CONDITIONS
1 CHAPTER I
INTRODUCTION The logic of public utility pricing Has significance
for more than the pricing of public utility services. The same logic applies to control of prices in general. The problems of general pricing controls, however, lie outside the scope of this study. Only public utility pricing in an economy characterized by a minimum of control will be con
sidered. Three basic questions will be considered. The nature
of the industry which Is to be regulated will receive first consideration. It will then be possible to examine the na
ture of the regulation. The economic logic of placing In dustries under control and of the regulation also will be
compared with transportation regulation policy on the fed eral level In the United States. Natural monopoly conditions will be the subject covered in Part One of the study. Attention will be directed at the
cost conditions which would result in a breaking down of op
timal equillbriiim conditions in unregulated markets. Such optimal equilibrium conditions will be derived from contem porary economic welfare theory. Put in another way, Part One will examine cost patterns to determine the types of cost patterns which would lead to non-optimal conditions in unregulated markets. Cost patterns and the physical produc tion conditions which a firm might encounter will be consid er 3 ered, as will industry cost patterns. Firm and industry
cost patterns then will be viewed from the standpoint of possible optimal equilibrium conditions.
Part Two will consider public utility pricing. An at tempt will be made to set up a definition of public utility consistent with the welfare standards. The legal definition will be compared with the economic definition. Having con sidered cost structures leading to a collapse of optimal equilibrium, applicable welfare standards and the definition of public utility, the problem of the pricing of public utility services can be examined. The pricing of public utility services will be broken into several areas. Long- run pricing and short-run pricing will be of greatest impor tance. New and expanding industries also will be examined, as will contracting and dying industries. Standards appli cable to pricing and investment will be the point of empha sis in the examination of these areas.
The last major area to be brought into the study will be the practical considerations Involved in public utility price setting. Practical problems involved in the applica tion of price setting standards developed from theoretical economic concepts will be examined. The feasibility of the use of theoretical concepts in actual regulation will be of major concern. The goals and pricing standards of federal public utility regulation as evidenced by the pertinent statutes and Supreme Court decisions dealing with federal 4 transportation regulation policy also will be examined.
Comparison of these goals and pricing standards with econom ic standards will receive major attention. The possible contribution of economic analysis to public utility price setting will be developed from the comparison.
Static concepts and standards will be used in this study. In the investigation of the cost structures, con ventional static cost theory will be the basis of consider ation. Similarly, the welfare standards will be those suited to the evaluation of static equilibrium conditions.
Use of static standards in a dynamic world results in in adequacy of treatment and gives rise to shortcomings in ap plication. However, dynamic theory in economics at the pre sent time is either of too sketchy a nature or inapplicable to the problems under consideration in this study.
Great difficulty arises just in the definition of eco nomic dynamics. To some people dynamics is apparently a synonym for good. Others define dynamics as arising when expectations are included in the analysis. Another defini tion applies the term dynamics to analysis in which every quantity is dated. Emphasis on time lags is the essence of dynamics to some. The nature of the achievement of equi librium is used at times as another definition. The secular growth of the entire economy in a macro-economic sense is the basis for the development of a substantial body of liter- 5
ature at the present time.-*-
-*-Por a discussion of various definitions of dynamics, see R. P. Harrod, Towards A Dynamic Economics, (London: Macmillan & Co., 1949), pp. 1-20.
Of the approaches mentioned above, the area of greatest
advance in dynamic theory has bean with the macro-economic
concepts of total output, employment, and price levels which are not suitable for application to the basic problems of
this study. Attempts at rigorous consideration of the dy namic aspects of the prices and outputs of individual com modities have been much less successful. They usually de pend on the postulating of special conditions such as lags in the response of producers to price changes used in the Q cobweb theorem. Dynamic welfare analysis is quite sketchy,
Ezekial, ,TThe Cobweb Theorem,” Quarterly Journal of Economics, Vol. LII (1937-58), pp. 255-26-SO. emphasizing the movement toward an equilibrium position.^
^See, M. W. Reder, Studies in the Theory of Welfare Economics, (New York: Columbia University Press, 1947), pp. 103-177.
In view of all the difficulties of applying the exist ing body of economic dynamic theory to the problem of public utility pricing, there.is little choice but to carry the analysis on a static level. One other alternative exists, 6 of course. The alternative is to develop a theory of eco nomic dynamics within the study itself. This is ruled out for two reasons. The development of a dynamic theory, if possible, would be a task in itself, overwhelming the con fines of this study. It is also impossible to say whether a rigorous, general theory of the dynamics of individual prices and output is possible. The result is that static analysis will be basic to this investigation. Such limita tion leads to possible shortcomings which must be kept in mind. In summary, this study is essentially a weaving togeth er of several strands of economic thought--price theory, welfare economics and public utility pricing. It is, thus, an attempt to bring together in a systematic study matters which have been discussed in a scattered, and often fragment ed, form in many books and journals. A major purpose of the study is to present a rigorous, thorough examination of the theory of public utility pricing from the standjjoint of wel fare economics and to consider the theory of optimum public utility pricing as to possible practical application.
The assumption upon which this study was undertaken was that comprehensive studies of public utility pricing theory are either polemical or based upon inadequate theoretical foundation or, at times, both while many adequate shorter treatments deal with more limited areas of the problem. It is hoped that a rigorous treatment of the theory with a 7 limited examination of practical application will furnish a rounded presentation of the theory and demonstrate the limi
tations on its practical application for those who could use
the theory in application to empirical studies relating to public utilities and transportation. CHAPTER II
ECONOMIC WELFARE CRITERIA Consideration of economic welfare criteria is included in this study because of the necessity of setting up cri teria to be used in the setting of public utility prices which would aid In, or at least not be detrimental to, the achievement of an economic optimum. Before discussing the effect of any attempt to achieve an optimum, however, it is necessary to give some consideration to such an optimum. Examination of such an optimum receives primary emphasis In the theory of welfare economics. The present state of wel fare economics and the conclusions it has to offer will be, therefore, the subject considered in this chapter. Because the general social welfare function is basic to an understanding of contemporary welfare economics, the social welfare function will be examined first. Principles which, given special assumptions, are deducible from this function then will be considered. There are two sets of such principles, the compensation principles and the mar ginal conditions for an optimum. These principles will be examined in the light of recent criticism. A choice will be made as to the welfare principles which will be used in the remainder of the study as guides for a system of optimum pricing of public utility services.
8 9
A. The Nature of Welfare Economics: The Social Welfare
Function."*-
-*-For a short summary of the current position of welfare economics see Kenneth EC. Boulding, "Welfare Economics," A Survey of Contemporary Economics, Volume II, ed. by B. P. Haley^ (Chicago: Richard D. Irwin, Inc., 1952), pp. 1-38. For a more extended treatment see Melvin W. Reder, op. cit.
Originating with pamphleteers and pleaders of various causes, the study of economics has from its beginnings dealt with welfare. Unfortunately, through much of the literature the criteria, the methods and the results become intertwined p and mixed to the detriment of lucid analysis.'5 Present day
2Reder, op. cit. , p. 13. welfare economics attempts to make explicit the standards or criteria by which the results of any action are judged, to have either a beneficial or an adverse effect economically. Any action, of course, may have one type of economic effect and just the opposite when examined from another point of view, e_.g;. political, ethical or military. For a short definition of welfare economics, Baumol provides one as satisfactory as can be found. "In it are examined the effects of economic actions and decisions on the welfare of the individuals who compose the economy or, more generally, on the welfare of any group to which the analyst chooses to direct his attention. In short, it might 10
be described as the theory of economic policy.
^W. J. Baumol, "Economic Theory and the Political Scientist," World Politics, Vol. VI (January, 1954), p. 266.
Starting with the very general and then proceeding to
the more specific is an approach which lends itself to many
areas. Welfare economics seems to be one of those areas.
Bergson’s general social welfare function makes an adequate point of departure for the study of welfare economics."
^A. Bergson, "A Reformulation of Certain Aspects of Welfare Economics," Quarterly Journal of Economics, Vol. LII (February, 1938)j pp. 510-554; id.j "’Socialist Economics," A Survey of Contemporary Economics, Volume I, ed. by H.S. Ellis, (Philadelphia: The Blaklston Co., 1949), pp. 412-448.
A general, undefined social welfare function may be set up as
W r W(x1# x2, x3, xn) in which the x's represent any variable considered relevant including any economic or non-eoonomio elements. In this general, unspecified version, it is stated merely that the welfare (W) of the group being considered is a function of many variables. No specific elements are mentioned as being included, excluded, important, unimportant, detrimental or beneficial. In this very general form the function is sterile and uninteresting.
Further meaning and interest arise when the function is 11 specialized, i.e.. when certain variables are explicitly con sidered to have a certain degree of relevance and when side relations are set up as constraints or boundaries. Such ex plicit specialization of the welfare function can be based upon whatever value judgments the specifier wishes to choose.
If he is an out and out Utopian he may wish to ignore various institutional relations re gardless of their empirical importance; Indeed he may go all out and repeal the laws of con servation of energy and widen greatly the technological productivities of the system. On the other hand, he may wish to take as fixed and immutable all social and economic institutions except those relating to the Central Bank. (Indeed those of fatalistic tem perament may regard the restraints to be so numerous as to leave no problem of choice.) In other words, the auxiliary constraints im posed upon the variables are not themselves the proper subject matter of welfare economics but must be taken as given.^
5Paul Anthony Samuelson, Foundations of Economics Anal ysis , (Cambridge: Harvard University Press, 1948)",™p p . 221- 2227 It should be noted that all the institutions, includ ing the Central Bank, may be defined in any way the speci fier happens to choose.
To get some content into welfare economics, most writers specialize the welfare function by means of a few general value judgments having a minimum of controversial content. The form in which the function is most commonly specialized Is
W - W [ j J , •.. ., ) > • • • •, Ug (x s i * • • • • * x s n ) | » where there are £ individuals and n goods and services. The amount of the jth good going to the ith individual is repre- 12
sented by and is the Ith person’s preference fane-
ft tion. Social welfare, thus, is considered to be some func-
6W. J. Baumol, Welfare Economics and the Theory of The State, (Cambridge: Harvard University Press, 1952), p . 127, n. 1.
tion of individual utilities which are in turn considered to
be some functions of the goods and services* In other words,
the welfare of the group varies in some way with, the indi
vidual welfares or utilities. Each individual’s welfare or
utility, in turn, depends in some way on the goods and ser
vices he receives or holds.
For any meaning to be attached to the equation It is
necessary to make some judgments as to the nature of the U
and W functions so as to give these functions form. The
individuals’ U, or preference, functions are given an or
dinal form rather than a cardinal form. The W function Is
also given an ordinal form. Very simply, such ordinal form means that statements may be made as to the IJ or W functions’
Increasing or decreasing but that statements as to the
amount of the Increase or decrease cannot be made.
Ordinality of the individual U functions means that any
Individual’s utility is measured by some arbitrary function
of the goods going to him per period of time,
U = U(x^, Xh ) . Subject to the requirement that
each of these functions change In the same direction with
any change in the quantity of some good going to the person, 13
any other arbitrary function, e.#g. U I V(x-j_, Xg, . ..., xn )> may be used as an index. In other words, the only restric
tion upon the choice among any number of arbitrary functions
is that each show utility as moving in the same direction
with any given change in goods received. Such definition Is
basic to the concept of ordinal utility and its use in show
ing whether the Individual is on a higher, a lower or the
same utility level as compared with an alternative position.
The ordinary graphic indifference curve analysis of
consumption is merely amor© special case of the general
utility index discussed above. Two dimensions measure the
various possible combinations of two goods. Utility is
measured in the third dimension. A surface is assumed to
show the utility of varying amounts of the two goods. Con
tours, or Indifference curves, represent those combinations
of the two goods possessing equal utility, or between which
the individual is indifferent.7 As movement is made up the
^Recently a new approach has been attempted which dis cards the concepts of indifference and utility. Instead a behavior map Is drawn up based upon the actual choices of individuals based on three assumptions. These assumptions are: (1) a consumer chooses a larger collection of goods rather than a smaller collection; (2) if a constuner chooses A over B and B over G, then he chooses A over C; and (3) every possible collection of goods is chosen in one and only one price-lncome situation. The behavior map resembles the .conventional indifference map. Instead of being indiffer ence curves, however, the contours represent boundaries be tween chosen and not chosen as compared with one particular collection. For a presentation see I. M. D. Little, TTA Reformulation of the Theory of Consumers’ Behavior," Oxford Economic Papers, Hew Series, No. 1 (January, 1949), pp. 90- 14 99; id., A Critique o.f. Welfare Economics, (Oxford: Oxford University Press, 1950), pp. 33-37 utility surface, a series of contours is crossed. It does not matter how each contour is labelled just so each contour line has a higher index than the contours lower down on the surface. No matter what the numbering system or scale hap pens to be, the results of the analysis^ would be the same.
It is of Interest to note that the study of methods of measuring utility has been given considerable attention In recent years.® Basically, they involve giving numbers of
®For a simple summary and interpretation see -A. A. Alchian, "The Meaning of Utility Measurement," can Bconomic Review, Vol. XLIII (March, 1953), pp. 25-50. ------4------rank to two choices and from this base placing values on
\ other choices. The subsequent measures deriyed from the two base choices are attained through the use of uncertain pros pects having certain probabilities. Of importance to this discussion is the fact that the measures for all subsequent choices depend upon what numbers or weights are given to the original two choices. The only requirement is that the most preferred of the original two choices be given the highest weighting.^ There are, therefore, a very large number of
®To Illustrate the basic idea, assuming G is preferred to B which is preferred to A, a weight of 1 can be given to A and a weight of 2 can be given to B. It Is then desired to find the weight or utility of C, or any other choice which may be preferred to both A and B. The individual can 15 toe given the choice between B as a certainty and of getting either A or C, each having certain probabilities, i.*®.* can take either B or the imcertain prospect of getting either A or 0. It is then necessary to get that probability of A or G which will make him Indifferent as between B and the uncertain prospect of A or G. The utility of B is then computed as being equal to the utility of chance of get ting A and the l-]c chance of getting C. The probabilities give a weighting to the utility of C. The eqxiation usedIs U(B) = p U( A) / (1-p) TJ( C) . If TJ (B ) , U(A) and p are known, U(C) is determined. potential scalings. Prediction as to what choices the indi vidual would make, assuming utility maximization, would be the same regardless of the scale chosen. These measuring techniques fit in with the above discussion of utility Indi cators, i_.e_. what Is Important Is the ordering or ordinal aspect of utility rather than absolute, or cardinal, meas urement of utility.
A theory of welfare economics can be constructed on the basis of the ordinal concept of utility. Conclusions can be reached which "would be true even if utility could be meas ured. "3-0
lOReder, ojc. clt. , p. 19.
Specification of the W function Involves the setting up of a rule as to how the individual utility, or T J , functions are to be combined to derive the social welfare. As stated, i_._e. W r W(TJ^_, Ug, ...., U s ), it can only be assumed that some kind of relationship exists between social welfare and the utility of the individuals making up the group. The 16 nature of the relationship furnishes subject matter for con siderable debate. Basically, the definition of the exact method of combining the individual utility functions re quires some value judgment. The value judgment commonly used in present day welfare economics is that social welfare is considered to be increased when any one person is better off, no one being worse off. More forma!!' y, the V/, or so cial, function can be said to increase when any individual
TJ function increases, no other U function showing a de crease. Nothing, however, can be said about the amount of the increase.
Two aspects of the welfare function, as presented to this point, deserve comment. One is the inclusion of all persons in the welfare function. Such inclusion involves a fundamental value judgment. While in the western world this standard is probably generally acceptable, with some excep tions for infants, etc., other cultures could conceivably limit the applicability to a more narrow group. An aristo cratic regime could place limitations on the basis of blood lines, some totalitarian regimes might place restrictions on the basis of party membership. Persons who are interested exclusively in their individual welfare habitually use such individual welfare as their norm with cavalier disregard for any effect upon others. An extreme Social Darwinism probab ly implies such a radical norm, i_.e_. it is the strongest who gain and should gaim even at the expense of others who 17
should not siirvive and be perpetuated anyway. Distribution of income so as to achieve optimum welfare according to the welfare function as specified to this point also deserves some attention. The individual utility indi cators cannot measure absolute utility changes, !..£• one group of goods received by an individual can be said to be superior to, or rank above, another group of goods but no statement can be made as to how much better one is than the
other. If no such absolute measures are possible for an in dividual, then comparison can not be made between Individ uals. No method can be devised for combining the Individ ual utilities. No statement can be made, therefore, as to optimum distribution of income, on the basis of the social welfare function as it has been specialized to this point. "Principles of income distribution cannot be deduced from the utility calculus either by the rules of logic or by em pirical demonstration. "H
"^Bergson, "Socialist Economics," o£. clt., p. 418.
To make possible any statements as to the optimum dis tribution of income, decisions must be made as to the ends income distribution is to serve. Whatever ends are chosen will give particular shape to the welfare function. The particular form of the welfare function determines, then, the optimum distribution of income according to the initial 18 evaluation of ends.
Bergson's contribution consists of pointing out the necessity of making explicit value judgments in setting up
the welfare function. To this point, three value judgments were used in specializing the general welfare function: (1) it is the individuals' preferences which are to count; (2)
these preferences can be evaluated on an ordinal basis; and
(3) the total welfare, W, is increased if any one person is
in a better position, no one else being worsened.
For application of the welfare function, as shaped by
the value judgments made to this point, to economic prob
lems, these problems can be broken into two areas. The best
allocation of a given quantity of resources among different uses in consumption and production is one area. The other is the welfare implications of reorganizations in which
changes occur in the resources available or in the total
output of an economy.^2 The welfare function can now be ap-
-*-^T. Scitovsky, "A Mote on Welfare Propositions in Economics," Review of Economic Studies, Vol. IX (Movember, 1941), pp. 77^88^1 Bc'iToVsEy places economic welfare propo sitions into these two general areas. plied to these problem areas. Recent criticism of the for mulations will follow. 19
B. Efficiency of Production and Consumption: The Marginal
Conditions for an Optimum By applying several assumptions which further special ize the welfare function, especially in making it more static, standards of efficiency in production and consump tion can be derived. These assumptions are: (1) each indi vidual has a utility or preference function which Includes all choices he must make including that between leisure and work. Each individual also owns a certain quantity of each good and each service, some of these quantities being zero;
(2) each productive unit or firm has a production or trans formation function given by the existing technology; (3) consumers are considered to be sovereign in that no re straints, eufa* sumptuary laws, are imposed on consumers; (4) an individual’s satisfaction depends only on the way in which he consumes and uses his inputs, :L.e. his satisfac tions are independent of other’s consumption and work pat terns; (5) inputs can be used indifferently in a number of different uses; and (6) products or inputs of any one class are assiuned to be homogeneous. These assumptions may be criticized, of course, on em pirical grounds, i,.e. disagreement with the objector’s view of the facts of life, or on ethical grounds. Assumptions
(3), (4) and (5) seem particularly vulnerable. Varying these assumptions would give a different shape to the wel fare function. ^ 20
4^For a more extended discussion of these assumptions, see Samuelson, o j d . ci t . , pp. 222-228.
With the welfare function as defined and as limited by the given constraints, the conditions for maximum welfare can be derived. Maximum welfare can be defined as a posi tion or condition in which it is impossible to improve any one person without putting someone else in a worse position.
Several writers using the same basic assumptions and defi nitions have derived the marginal conditions for maximum welfare. While the number of conditions and their specific formulations vary, they are equivalents. This study will use the seven marginal conditions for maximum welfare stated by M. W. Reder.14
-*-4Reder, oj3. cit., pp. 21-38. Other writers include A. P. Lerner, The Economics of Control, (Hew York: The Mac millan Co., 19467* PP* 72-77; Oskar Lange, ,!The Foundations of Welfare Economics," Econometrlca. Vol. X (1942), pp. 215- 228; A. Bergson, "A Reformulation of Certain Aspects of Wel fare Economics," ojo. clt.
1. The marppinal rate of substitution for any two pro
ducts must be equal for any consumers consuming both.
(Allocation of Products)
It is Intuitively rather obvious that if any two con sumers have unequal marginal rates of substitution for two products both can Improve their position by trading. Assum ing Individuals A and B each have certain quantities of pro- 21 ducts x and y with A Ts marginal rate of substitution being 1 x for 2 y's and B's being 1 x for 3 y ’s, then both could benefit by trading. A could give up x ’ s for B's y's at any barter rate between 1 x for 2 t/;'s and 1 x for 5 y's. The optimum would be reached when thoir marginal rates o f sub- 1 5 stitution became equal.
15iphis and all following numerical examples follow Bouldlng, "Welfare Economics," ojo. clt. For a mathematical proof of these marginal conditions see Appendix A.
Geometric analysis of this problem helps in clarifi cation and at the same time points out a basic problem in welfare economics. Figure 1, a standard diagram, shows preference surfaces for two Individuals, A and B. Together they have AM (zBN) of x and AN (:BM) of y. ^ ^-s ^he point of origin for A's preference surface represented by indif ference contours 1^, 11^, III^, signifying higher levels of preference, as measured by some index of utility, in that order. B is the point of origin for B ’s preference surface as represented by indifference contours Ig, IIg, Illg, sig nifying higher levels of preference in that order. Assuming that A possesses AF of x and AG of y, B then possesses BH of x and BJ of y. This point, R, would be at one intersection of 1^ and Ilg at which the marginal rates of substitution are unequal, i.e.. the slopes of the indif ference curves are unequal. Trading could take place along 22
Figure 1
Utility Surfaces
N
0
Y nr
A 23
I ^ until point P would bo reached. At point P, A would have the same marginal rate of substitution between x and j as B would have,’" i_,e_. is tangent to Illg. B would be better off than he was at the initial point, having moved from in difference curve IIg to Illg. A would be just as well off as initially, remaining on indifference curve IA . Trading also could take place moving along Ilg until point P* is reached. A would be better off than in the initial posi tion while B would be just as well off as initially. In general, if 00' represents the points of tangency of A's and B*s indifference curve systems, they would be hotter off on some point on CC than on a point off GO'. In the specific example above, a movement from R to any point on CC1 between
P and P' would represent an Improvement for at least one of the traders without hurting the other, or an improvement for bo th. Figure 1 can be used to illustrate the nature of con temporary welfare economics. It is known that along CC’ the marginal rates of substitution for the two consumers as be tween products x and y; are equal. For an optimum to be achieved, therefore, it is necessary that exchange result in some position on CCr . Some point on GO’ Is better for one, without hurting the other, or for both, than any point off
CCr. Movement along CC', however, represents a betterment of one individual at the expense of the other. Because of the impossibility of interpersonal comparison of utility, 24 points on CCT cannot be compared with one another. In other words, no welfare judgment can be made as to income distri bution since each point on C C represents a different dis tribution of existing output benefiting one and hurting the o ther. 2. The marginal rate of transformation for any two pro ducts must be the same for any firms producing; both.
(Allocation of Production) Assuming the possession of a fixed amount of various inputsj any firm has a choice of the amounts of any products it wishes to produce, i.*®.* it has a transformation function which limits the amounts of the various products which can be produced with the inputs. Assuming production of two products, every firm producing the two products must have the same marginal rate of transformation. The marginal rate of transformation is the amount of one product which is lost
If resources are shifted to produce an infinitesmal addi tional amount of another product.
If the marginal rate of transformation between products x and 2 1 is one x equal to two - This reallocation of production should continue until the marginal rates of transformation are equal, !_•£• just as much is lost in shifting as is gained. 3 • marginal rate of transformation be tween any pro duct and any input must be the same for any two firms producing the produot and using the ingut. (Allocation of Inputs) Any input must be allocated between firms so as to max imize the potential output. This is achieved when the ad ditional output gained from using the last infinitesmal amount of the input is the same for all firms. In other words, the marginal physical productivity of an input must be equal for all firms using the input to produce the same product. If firm A has a marginal rate of transformation of one unit of input a for two units of product x while firm B re ceives three units of x •f’or> one unit of a, input a should be shifted to firm B. The first unit shifted means a loss of two x'3 a 'i: A anc^ a gain o f three at B, or a total gain of one. The reallocation should continue until the marginal rates of transformation are equal and the gains just balance the losses. 4. The marginal rate of substitutlon for any two Inputs in the production of any product must be the same for any firms producing the product and using these Inputs. (Input Substitution) 26 Inputs must be so combined as to achieve the optimum output. Mien inflnitesmal amounts of inputs can be substi tuted for each other, output remaining the same, at the same rates in all firms using the inputs to produce the same pro duct, this optimum can be achieved. To demonstrate this marginal condition an assumption can be made that firm 1 can substitute one unit of input a for two units of input b, while firm B can substitute one unit of input a for three units of input b, both outputs re maining constant. One unit of a can be shifted from firm A to firm B, while one unit of b_ is withdrawn from firm B to other uses and two units of b shifted from firm B to firm A, outputs remaining constant in firitis A and B with a gain from other uses of input b. Such substitution should continue until the marginal rates of substitution are equal for the two firms. At such point of equality a fixed output can be produced with a minimum use of inputs. Any infinitesmal shifting of inputs will result in output gains just being offset by output losses. 5. The marginal rate of transforma11on between any two products must be equal to the marginal rate of substi tution of the products for any two consumers consuming both. (Substitution of products) Some relationship must exist between the conditions of production and consumption for an economic welfare optimum to be achieved. Products must be substituted for each other in production, using the same number of Inputs (marginal rate of transformation), at the same rate as the consumers can substitute these products with the same amount of satis faction (marginal rate of substitution). Proof of this condition can be shown by using a situ ation in which the community can marginally transform one unit of product a into two units of product b using the same amount of inputs, while some consumer can substitute one unit of product ja for one unit of product b, satisfac tion remaining constant. The consumer could consume one less of product a and two more of b with the same amount of resources used by the community but the consumer’s satisfac tion increased. Such substitution of products should be carried on until the marginal rate of transformation for the community is just equal to the marginal rates of substitu tion for all consumers. At this point the last inflnitesmal switch of products will leave resource use constant and the consumer’s satisfaction constant. 6. The marginal rate of substi tutlon between the amount of any product received in aiding in 1ts production and the time spent in rendering such ai d for all input owners must be equal to the marginal rate of transfor mation between input time used ai ding in production and the product for any input used. (Leisure and Product substitution) The use of input time related to the rate of remuner- 28 ation and the marginal rate at which the input is trans formed into the product is of great importance in reaching a welfare optimum. If the marginal rate of-substitution be tween working time and income, measured by some numeraire, Is equal to the rate of remuneration which in turn is equal to the marginal rate of tranformation between input time and the product, welfare would be maximized. An Illustration demonstrating this condition can start with the assumption that the rate of remuneration and the marginal rate of transformation are both one time unit of input a for three units of product x, with a marginal rate of substitution of one time unit worked equal to two units of product x on the part of input a. It then would be ad vantageous for the Input to substitute leisure time for working time. The input requires only two units of x to substitute for one time unit of work. But, by working one more time unit It can receive three units of the product. The rate of remuneration, thus, would more than compensate for the displeasure of a unit more of working time. The producer who made the payment would be at least as well off as he was prior to the increased working time. Such sub stitution should continue until the displeasure incurred through the last moment of working time Is just offset by the remuneration received for such work, with the producer not suffering a loss. While this analysis seems to apply primarily to labor, 29 it applies as well to any input that is used. The input could he used in rendering services to its owner directly rather than being hired out at a rate of remuneration. With labor, the rendering of direct services usually takes the form of leisure time activity. 7 * The marginal rate of subs ti tutlon be tween resource control (supply of funds, as measured by 3 ome numer aire ) at any two different moments of time (t^ and t_g) must be equal for any pair of individi-ials or firms or any individual and firm. (Resource control through time)Is ■'■6Por a more detailed breakdown, see Appendix A. Any individual possessing a certain control over re sources, i_.e_. fund of purchasing power, can allocate his expenditure through time by borrowing or lending. Similarly every firm can control inputs through time to receive a cer tain flow of outputs over1 time. Every pair of individuals and/or firms will at the optimum have equal marginal rates of substitution between resource control for any two dif ferent moments of time. For verification of this condition it is assumed that some individual A has a marginal rate of substitution of one unit of x at t^ for 1.3 units of x at while some other individual or firm, B, has a marginal rate of substitution 30 of one unit of x at t^ for 1.4 units of x at t2 * For an optimum to be reached it would be necessary for A to lend units of x to B. More than 1.3 units of x could be deliver ed at tg by B for each unit borrowed at t^_ without being worse off, while A would be better off. Such lending and borrowing should continue until the intertemporal marginal rates of substitution are. equal. It should be noted that no conditions of uncertainty are assumed to exist* The same was true in the consideration of all the marginal conditions which have been discussed. The seven listed marginal conditions, it should be pointed out, are primary conditions which could determine a minimum as well as a maximum. To determine the maximum, certain second order conditions must be met. These refer to the nature of the preference and transformation functions. Stated simply, the transformation functions must be contin uous and concave to the origin and the preference functions must be continuous and convex to the origin. In other words, in the region of operations there must exist a dimin ishing marginal rate of substitution for each consumer, a diminishing marginal productivity for each input and a di minishing marginal rate of transformation of one product for another for each firm. In addition to the marginal conditions and the second order conditions, Hicks specifies what he calls the total conditions. These require that no abandonment of the pro 31 duction of one product or service, or the introduction of a new one would improve anyone's position. ^ 17J. R. Hicks, "Foundations of Welfare "Economics," Economic Journal, Vol. XLIX (December, 1939), pp. 69 6-712. Fulfillment of these marginal conditions when goods and services are priced at marginal cost is the basis of mar ginal cost as a pricing standard. Where prices are set by some authority, marginal cost is the pricing principle which results in most efficient use of resources. Where no authority sets the price, the meeting of the marginal conditions under perfect competition results in its being used as an efficiency standard. The seven mar ginal conditions are met under perfect competition because the prices of all products and inputs are given to pur chasers and the rate of remuneration is given for inputs. The first marginal condition is met because under per fectly competitive conditions any two products will have prices which are given identically to all consumers. Each consumer adjusts his expenditures so that the marginal rate of substitution between products is equal to the ratios of the prices. As the ratios of the prices are the same for all consumers, the marginal rates of substitution will be the same for all consumers. As competition results in given sale prices to All pro- 32 ducers, the second marginal condition is satisfied. In the firm’s maximizing position, the marginal rate of transfor mation will be equal to the ratio of any two products’ prices. As the price ratios are the same for all firms, the marginal rates of transformation will be the same. Firms operating under perfect competition have input and product prices given to them. At optimum output each firm will have the marginal rate of transformation of input into product equal to the ratio of the input and product prices. Each firm having identical prices means that all firms_ have the same marginal rate of transformation, satis fying marginal condition three. Condition four is the classic equi-marginal producti vity ratios concept. All firms will make the marginal rates of substitution between inputs equal the input price ratios. As perfect competition results In Identical given input prices to all firms, all firms will have equal marginal rates of substitution for the inputs. Marginal condition five is central to the case for mar ginal cost pricing. Under pure or perfect competition all prices are equal to marginal cost at each firm’s profit max imization output. The marginal rates of transformation for all firms producing any two products are equal to the ratios of the marginal costs. At the same time the marginal rate ^®If the marginal cost of x Is $1 and that of £ i-s $2, 33 the marginal rate of transformation is g, the inputs used to produce the marginal unit of x are able to produce |r of the marginal -unit of of substitution as between two products is equal to the ratio of thair prices for all consumers, If they are in their optimum consumption position. The marginal rate of transformation for all producers is equal, therefore, to the marginal rate of substitution for all consumers.^ ■^It Is at this point that the question has been raised as to whether an equal proportionality between marginal cost and price for all firms in the economy would not result In the satisfaction of the seven marginal conditions, ji.e,. if the same degree of imperfection in all markets would not satisfy the marginal conditions. See Ragnar Frisch, nThe Dupuit Taxation Theorem,” Econometrica, Vol. VII (April, 1939), pp. 145-150. However, as Lerner and Reder point out, the only pro portion which would satisfy all the requirements would be unity. Otherwise there would be restriction In the use of some input, therefore not satisfying condition six. Lerner, op. cit., pp. 98-103; Reder, ojo. cit., p. 42, Under conditions of perfect competition, marginal con dition six is satisfied because the rate of remuneration is given to both purchasers and owners of Inputs. Producers will hire Input time until at the profit maximization posi tion the marginal rate of transformation of intput time into the product is equal to the rate of remuneration. Input owners, in turn, will hire out the Input so as to equate the rate of remuneration to the marginal rate of substitution between income and the contribution of direct services by the input to the input owner. In the case of labor the di rect services can be considered as leisure. 34 Optimum resource control over time is achieved under perfect competition bj'- the interest rate which is given to all individuals and firms. Each will borrow or lend until the ratio of a dollar at the earlier time (t-^) to the dollar plus interes t at any later time (tg) is equal to the mar ginal rate of substitution of resource control between t-^ and tg. Marginal condition seven, thus, is satisfied under perfect competition. One point of considerable importance should be noted. Perfect competition i3 considered as leading to the optimum only because the seven marginal conditions are satisfied. These marginal conditions are satisfied if prices of all products are equal to their marginal costs. In perfectly competitive markets prices are equal to marginal costs. Therefore, perfect competition leads to optimum allocation of resources and goods. Perfect competition is not set up as an ideal for its own sake or because prices equal minimum average costs.^ ITor is perfect competition held as desir- fact, in the short-run, optimum prices may be above or below minimum average costs. able because pure profits do not exist in the long-run. Prom the standpoint of welfare economics, any desirability of perfect competition results from the satisfaction of the seven marginal conditions for an economic optimum, thus re sulting in optimum allocation of resources and goods. 35 G. Reorganizations: The Compensation Pr’inciples In the discussion of the welfare function above it was assumed that economic welfare is considered to be Increased if any one person Is bettered and no person is worsened. Most economic policy decisions, however, involve putting some people in better positions at the expense of others who lose. Tariff changes are good examples of such policy de cisions. The compensation principle was devised for the evaluation of such changes. The compensation principle is a device for relating the individual U functions so as to de termine what is happening to the social, or W, function when there is a basic change in the economy which affects the x ' s in possession of individuals, i,.©,. changes the quantities of goods and services received by individuals. In its simplest form the compensation principle states that If those who benefit from a reorganization are able to pay bounties to those who lose in sufficient amount to com pensate the losers for their loss, welfare is considered to be Increased, If compensation is actually paid. "This for mulation is certainly not debatable except perhaps on a philosophy of systematically denying people whatever they want. n ^-'- 21k . J. Arrow, Social Choice and Individual Values, (Sew York: John Wiley and Sons, 1951), p. 34. This formulation was extended by Kaldor and Hicks to 36 Include situations where compensation is not actually paid.22 The rule states that if a reorganization is effec- oc N. Kaldor, "Welfare Propositions in Economics," Econ omic Journal, Vol. XLIX (September, 1939), pp. 549-555; Hicks, "The Foundations of Welfare Economics," o j d . cit. ted, e_.£. change In tariffs, and those who gain could com pensate those who lose for their loss, welfare is considered to be increased. Whether compensation should be paid or not is, according to Kaldor, "a political question on which the economist, qua economist, could hardly pronounce an opin ion."25 PS Kaldor, ojs. cijb., p. 550. A difficulty inherent in the Kaldor-Hicks formulation was pointed out by Scitovsky.2^ It is possible that a con- 24 T. Scitovsky, "A Note on Welfare Propositions In Economics," oj). cit.; id., "A Reconsideration of the Theory of Tariffs," Review of Economic Studies, Vol. IX (1942), pp. 89-110. tradiction might arise from the use of the Kaldor criterion. A reorganization may take place which moves individuals in an economy from position A to position B, each of these rep resenting a certain quantity of goods distributed in a cer tain way. Under some conditions it is possible that com pensation payments in position B will make everybody at 37 least as well off as in position A and at the same time com pensation payments in position A could make everyone at least as well off as in position B. Thus the contradiction, A would be deemed superior to B and B deemed superior to A. To clear up such difficulties a new compensation principle, really a double criterion, was proposed by Scitovsky. Under the double criterion, potential results from com pensation payments must be viewed from both positions. With a potential reorganization resulting in a change from po sition A to position B, it would be necessary first to see whether compensation payments from position B would make some individual better off than in position A. It would then be necessary to see whether compensation payments from position A would make someone better off than in position B. If the first is possible and the second Impossible, position B could be said to be superior to position A. If the second is possible and the first impossible, position A would be considered superior to position B. If both are possible or both impossible, no welfare judgment could be raade,^^ ^Sprom each position one question is asked, "Is it pos sible through a series of compensation payments to Improve someone as compared with the alternative position?" The possible answers to the two questions are: (1) from B it is possible to improve over A, or it is impossible; (2) from A it Is possible to improve over B, or it Is impossible. If answer (1) states that it is possible from B to Improve over A and answer (2) states it Is impossible from A to improve over B, B is superior to A. If answer (2) states that it Is possible from A to improve over B and answer (1) states It is impossible from B to improve over A, A is superior to B. 38 If answer (1) states it is possible from B to improve over A and answer (2) states it is possible from A to improve over B, no welfare judgment can be made. If answer (1) states it is impossible from B to improve over A and answer (2) states it is Impossible from A to improve over B, no welfare judgment can be made. Criticisms have been made of these concepts. These criticisms will be discussed in the following section in which the welfare concepts which will be used in later parts of this study will be pointed out. D. Welfare Measures To 3e Used In "Phis Study The setting of public utility prices will be a major concern In the remainder of this study. To set any prices some standards must be applied. As this study is concerned with the achievement of economic efficiency, welfare econom ics should disclose the principles which apply. Several economic welfare concepts have been presented In this chap ter. Criticism of these concepts must be considered in determining their validity. A rigorous logical examination of the general welfare function and its particular forms as put forth in the com pensation principle was made by K. J. Arrow.Arrow an- 26Arrow, o]o. clt. alyzed a basic problem in very general terms. The problem was that ..of finding a method or procedure by which a social ordering of various alternatives can arise out of the order- 39 ings of these alternatives made by all the individuals. In the economic sphere this means the choice of some social welfare function of the Bergson type, some method of relating the individual IJ functions to determine the W, or social, function, thus making the determination of the social optimum possible. It also includes the compensation principles as particular forms of the social welfare func tion. Arrow’s basic conclusion was that no social welfare function can be found which meets five ’’apparently reason able conditions,”2,7 and which is also transitive.28 27Ibid., p. 26. 28 By transitivity Is meant that If X Is preferred to Y and Y preferred to Z, X is preferred to Z. The five conditions applied to the nature of the indi vidual or social orderings are: (1) Out of any possible number of alternatives being ordered there are at least three alternatives which can be ordered by individuals in any logical way. (2) If any alternative rises or remains the same in every individual's ordering, it will rise or at least remain the same in the social ordering. Thus, If A Is preferred to B in the original social ordering, it will also be preferred in the new ordering arising from A's moving up or remaining the same in all individual orderings. (3) The social choice made among various alternatives 40 depends only upon these possible alternatives and not upon any alternatives outside the aroa of choice. This means that irrelevant alternatives have no effect, on the social choice among those choices which are possible. Thus, if individuals have a possible choice among alternatives A, B, C and D and we know what the individual choices are, the social ordering of A, B and C will remain unchanged even if D no longer remains a choice for some reason, the death of a political candidate immediately following the vote would not affect the ordering of the remaining candidates. (4) The social welfare function will be an indicator or reflection of the true individual desires. In other words, the social ordering is determined not by social code or custom but arises from the individual desires. (5) The social welfare function is not to be dictator ial in the sense that the choice of one individual deter mines the social choice regardless of all other individual choi ces. Using symbolic logic Arrow showed that no general social welfare function could be set up satisfying these conditions and possessing transitivity. 29 By restricting ^Arrow, op. cjt. , pp. 51-59. the individual orderings to single-peaked preferences, how ever, a social welfare function could arise satisfying these 41 five conditions.50 Single-peaked preference would mean that 5QIbld., pp. 75-80. an individual prefers one alternative very much over all others and that there is decreasing preference for others with more distance or more dissimilarity in both directions from the most preferred. The result of this study has been to put the social welfare function viewpoint into discard as a basis for policy recommendations unless more specific restrictions are placed on the individual orderings. Very similar results befall the compensation principle when no compensation is paid. Arrow showed that the Scit- ovslcy compensation principle would not be transitive when there were three or more alternatives.31 Only when there 53-1 bid. , pp. 44-45. were two alternatives would it have any value. The case of only two alternatives is rarely found. Because of the deduction of the marginal conditions of efficiency of production and consumption from an elementary form of the social welfare function, the question arises as to the degree of validity of these conditions. The seven marginal conditions cannot be, of course, the complete con ditions for the optimum. For the determination of the opti mum, the marginal conditions must be supplemented by condi- 42 tlons relating to distribution as between individuals. Ac cording to Samuelson, if later attention is paid to distri bution considerations and if it is further assumed that social welfare is increased when individuals become better off and that more goods and less inputs are desirable, these conditions have validity. Such validity exists in the sense that if any of the marginal conditions are not met, "there exists a physically attainable position that makes everyone better off."3S 3^Paul A. Samuelson, "Comment," A Survey of Contempo rary Economics Volume II, ed. by Bernard P. Haley, (Chicago: Hi"chard D. Irwin, Inc., 1952), p. 38. Fortunately, this study need make no use of the general welfare function. The compensation principle when used will assume actual compensation payments being made. The meas urement of the ability to make such payments will be through Hicks’ compensating variation which will be explained when it is used. The seven marginal conditions will be used as standards of efficiency through the emphasis on pricing at marginal cost. It will be assumed that prices varying from marginal cost indicate a possible movement to a position in which everyone is better off, or at least some people are better off and no one Is worsened. Emphasis should be put on one important consideration Involved in the welfare principles. This consideration is 43 til© static nature of tiie assumptions from..which the welfare principles were derived. Of particular significance are the assumptions of a given technology, given preference func tions and possession of a given amount of goods and services by the individuals. Meeting the seven marginal conditions would maximize welfare under the given conditions. However, it is quite possible that violating one or more of these marginal conditions may result in greater welfare in the future than would be possible under rigid adherence to the marginal conditions for maximum welfare. The introduction of such dynamic considerations requires some modification of the principles. 'V2)^ 3?Samuelson, Foundations of Economic Analysis, op. cit., p . 253. CHAPTER III LAWS OP COSTS AMD RETURNS In the preceding chapter pricing at marginal cost was set up as the principal guide to optimum pricing. The pos sibility of reaching such an optimum under various condi tions is of great Importance to this study. Before investi gating equilibrium under different cost and supply patterns it is necessary to give some attention to these cost and supply patterns. This chapter has as Its purpose the ex plicit statement of cost and supply patterns under various possible conditions. The chapter to follow will contain a consideration of equilibrium positions and their evaluation from the standpoint of the welfare principles. Presentation of the material In this and the following chapters Is justi fied on the grounds of minimizing confusions. Terminology and definitions will be developed which will be used In later portions of the study. Without a rigorous considera tion and explicit definition, confusions as to meaning would develop In later sections of this study. Of special signif icance in this regard is the relationship between, firm and Industry supply and cost patterns. The problem of cost and supply patterns will be broken into two classes. These are the firm cost and supply pat terns and the industry cost and supply patterns. The pos sible patterns and their causes will be considered for both firm and industry. Before either of these patterns can be 44 45 analyzed it will be necessary to consider the firm's pro duction function and input adjustments as the groundwork from which to derive firm and industry cost patterns. A. Firm Cost Patterns 1. Input Adjustments A firm is here defined as a managerial or accounting unit.^ Each such firm possesses a given production func- ^The problem of defining a firm has been deliberately slighted as being outside the scope of this study. For a discussion of problems Involved in such a definition see R. H. Coase, "The llature of the Firm," Economlca, Vol. IV N.S. (1937), pp. 386-405 reprinted In Readings in Price Theory, ed. by George J. Stigler and Kenneth E. Boulding, (Chicago: Richard D. Irwin Inc., 1952), pp. 331-351. tion. The production fiznction is defined by the relation ship between output per unit of time and all possible com binations of inputs per unit of time under the given techno logical possibilities.^ Symbolically the production func- p The term inputs is used because of the general nature of the meaning as anything purchased for use in the produc tion process. The more conventional term, factors, connotes the traditional four way classification. tion can be presented as P = f(a, b, c, ...... n) where P Is the output per unit of time and a, b, c, ...., n are amounts of various inputs per unit of time. Some of these inputs may be zero, of course. 46 Graphic presentation of the production function of a firm commonly employs a two input model. This is more easi ly understood and followed by those unfamiliar with the more rigorous mathematical analysis. The output resulting from various combinations of the inputs can be illustrated by a two dimensional contour map equivalent of a three dimen sional model. In Figure 2 the series of contour lines rep resents the intersection of a surface by planes cutting the output axis perpendicularly at various heights. Each con tour line represents a given output and is labelled accord- ingly. P q thus represents a fixed output, P q , which can ^This technique can be found in many textbooks. One of the best is Kenneth E. Bouldlng, Economic Analysis, (New York: Harper & Bros., 1949), pp. 671-698. For a more rigor ous treatment see R. G. D. Allen, Mathematical Analysis for Economists, (London: Macmillan & Co., 1948), pp. 369-371, 284-288. be produced by any one of many combinations of quantities of inputs si and b. Similarly for contours P_^ to P^. These can be called constant product product curves. Certain assumptions are implied by the form of the pro duction function illustrated in Figure 2. First is the as sumption of continuity. That Is, inputs a and b can be sub stituted for each other in infinitesimally small amounts, product remaining constant.4 The second assumption is Im~ 4For a more rigorous definition of continuity, see 47 Figure 2 The Production Function 48 Allen, ojc. clt. , p. 269. plied by the convexity of the constant product curves to the origin. Only if diminishing marginal productivity operates is the constant product curve convex to the origin. As in put a is substituted for input b, output remaining constant, it takes more and more of input a for a given unit of input b. A third assumption is that the various outputs are not dependent on input quality variations, i_.e_. inputs are as sumed to be homogeneous. Restrictions can be placed on the production function by altering the curves accordingly. Discontinuity, or limi tations on proportions in which inputs can be combined, can be illustrated by corners on the constant product curves. With only two inputs, Figure 2 represents the long-run situ ation in which any variation in inputs a and b can be made. In the short-run in which one of the inputs is fixed, the possible output and input combinations can be found by erecting a perpendicular to the pertinent axis at the point representing the fixed amount of input, Q_._g. point b_Q in Figure 2. It is to be expected that output along such a perpendicular would follow the law of non-proportional re turns. As the emphasis of this study is on long-run results, it is not necessary to investigate this law.^ ^The law of non-proportional returns has often been treated superficially. The best statement probably is 49 John W. Cassels, "On the Law of Variable Proportions," Explorations in Economics, (New York: McGraw Hill Book Co., 1936), pp. §"23-236, reprinted in Readings in the Theory of Income Distribution, ed. by William Fellner and Bernard F. Haley, (London: George Allen & Unwin Ltd., 1950), pp. 103- 118. 2. The Least Outlaw Combinations Given the physical production function, it is possible to determine input relationships at various outputs with any set of given input prices. These input relations can be ex pected to be optimum for the firm. The firm, in other words, can be assumed to combine inputs to produce the maximum pos sible output with a given expenditure, or, conversely, to minimize expenditures at a given output. A firm having a production function illustrated In Figure 2 will also have a set of given input prices. With the given input prices the firm will have a series of poten tial total expenditures. In Figure 2 let 0a^ be a possible total expenditure if only input a were purchased. Let 0b^ be the same total expenditure if only Input b were purchased. The slope of b-j_a^ then equals the ratio of the price of b to the price of a. Further, along b^a^ expenditure remains constant. Thus, it can be called a constant outlay line.® ®The above is true if the firm is an atomistic pur chaser of inputs with input prices given to the firm. If the firm has monopsonistic powers in the input markets, the constant outlay curve would be concave to the origin, i.e.. the price of an input would increase as more would be purchased. 50 With movement along bqab, input relationships and out puts vary. For the given expenditure Oa-^(IOb^) the maximum possible output is represented by the highest constant pro duct curve touched by b^a^. This curve is P-j_ which is tan gent to the constant outlay line. At the point of tangency the marginal rate of input substitution (slorje of constant product curve) is just equal to the ratio of input prices (slope of constant outlay line). This is stating in another way the old rule that the ratios of price to marginal pro ductivity for all Inputs should bo equal to maximize output TV for a given expenditure. ^This problem is the equivalent of maximizing P - f(a, b) subject to constraint apa / bpb - constant. The constant is equal to the total outlay and pa and pb are the respective input prices. Using the method of the Lagrangean undetermined multiplier (-m), there is a new function, G = f (a, b) -m(apa / bpb ) . (1) For a maximum, = f a( a, b) - rnpa = 0 and (2) = fb (a, b) - mpb = 0 . (3) Rewriting (2) and (3), 1 _ Pa_ Pb m “ fa (a, bj “ fb(a, b) * The undetermined multiplier (m) is the factor of propor tionality of price and physical marginal productivity with fa (a, b) and fb (a, b) being the physical marginal produc tivity of inpxits a and b, respectively. Line 0A, the scale line, represents the input adjust- Q ments in the long-run as scale is changed. With all in- 51 * % i t h a production function P — f ( , Xg, . ... f x^), a particular scale of plant Is said to exist when one or more Inputs have fixed values, x^ z, xfj? (i — 1 , 2 , . ... p j ) or P — f (x-^, ••••> j ^ *•... ? ^n ^ * Of two scales , (1) P — f (x^, . ... p j * xk> •... p ^n^ and (2) P _ f (x^, x j * f * * * * * xn ^* (1 ) > (2 ) if xj ? x? (i = 1 , 2 , j) 1 2 and xi"l > x. x for at least one —i. puts variable and divisible, the optimum adjustment can be made at any output. The short-run adjustment with an input fixed can be illustrated by assuming input b to be fixed in supply at bQ. Only at output Pg would the Inputs be adjust ed optimally. At any other output total cost would be high er along the bg line than along the scale line OA. General izing, it can be said that, for any scale of plant, short- run total costs equal long-run total costs at only one out put. At any other output the short-run total costs are greater. Figure 2 illustrates a scale line which is linear and which cuts the origin of the coordinates. There is no reason why any scale line should necessarily possess these characteristics. Instead of being a straight line, the 52 scale line could follow a twisting, devious path or any of many possible patterns. A linear scale line indicates the maintenance of the same input proportions at all potential outputs. The proportion could be determined by the slope of the scale line. As OA has a slope of approximately one, approximately bisects the quadrant, inputs a and b would be added on an approximate one to one ratio. Any other ratio could conceivably exist, however. If the scale line varied from the straight line pattern it would indicate variation of proportions in which a. and b are used. Sim ilarly, there is no necessity for the scale line to cut the origin. It could begin at some positive value of either a or b. Such scale line origin could occur if there were some minimum amount of one of the inputs necessary to produce an infinitesimal amount of the product. Of more importance to this study is the rise in the scale line as it goes up the production surface. The ques tion Involved is the rate of increase of output as inputs are increased. For analytical purposes it is assumed that the scale line follows the linear pattern illustrated in Figure 2. The production surface can be cut along the scale line perpendicularly to the aOb axis and a profile view would result. This profile view shows the increase in out put as the two inputs are varied optimally. Each potential output could be plotted on a graph above the amount of input a which could be purchased with the ex- 53 penditure incurred in producing that output optimally. For example, output P-^ was produced optimally with a total ex penditure sufficient to purchase a-^ of input a. A chart could be made plotting a point at a height P-^ above a^. This could be done for each of the series of potential out puts, £ . £ . Pg over ag, Pg over a^, etc. The amount of a at any point multiplied by the given price of a would make the horizontal axis a cost axis. The result would be a curve relating output to total costs, the firm making adjustments optimally. Such a curve would show the cost patterns which derive from the firm's production function. 2. Patterns of Costs to Scale A chart plotted according to the description of the previous section would have costs along the abscissa and output along the ordinate. Conventional usage, however, has output along the abscissa. To avoid confusion this study will use the conventional presentation.® ^Derived from Jacob Viner, "Cost Curves and Supply Curves," Zeitschrift fur Hatlonalokonomie, Vol. Ill (1951), pp. 23-46; reprinted in Readings in Price Theory, ed. by Kenneth S. Bouldlng and George J. Stigler, (Chicago: Richard D. Irwin Inc., 1952), pp. 198-226. Figures 3, 4 and 5 give the three basic long-run cost patterns which may be expected. Figure 3 shows the total cost pattern (LTC) characterized by a decreasing slope at low outputs and later changing to an increasing slope. As 54 will be seen later, the region of increasing slope is of greatest interest in this study. Figure 4 shows the total cost pattern (LTC) having constant slope. Figure 5 shows the total cost pattern (LTC) having a decreasing slope. LTC in Figure 3 is made up of a series of short-run total cost curves, ©.•£• ^TC-^, STCg. The smoothness of LTC derives from an assumption of continuity, _i.e_. it is assumed the scale of operations can be varied in infinitesimal amounts. Two short-run total cost curves are drawn to show the general relationship of short-run and long-run curves. Each short-run cost curve intersects the ordinate above zero. The point of intersection measures total fixed cost. OS represents the total fixed cost with a scale of plant having total cost curve STC-j_. The larger the plant, the greater the fixed cost. Each short-run cost ourve touches the long-run cost curve at only one point. At any other output short-run costs are greater. This derives from the analysis of Figure 2. In Figure 2, with inpxxt b fixed at bQ total cost was higher than along the scale line at every output except Pg. At output P3 total cost with input b fixed at bQ was equal to total cost with all input quantities variable. Output OM of Figure 3 could be considered the equivalent of the inter section of the scale line and output Pg in Figure 2. Average and marginal cost patterns can be derived from the total cost curve by considering two different slopes. Figure 3 Increasing Costs to Scale LTC 5TC 5TC, O u t p u t l a c SAC SAC, 56 Figure 4 Constant Costs to Scale | Output o R 57 Figure 5 Decreasing Costs to Scale STC ST C LTC Ou t p u t SMC SAC SMC SAC LAC LMC Ou t p u t The slope (first derivative) of the total cost curve at any output is the marginal cost at that output. Therefore, LMC in Figure 3 decreases through a certain region, then begins increasing. In Figure 4 LMC is constant because LTC has a constant slope. In Figure 5 LMC decreases through the en tire region of possible output. Average cost for any output is equal to the slope of a straight line from the origin to the point on the total cost curve representing that output. Thus, In Figure 3 the slope of OF gives the average cost at output OR. Similarly for any other output. It should be noted that the slope of ON is equal to the slope of the total cost curve at output OR. OR Is the output, therefore, at which marginal cost Is equal to average cost. Average cost at that point is a minimum. l^Let G(x) be the total cost function. To minimize average cost, 0 ax Rewri ting,G T (x) = J112L) or marginal cost x equals average cost at minimum average cost. Average cost in Figure 3 decreases, reaches a minimum and then increases in accordance with the slopes of the respec tive lines from the origin. As LTC In Figure 4 is a straight line from the origin, average cost In Figure 4 is a horizontal line equal at every point to the marginal cost. Following the same rule, LAC in Figure 5 is decreasing 59 throughout. The short-run average and marginal costs are derived in the same manner. Because of the assumption of the oper ation of the law of non-proportional returns, they all have the same general form. Short-run marginal cost curves fall and then rise. Short-run average cost curves also fall and then rise. Short-run average and marginal costs are equal at minimum average costs. What is more important is the relationship between the short-run and long-run average and marginal cost curves. Each short-run total cost curve touches the long-run total cost curve at one point. At the pertinent output, there fore, short-run and long-run average costs are equal. At every output the long-run average cost is equal to a short- run average cost corresponding to some scale of plant. Fur thermore, the point of equality of the long-run total cost with some short-run total cost is also a point of tangency. At the pertinent output, therefore, short-run and long-run marginal costs are equal. At any output the average and marginal costs corresponding to some scale of plant are equal to the long-run average and marginal costs respec tively . At output OM, for example, STG]_, the total cost func tion for the plant of scale 1, is tangent to the long-run total cost. At OM output SAC]_, the average cost for scale 1, is equal to the long-run average cost. At any other out- 60 put SAC^ is greater than LAO, the long-run average cost. The marginal cost for plant of scale 1, SMC-^, is equal to LMC, the long-run marginal cost at output OM. Generalizing on the basis of previous assumptions, at every output the long-run average cost is equal to the short-run average cost of some particxilar scale of plant. At the same output the long-run marginal cost is equal to the short-run marginal cost of the same particular scale of plant. 1 1 -^Stated symbolically, for a long-run total cost func tion F(x), there is at some output a short-run total cost function, G(x), such that: F(x) - G(x) x x d [F(x)I _ d [g (x )j # d x ~ 55c Similarly for each output such a relation will hold between F(x) and some short-run total cost function. At minimum long-run average cost there is some short-run total cost function, H(x), such that:' F(x) _ H (x) _ d CF(x )] _ dCH(x)) x x dx dx d fF(xjl d|H(x)1 and j x J - L x J - 0. dx dx It is now possible to summarize the three basic types of cost structures using marginal cost patterns as the basis for the classification. The reason for the choice of mar- ginal cost will be made clear in Chapter Pour. The use of marginal cost as the standard helps clear up some confusions on the subject. These confusions result from the failure to state explicitly which type of cost is being considered. Marshall, for example, regularly labelled "cost curves and supply curves alike with the symbols ss, conventionally used for supply curves, and thus diverting the attention of his readers, and perhaps also occasionally his own attention, from the necessity of selecting from among the many possible types of cost curve that one which in the given circumstance alone has claims to being considered as also a supply curve. •^Viner, o£. clt., p. 198. First is the case in which marginal costs eventually begin increasing. This can be called the case of increasing costs to scale. Figure 3 illustrates the increasing costs case. Next there is the situation in which marginal costs are constant through the entire region of possible oper ations. Marginal costs are equal at all outputs to average costs in this case. It can be calledthe case of constant costs to scale. Figure 4 illustrates this situation. Last, it is possible to have marginal costs decreasing through the entire region of potential operations. Decreasing costs to scale is the label for such a situation. It is illus 62 trated in Figure 5. Each of these cost conditions will be examined in Chapter 4 for their influences on the nature of the possible equilibrium conditions. 3. Economies and Diseconomies to Scale In the preceding presentation input costs to the firm were assumed to be fixed, the magnitude of the firm’s operations did not affect inpirb prices. The variations in costs at the various possible outputs must be explained by economies or diseconomies internal to the firm's operations. In other words, any explanation for a particular firm cost pattern must be stated in terms of the effect upon the costs of the various possible levels of operations. Considerable controversy exists in this area of economic theory. Controversy is not present in explaining short-run cost patterns. The cost pattern showing costs at various pos sible outputs with a fixed plant is explained by the law of non-proportional returns. Two general explanations are given for internal econ omies of scale. One explanation emphasizes Indivisibilities of inputs which bring cost decreases.13 other states ■^G-eorge J. Stigler, The Theory of Price, (Hew York: The Macmillan Co., 1947), pp. 128-142 and p. 202. This line of thought stems from a discussion in p. H. Knight* Risk, Uncertainty and Profit, (Hew York: Houghton Mifflin Co., 1921), pp. 177-178. that it is increased efficiency which brings about cost 63 •j 14 decreases, T -^Edward H. Chamberlin, The Theory of Monopolistic Com- petitlon, (Cambridge: Harvard University Press, 1948X7 Appendix B, pp. 230-259. J. M. Clark also considers this problem in some detail. His discussion is in more special terms emphasizing examples of specific cases. Included are examples of both indivisibility and efficiency changes. See J. M. Clark, Studles in the Economics of Overhead Costs, (Chicago: University of Chicago Press 7 1923X# PP. 70-13?". The indlvisibillty explanation states that certain equipment cannot be increased or decreased in size by in finitesimal amounts as scale of operations changes. Rather, such equipment is available for efficient use only at cer tain discrete outputs. A certain machine, for example, Is available only in a limited number of sizes and capacities. Each size can be used efficiently at only one output, or, perhaps, over a limited range of outputs. Cost decreases, in this way of thinking, arise from the availability with scale changes of Inputs which are peculiarly suited to pro ducing at the given output. Three criticisms can be made of the indivisibility ex planation for economies of scale. As divisibility of inputs is assumed in most theoretical presentation, the indivisi bility explanation violates an initial assumption. Further more, indivisibility Is defined in terms of efficiency. "An indivisible productive service is defined as one which is not equally efficient In all sizes (measured in terms of output). An indivisibility explanation, change from 64 ^Stigler, o£. cit., p. 202, n. a on© track to a two track railway, thus becomes a special case of efficiency changes. Also, in theoretical presenta tion the only difference between divisibility andindivisi bility is the difference between a smooth curveand an ir regular curve. The .... objection .... to explaining the shape of a plant curve in terms of ’indivis ibility' is that it has no meaning. If a . factor is indivisible, that is the end of the matter: there is no way of finding out how dividing It would affect its efficiency. If by divisibility is meant merely the sub stitution of a smooth curve for the actual scalloped one, the substituted curve must at least be a reasonable fit to the one it replaces, and not involve an arbitrary as sumption which carries it off on a tangent.1® 1®Chamberlin, o£. clt., p. 244. Explanations of a firm's cost variations by efficiency changes describe cost decreases as resulting from ”(1) in creased specialization made possible in general by the fact that the aggregate of resources Is larger, and (2) quali tatively different and technologically more efficient units or factors, particularly machinery, made possible by a wise selection from among the great range of technical posslbil- 1 V Ities opened up by the greater resources." 17Ibld,, pp. 235-236 65 Increasing specialization has received extensive con sideration in economics. Adam Smith5s famous pin-making example illustrates this beautifully.-'-® There would seem to •'-®Adam Smith, The Wealth of Nations, ed. by Edwin Cannan, (New York: Random House, 1937), pp. 4-5. be no reason for further pursuing this explanation for e conomie s. Efficiency increases caused by greater possible choice among qualitatively different input units pose another prob lem. Under any given technological conditions the same operation may be performed by several different methods each having different efficiency as measured by cost of inputs related to output. A welding operation, for example, may be performed by hand or with the use of an automatic or semi automatic welding machine. Hand-welding becomes much less efficient than machine welding once a certain output is reached. Sheet metal may be formed by hand or by presses and similar metal-forming machinery. Usually machinery becomes less costly than hand operations after a certain output is reached. Many similar illustrations can be found in many standard economic vmrks.’'-® 19A good discussion of such economies can be found in E. A. G. Robinson, The Structure of Gompetltive Industry, (Cambridge: Cambridge University Press, 1950), pp. 25-34. 66 A criticism made of the use of qualitative differences in inputs as an explanation for economies is the violation of the assumption of homogeneity of inputs. This criticism, however, is difficult to appreciate. Qualitative differ ences in inputs could be inferred to mean only different classes or types of inputs each homogeneous in its own class. In the example above, use of welding machines at large outputs and hand welders at small outputs only implies use of different inputs each of which is a homogeneous class.^ 20With a production function P = f(a, b, ...., n), in puts a through d may be used at small outputs and inputs k through n may be used at large outputs to achieve maximum efficiency at these outputs. The inputs are homogeneous within classes, though there are qualitative differences between classes. Qualitative difference would seem to be the basis for the separation of classes of inputs. Increasing costs with output expansion commonly are attributed to physical impediments and by increasing diffi culties in coordination and management. J. M. Clark fur nishes the classical illustration of the plow three times the size of an ordinary two-horse plow which required fifty horses to move because surface resistance increased so rapidly. ^ With any given technology there will be some ^Clark, ojc. cl t., p. 116. output at which these frictions will increase as productive 67 equipment is increased in size and capacity. Coordination and management are commonly held to become subject to increasing frictions as the size of the produc tive unit is increased. Cost would be affected by these in creasing frictions. E. A. G. Robinson states, "A mistake made by a platoon commander demands only an instantaneous 'As yo\i were 1 1 A mistake by an Army Commander may require days of labor to set right. "22 ^ E . A. G. Robinson, ojc. ci t., p. 45. Failure of physical frictions to raise costs can be shown easily. Instead of blowing up the size of an input, _e.£. tripling the size of a two-horse plow, it Is possible to add successive units of the most efficient sized unit. Barring other Inefficiencies, this would result in costs leveling off Instead of rising. Kaldor criticizes the coordination and management ex planation as not explaining diseconomiesAccording to Kaldor, "The Equilibrium of the Firm," Economic Journal, Vol. XLIV (1934), pp. 60-76. Kaldor, coordination as an adjusting, changing, contracting activity "is an essentially dynamic function. "24 As the 24Ibid., p. 70. 68 assumptions of the theory of the firm developed to this point are static, the dynamic coordinating activity cannot be held responsible for diseconomies. By Kaldor’s line of thouight, it is not possible to get decreasing returns to management for all inputs are assumed to be perfectly elas tic In supply Including management. If management is as sumed to consist of the supervision of routine operations In a static situation, "An army of supervisors may be just as efficient(provided It consists of men of equal ability) as one supervisor along.1'2® 25Ibid., p. 68. Although Kaldor’s view about the dynamic character of the coordinating activity seems well taken, criticism can be made as to the equal efficiency of one or an army of super visors. If supervisors are subject to some degree of human error, then Increasing the levels of supervision would mul tiply the human error. As more people are Involved in the communication of accounts, routine orders, material flow and other such routine supervisory Information, error would be compounded. A normal error of 2% per person would be com- pounded to an error of 18.3% with ten people involved. 2®Total error would equal l-(l-e)n where _e is the indi vidual error and n is the number of people involved. 69 Homogeneity would act here to ensure the compounding of the error. If the supervisory input were homogeneous, the error would be biased in one direction thus eliminating the possi bility of offsetting errors. The common classroom experi ment of the person to person (all identically biased) repe tition through a classroom of a previously prepared state ment can be used as an explanatory analogy. When the final person hears the statement and writes it down a considerable difference from the original is to be expected. Another view concerning economies and diseconomies to scale exists. Samuelson objects to the intuitive or self- evident truth of the ubiquitous constant costs or, what is the same, linear homogeneous production functions Implied by Kaldor's line of thought. Samuelson states, It is a scientifically meaningless assertion that doubling all factors must double product. This is not so because we do not have the power to perform such an experiment; such an objection is of course irrelevant. Rather the statement is meaningless because it could never be refuted, in the sense that no hypothetically conceivable experiment could ever controvert the principal enunciated. This is so because if product did not double, one could always conclude that some factor was 'scarce'.^7 ^Samuelson, ojc. ci t., p. 84. Cf. Joseph A. Schum peter, History of Economic Analysls, (New York: Oxford Uni versity Press, 1954), pp. 1040-1041, nn. 33 and 35. It should be noted that , following the same line of thought, the compounding of individual error to equal l-(l-e)n pan- no t be refuted nor confirmed by any conceivable experiment. Following Samuelson's suggestion, the remainder of this 70 study will assume production .functions drawn up as reflect ing the particular circumstances pertaining to the firm. These circumstances will give some firms decreasing costs to scale, some constant costs to scale and others increasing costs to scale. If any such patterns cannot exist, analysis or conclusions drawn respecting firms possessing such pat terns have no meaning In empirical applications. Tools which at least have value In explanation, however, will he at hand. B. Industry Supply Patterns Cost patterns related to scale changes are pertinent to the consideration of individual firm equilibrium, adjustment to given demand, technical and input price conditions. Con sideration must also he given to Industry supply patterns. The order of the discussion will be to consider possible industry supply patterns first, then to proceed to the reasoning or bases for the possible patterns. 1. Possible Industry Supply Patterns Industry supply curves consist of more than the sum of firm supply curves, i_.e_. marginal cost curves. The industry supply curve must take Into account not only the firm sup ply curves at any given industry output but also the nature of industry supply with the entrance or exit of individual firms. Analysis of Industry supply conditions must consider the effect of industry expansion or contraction upon input 71 prices to all firms. Such analysis also must explain the limitations to the number of firms in the industry and the profitability of the firms in the industry. The effects upon inputs is of particular importance for it is through such input price effects that resources are allocated be tween industries. The term supply curve, it should be noted, implies the existence of pure or perfect competition. A characteristic of monopoly is the absence of any true supply curve in terms of a particular output for a particular price. Rather, the monopolist looks to the demand curve facing him to determine the price to be charged at any output. The discussion of industry supply patterns, therefore, automatically assumes the absence of price-setting power on the part of the firm, .1.e_. assumes pure or perfect competition. No firm, further more, can have any effect on input prices. Any change in output or input prices must come from outside the individual firm. Input prices, the basis for varying industry pat terns, can be changed only by entrance into or exit from the industry by firms. As discussion of industry supply re quires the existence of many small firms, there must be an assumption of some limitation to the size of the firms. This limitation will be assumed to arise because of rising marginal and average costs coming from diseconomies of com pounded Individual error overcoming any economies which might prevail. Such an assumption does not represent any 72 acknowledgment of the probability of such firm cost patterns. If such patterns do not exist, it means merely that the equilibrium results with such firm and industry patterns to be considered In the next chapter cannot be found. Rising firm average and marginal costs can still be of use for analytical purposes. In the chapter to follow, equilibrium of firms and in dustries will be considered. The method to be followed will be to see If the assiimption of competitive conditions under the various firm and Industry supply conditions involves any contradiction. Industry supply patterns derive from the different in put prices which would exist at various industry outputs and, thus, also different numbers of firms In the industry.^® At ^®The discussion here follows from Joan Robinson, "Rising Supply Price," Economica, Vol. XIII IT.S. (1941), pp. 1-8, reprinted in Readings in Prlce Theory, ed. by Kenneth E. Boulding and George J. Stigler^ (Chicago: Richard D, Irwin, Inc., 1952), pp. 233-241. any one particular long-run industry output there would be a certain number of firms In the industry. Each of these firms would be faced with given Input prices. The given In put prices and the firm’s production function would deter mine each firm’s cost pattern. Here, the production func tion is specified to lead to rising average and marginal costs. At any other industry output there would be a dif ferent number of firms in the industry. Because of the 73 effect upon the demand for Inputs there also might be dif ferent input prices. The industry supply patterns would arise out of the absence or presence of different input prices at different industry outputs. The reason for the possible input price variations will be discussed in the next section. Figure 6 illustrates the possible industry supply pat terns. Rising supply price is illustrated in Figure 6a. At the total industry output of OM a particular firm produces Om. The particular firm lllustreated has a minimum long-run average cost of Oa. Every other firm in the industry also is operating at the minimum long-run average cost of Oa though not necessarily at Om output. Firms having a higher minimum average cost could not operate. Any firm having a lower minimum average cost would be underpaying some input. Under competitive conditions the price of this input would be bid up.^® Every firm in operation, therefore, would have 29j0an Robinson considers this condition as entrepren eur’s rent. The conclusion is the same. See Joan Robinson, The Economics of Imperfect Competition, (London: Macmillan & Co., 1933), pp. 123-126. identical minimum long-run average cost in a competitive equilibrium situation. SS is the sum of the marginal costs of all firms operating with their optimum plants and is, therefore, the short-run industry supply. If total industry output were raised to OM^, costs 74 Figure 6 Industry Cost Patterns nc, AC, AC LRS Qu tpu T nc AC ci LRS Output MC AC MC, AC Q. LRS 0 75 would Increase for every firm, under the increasing industry cost conditions illustrated in Figure Sa. Each firm would, under the illustrated conditions, operate with a minimum long-run average cost of Oh. This would result from the increases in input prices as more firms entered the in dustry. The greater number of firms in the industry re sults in the short-run supply, S'S*. LRS, the industry supply curve, shows the pattern of the equilibrium long-run average and marginal costs to each firm in the industry as total industry output changes. Such a long-run industry cost pattern is commonly referred to as illustrating an in creasing cost industry. Figure 5b shows constant industry costs. In such an industry, input costs remain the same at any total industry output, i,.^. a firm is unaffected by the entrance or exit of other firms in the industry. The firm operates with the same long-run cost structure. The sum of optimum plant mar ginal costs is moved to the right as more firms enter, e_.g,. the short-run industry supply moves from SS to S'S’. In Figure 6c entrance of new firms into the industry lowers input costs. At industry output OM the illustrated firm will have minimum average cost, equal to marginal cost, of Oa at output Om. The sum of the firms' marginal costs is illustrated by SS. If the industry output expands to output 0M]_, the illustrated firm will operate at minimum average cost of Ob at output Om. All firms in the industry 76 would have the same minimum average cost. The sum of the firms1 marginal costs, or short-run industry supply, is il lustrated by S'S1. Fig^^re 6c illustrates decreasing indus try costs or decreasing supply price. The nature of equilibrium under the different industry cost conditions will be investigated in Chapter 4. The pos sibility of an optimum equilibrium condition under each of the industry cost conditions will be emphasized. 2. Causes of Industry Supply Patterns®^ ^T his analysis follows Joan Robinson, "Rising Supply Price," ojo. cl t. , pp. 233-241. Changes in input prices were held to be responsible for the different industry output patterns. It is now necessary to investigate the causes of variation in input prices with changes in industry output. While static assumptions are used, changing consumer tastes, which is a dynamic charac teristic, will be used for purposes of exposition. Industry cost changes occur when Inputs shift In res ponse to consumer demand changes, money income in the econ omy remaining constant. As a result of the shifts of in puts, all firms in the economy may be subject to input cost changes which are beyond their control. These changes can be called external economies or diseconomies as contrasted to cost changes dependent upon the scale of plant which an 77 Individual firm chooses. External economies or diseconomies result from changes in demands for inputs. To determine the final effect upon costs to the industries Involved both the demand change and the nature of the supply curves of the inputs must be con sidered. The resulting cost changes, In other words, depend upon the nature of the input demand changes and also upon the elasticities of supply of the affected inputs. The Im pact upon costs of the industries Involved depends on the proportions in which the industries combine the inprits which have had their prices changed. Assuming a competitive economy with a given total in come, a change in consumer tastes would increase the demand for one commodity and decrease the demand for a series of other commodities. The expanding industry will have an increase in demand for Inputs and will, thus, absorb inputs in some proportions. The contracting industries will have a decrease in demand for inputs and will release inputs in some proportions. The proportions In which inputs are re leased can be called the average proportions. Differences in the proportions In which inputs are absorbed from the average proportions and the supply elasticities of the af fected inputs determine industry cost changes and, there fore, industry supply patterns. If the absorbing industry combines inputs in the same proportions as the average proportions, the absorbing in 78 industry must have constant costs. This because the total demand for every input will remain the same. Input prices will be unaffected by the transfer between industries. Absorption in the same proportions as inputs are released seems, however, quite improbable. It Is, furthermore, when differences prevail that the more interesting results are achieved. This leads to a consideration of possible elas ticities of supply. If all input supply curves are either positively sloped or perfectly inelastic, increasing industry supply price must prevail. The inputs used by the expanding Industry in smaller proportions than the average would decline in price while those inputs used in larger proportions than the aver age would rise in price. Costs to the expanding industry would be increasing, therefore, because the Inputs which the expanding industry requires most would rise in price, other inputs, used less, would fall in price. For a decreasing Industry cost to prevail some Inputs must be supplied at constant supply price. To illustrate, 31 an extreme case given by Joan Robinson can be employed.. 31Ibid., pp. 239-240. This example assumes that inputs are supplied only with either a constant supply price or with completely inelastic supply. If the absorbing Industry purchased inputs having constant supply price In greater proportions than the aver 79 age and Inputs with inelastic supply in smaller proportions than average, the absorbing industry would have decreasing costs. This would result because some input prices would fall, some would remain constant, but none would rise. The expanding industry could purchase its composite bundle of inputs at a lower total cost. The same result could be achieved if some inputs had constant supply price, while other inputs had an increasing (positively sloped) supply price. It should be added that an industry purchasing only inputs having constant supply price would have constant industry costs. If inputs were supplied with decreasing supply prices, decreasing industry costs could easily arise. The possi bility of decreasing input supply price has been debated at some length, especially with regard to labor.^2 Reconsider- rz p Clarence D. Long, "The Labor Force and Economic Change,” Insights Into Labor Issues, ed. by Richard A. Lester and Joseph Shister, (New York: The Macmillan Co., 1948), pp. 329-355. ation of the debate would not be pertinent to this study. It might be pointed out, however, that an industry which was composed of one firm having decreasing costs to scal9 would supply a product at decreasing supply price.^ This firm’s ^This would be a one firm industry because such cost situations lead to a monopoly, as will be shown in Ch. 4. 80 product could be an input to other industries which further processed the product. Despite any difficulties which might prevail it will be assumed that decreasing industry costs can exist. ^he effect of such decreasing industry costs upon equilibrium will be considered in the following chapter. A difficulty arises if there is a possibility of dif ferent proportions of released inputs depending upon which specific industries ara affected by decreased demand. It is possible that the change in tastes which results in shifts In input allocation may take many different forms. A priori It is just as possible for Industries B and G to be the largest input releasing industries as it Is for industries D and E. Each of these, and all other, possibilities may have a different effect on Industry costs. An industry cost pattern must assume, therefore, that there is some given average proportioning of inputs for all other Industries and that the releasing industries release their inputs in these average proportions. The difference of the input propor tions in the industry being considered from the average proportions determine the industry cost pattern. One more point should be mentioned. All movement of Inputs between industries was assumed to have zero transfer costs. The inclusion of transfer costs would complicate the analysi s. C• Summary Firm long-run cost patterns derive from the production 83. function. Given input prices, the firm adjusts the inputs optimally at any given output. The optimum adjustment re sults in equal ratios of price to marginal product for all the inputs. At each possible output there is a unique op timum adjustment which gives the total expenditure at that output. There are three possible long-run firm total cost patterns. Average and marginal cost patterns can be derived from these three total cost patterns. Using marginal cost as the standard, the three cost patterns can have increas ing,- decreasing or constant marginal costs. Which happens to prevail depends on possible economies or diseconomies to scale resulting from efficiency changes. Arguments that production functions must be linear and homogeneous, i_.e_. doubling all inputs must double output, are considered as untestable propositions. Economies and diseconomies to scale will be assumed possible, thus the three firm cost patterns. Industry cost variations are based on the exit or entrance of new firms changing total Industry output. An assumption Is made that variation In the output of an indi vidual firm is Insignificant for the entire industry. The entrance or exit of firms Is In response to changing con sumer demand because of changes In taste. Industry expan sion or contraction because of entrance or exits of firms results in shifts of inputs. An industry cost pattern depends upon: (1) the proportions In which the industry com 82 bines the several inputs as compared with these proportions for all industries on the average; and (2) the supply func- tions of these inputs. Attention should be called to the static nature of the analysis in this chapter. This is especially true of the discussion of the relationship of the firm and industry. Firms were assumed to possess similar cost conditions; so similar, in fact, that minimum average costs were the same for all firms. A primary characteristic of economic devel opment, however, is the tendency of firms to break out of the confines of such industry equilibrium which, as it is tied to the equilibrium of all industries, means the distur bance of the general equilibrium of the economy. According to Schumpeter, .... this is .... obviously due to the unremitting efforts of people to improve according to their lights upon their productive and commercial methods, :l.e_. to the changes in technique of production, the conquest of new markets, the insertion of new com modities, and so on. This historic and irreversi ble change in the way of doing things we call ’in novation’ and we define: innovations are changes in production functions which cannot be decomposed into infinitesimal steps.34 A. Schumpeter, "The Analysis of Economic Change," Review of Economic Statistics, Vol. XVII (May, 1935), p. 4, reprinted In Readings in Business Cycle Theory, ed, by Gottfried Haberler, (London: George Allen and Unwin Ltd., 1950), p. 7. It is, however, because of the irreversibility and the impossibility of decomposition into infinitesimal steps that 83 the analysis of such change is not amenable to rigorous theoretical consideration. The discussion here, therefore, remains primarily static, though the importance of such dy namic elements must not be disregarded. CHAPTER IV equilibrium c o n d i t i o n s a n d m a x i m u m w e l f a r e The various cost situations have pertinence only in influencing equilibrium conditions. In this study the focus of Interest is upon the impact of the cost conditions upon possible optimal long-run equilibrium. As two sets of cost structures, of the firm and of the industry, have been de rived, each will be considered in turn. In the discussion of equilibrium perfect freedom of entry will be assumed. Such freedom of entry can be defined as the ability of a firm to enter under the same conditions as are existing for any other firm in the i n d u s t r y . T h e r e are no impediments ^-Robert Triffin, Mono polls tl c Compe tl tion and General Equilibrium Theory, (Cambridge: Harvard University Press, 1940")"," pp. 117-125. such as patent laws, product differentiation or artificial limitations in the supply of inputs. A negatively sloped market demand curve will also be assumed. As a result, no unusual demand conditions could limit the possibility of an optimal equilibrium. Any impediments to the attainment of an optimum come only from the existing cost conditions. As a competitive equilibrium situation leads to an op timum, the approach to be used In this chapter will be to see first of all if the cost conditions are compatible with a competitive equilibrium. If the competitive equilibrium 84 85 is possible, the optimum solution exists. If a competitive equilibrium is not possible, comparison must bo made between the resulting equilibrium and the optimum. As price equal to marginal cost exists at an optimum, the comparison can be based on the divergence of the price from marginal cost. Explicit definition of a stable competitive equilibrium is necessary, if it Is to be a focal point of the investiga tion, The Marshallian definition of stability in equilib rium will be used. Such stability exists when the price or output, if displaced from the equilibrium point will return to the original equilibrium point.^ Competitive equilibrium ^Alfred Marshall, Principles of Economics, Eighth edition, (New York: The Macmillan C o 1949), pp. 545-346. is defined as existing if two conditions prevail. First, a price must be given to the firm. This is equivalent to the conventional perfectly elastic demand curve over which the firm has no control. Second, long-run competitive equilib rium exists only if there Is an absence of pure profits in the long-run. This means that the given price Is equal to average cost at the equilibrium output. It also means the exhaustion of product, i^.e,. each resource paid according to its marginal productivity exhausts total output. Marginal cost need not be considered explicitly as price is auto matically set equal to marginal cost under perfect compe tition. 86 A. Equilibrium of the Firm For a determinant equilibrium of the firm in general there must be discovered some output at which two conditions exist. First, at this output marginal cost must be equal to marginal revenue. Second, the marginal cost curve must have a greater slope than the marginal revenue curve.® This is ®With R(x) as the total revenue function, C(x) the total cost function and P(x) as the profit function, P(x) = R(x) - G(x). For a profit maximum, P*(x) r R*(x) - C(x) = 0, or R*(x) r C'(x), (1) and Rn(x) - C”(x) < 0. (2) As the price is given to the competitive firm, R T(x) = constant and R"U) = 0. (3) Substituting (3) in (2), -C"(x) < 0 or C"(x) > 0. For a monopolistic firm, R"(x) < 0. (4) Substituting (4) in (2) It can be seen that C"(x) can be positive. It can also be negative but must be less nega tive than R n(x). the equivalent of saying that the marginal cost curve inter sects the marginal revenue curve from below. Only when both these conditions prevail Is profit maximized. The two general conditions must be met whether the mar ket In which the firm operates is competitive or monopolis tic. Despite text book slighting of the second point, It is this second condition, which is most important in application to the firm. If the firm operates in a competitive market, its demand curve is perfectly elastic and is illustrated by a horizontal line, i.e. the slope is zero. To meet both re quirements the marginal cost curve must be rising at the equilibrium output. A monopolistic firm is presented with a negatively sloped demand curve. For equilibrium in such monopoly conditions, the marginal cost curve must lie below the marginal revenue curve to the left of the equilibrium output and above the marginal revenue curve to the right of the equilibrium output. If this condition prevails, it does not matter whether the marginal cost curve is sloped posi tively or negatively. Each possible firm cost pattern must now be considered from the standpoint of the equilibrium conditions. The three possible firm cost patterns will be considered in turn. 1. Increasing Costs to Scale Figure 7 Illustrates the conventional case of the firm having increasing costs to scale. Individual plant cost curves are ommitted for purposes of simplicity. Reference to Figure 7 will Indicate why the competitive equilibrium requirements hold true and why there Is no problem of com petitive equilibrium with increasing costs to scale. Given a price of OP, OM is the equilibrium output of the firm. At this output the last unit produced contributes just as much to the firm’s revenue as It adds to the firm's cost, or mar ginal cost is equal to marginal revenue. At any other out- 88 Figure 7 Increasing Costs to Scale AC AR MR Ou t p u t 89 put profit would not be maximized. If the firm produced more than OM, cutting output to OM would decrease costs by a greater amount than revenues would be decreased. If some output less than OM were produced there would be an advantage in increasing output. Thus, the equilibrium would be stable because of the adjustment back to OM, if the equilibrium were temporarily displaced. Figure 7 has average cost equal to average revenue. Thus, competitive equilibrium as defined above is possible in this cost case. Input payment according to marginal pro- diictivity at such equilibrium has been proved by Hicks.^ *J. R. Hicks, The Theory of Wages, (London: Macmillan & Co., 1932), pp. 237^538. x = f(a, b, c, ...... ) (Production function) Total Cost r apa / bp-j-, / ..... pa, p-^, etc. are input pri ces Average Cost = z r l(apa / bpfe / ....) (1) x z z px , i.»e> average cost equals selling price. For z to be a minimum, 3z 9 z m must all r 0. 'WZ* ab fa z apa t bpb ^ .....3 ' !pa - |s §f = ! p a - h I f _ 1/ _ d X , - x (pa " z as:5* But since 9z _ n _ 5x _ 3x, and similarly for aa * a “ da 3a other inputs b, c, ...... Substituting in (1) x = a b ^ ...... 90 The means by which the no profit equilibrium arises will-be explained when industry equilibrium is considered later in this chapter. Suffice it for the moment to show that a competitive equilibrium is possible in the increasing cost case. The cost conditions do not bar a competitive solution which results in the optimum. Reference must be made to increasing costs through all scales. Such a cost situation would result in industries made up of one man firms.® The result would be equilibrium 5Paul H. Douglas, The Theory of Wages, (New York: The Macmillan Co., 1934), pp. 55-56. at minimum average cost as in Figure 7 (though falling aver age cost at outputs smaller than equilibrium would be ab sent) • The relationship between long-run marginal cost and marginal revenue at the first intersection point (output OK in Figure 7) represents a situation neglected, in convention al equilibrium analysis. At output OK marginal cost is equal to marginal revenue. However, it is the profit minim ization point rather than maximization point.® Marginal ®To minimize P(x) = R(x) - C(x), P'(x) r R 1 (x) - C'(x) r 0 and P"(x) r R"(x) - C"(x) > 0. cost at this point is negatively sloped. The firm could 91 better Itself, from a profit standpoint, by producing any where other than at output OK. By decreasing output units having an additional cost greater than additional revenue would not be produced. If output were increased, each ad ditional unit would return a greater additional revenue than additional cost Incurred until output OM Is reached. At OM the last unit produced would return an additional revenue just equal to the additional cost incurred. It is of in terest to note that a theoretical structure based upon minimizing assumptions (which could hardly be termed eco nomic) would take output OK as the equilibrium point. 2. Decreasing Costs to Scale Figure 8 illustrates the conditions faced by a firm having decreasing costs to scale. Assuming some price OP as given to the firm there could be no stable equilibrium. The primary condition of equality of marginal cost and marginal revenue can be attained at output OM. OM, however, is the profit minimization point. The firm could improve its posi tion by either decreasing or increasing its output. If out put were decreased, total cost would be decreased more than total revenue would be decreased. This is true because at outputs smaller than OM marginal cost exceeds marginal reve nue, By increasing output total cost would increase less than total revenue would be increased. At any output greater than OM marginal cost is less than marginal revenue. Assuming a profit maximization motive behind the firm’s 92 Figure 8 Decreasing Costs to Scale AR MR AC ----MC AR( MR, Ou t p u t 93 operations, a firm would expand output until it captured a substantial part of the market for the product. Upon the capture of a substantial part of the market the demand curve to the firm would no longer be perfectly elastic. The firm's operation would have some impact upon the price of the product. Monopolistic powers would exist. An industry with firms having such cost characteristics would result in either a single monopoly or some type of oligopoly solution equivalent to monopoly. At the birth of such an industry all firms would see very great advantages in expanding in definitely. Potential profits would appear limitless. Eventually some firms would become significant producers in terms of impact on the market for the product. With the demand curve sloped for the various firms, profits would not seem limitless. Not only would the firms find that their own activities influence the market situations, but they would find that competitive activities by any firm would affect demand conditions faced by all other firms. This is, of course, the standard definition of oligopoly. The result would be either price warfare with the advantage going to the firms having largest scale of plant or some type of im plicit or explicit oligopoly agreement. Price warfare would lead, in turn, to either a one firm monopoly or to an olig opoly agreement. Natural monopoly conditions can be said to exist when decreasing marginal cost prevails at individual firm outputs 94 of a size, relative to entire industry output, that no one firm can influence price. Under such conditions each firm sees its advantage in expanding its scale of operations. Such expansion eventually results in a stabilized oligopoly or a pure monopoly. In any case, pricing above marginal cost will prevail. A deviation from optimum welfare condi tions will exist. AR-j_ and MR^ in Figure 8 illustrate a possible demand condition for a pure monopoly. AR^ is the total demand for the product at various prices. Monopoly output would be set at ON with a price of OR. MR-^ cuts MC from above, meeting the secondary requirements for stable monopoly equilibrium. If price deviated from OR, there would be a quick movement back to the OR price. If price increased, purchases and output would be less than ON. The firm could profit by pro ducing more and lowering the price as each additional unit produced would add more to revenue than it would add to costs. If price went below OR, the firm could gain by raising the price and producing less. Each unit of output greater than ON contributes less to revenue than it adds to costs. It is of Interest to note the similarity of the average and marginal cost curves in the decreasing cost situation and the cost curves illustrated in Figure 7 at outputs smaller than OL. The portions of the conventional U shaped cost curves to the left of minimum marginal costs are exact- 95 ly like the decreasing cost case. It could be assumed that eventually the curves in the decreasing cost case would be gin rising and thus take the form of the U shaped cost curves. If the forces resulting in Internal diseconomies are universal this would be true. Care should be taken, therefore, in the definition of the decreasing cost situ ation. It should be defined in terms of the conceivable scale of operations given a certain market. There is no assurance that eventually operations might not be at a scale at which diseconomies would come into play. If this line of thought has any validity, decreasing cost would be a special case within the conventional cost picture presented by the II shaped curves. It would exist because under the given technology economies are present for a great potential range of operations. Consumer demand under these conditions must not require outputs so large that diseconomies would become dominant. The possibility of eventual dominance of Internal dis economies requires that consideration be given a particular situation. This situation would occur when decreasing mar ginal cost exists at individual firm outputs so small that each firm would see the price as given but upon expansion one or a few firms which still existed would encounter dis economies and rising marginal costs. In such a situation pure or perfect competition could not exist. But the pure monopoly situation with the monopoly equilibrium at an out- 96 put at which, decreasing marginal cost prevailing, as illus trated in Figure 8 , would not exist either. Rather, one or a few firms would find an equilibrium situation at plant outputs having increasing marginal costs. Either a pure monopoly would exist with a long-run output at a point at which marginal revenue equaled the increasing marginal cost or an oligopoly situation would exist. Under the oligopoly situation the market would be shared on some basis by the firms in existence, each of these firms having increasing marginal costs. Natural monopoly may have four different results, therefore, each arising from the existence of decreasing marginal costs at the low output levels necessary for com petition. A pure monopoly situation with decreasing mar ginal cost at the equilibrium output, as Illustrated in Figure 8 , may exist. A pure monopoly with a U shaped aver age cost curve and increasing marginal cost may exist if one firm could handle the demand while a greater number of firms could not exist because the market was not large enough. An oligopoly situation with several firms dividing up the ex isting market on some basis is another possible result. Such an oligopoly situation may have either increasing or decreasing marginal cost.^ ^Such oligopoly situations would most likely be caused by a fear on the part of all firms to fight a price war be cause of lack of assurance of winning. 97 3. Constant Costs to Scale A firm with constant costs to scale brings problems which are quite different from either increasing or de creasing cost firms. This can be seen by the application of the stated equilibrium requirements to this cost condition. Under competitive conditions the price is given to the firm. A perfectly elastic demand curve given to a constant cost firm can have three possible positions relative to the aver age cost curve (equal at all outputs to the marginal cost curve).® The perfectly elastic average revenue curve (equal ®This treatment follows Samuelson, ojo. clt., pp. 78-80. at all outputs to the marginal revenue curve) may parallel the cost curve above it or below it. The revenue curve may also coincide with the cost curve. If the revenue curve lies above the cost curve, exist ing firms will expand and new firms will enter. With free dom of entrance assumptions, there should be a final adjust ment such that the coincidence solution results. This coin cidence would come about through changes in either the pro duct price or input prices and product price, deiDending on the nature of the external forces resulting from industry output changes. These industry cost changes will be con sidered below. Similar results would be attained if the revenue curve paralleled the cost curve below. Because of losses' the adjustment would come about because of contrac 98 tion of scale and the exit of some firms. The result would be, in any case, the coincidence of four curves as illustrated in Figure 9. Obviously, marginal cost is equal to marginal revenue at all outputs. The sec ondary conditions for a determinant equilibrium are not met. The marginal cost curve does not cut the marginal revenue curve from below. This means that there is continual incen tive on the part of the firm to expand output. At first sight a monopoly result might be expected as wild expansion and cut-throat operations occurred in the industry. A one firm inonopoly might well be the final result of the constant cost case. Assuming free entrance, the impact upon the consumer, however, would be that of a competitive solution, i..e^. price equal to marginal cost. Price could not rise above the optimal price of OP. Even if one firm succeeded in taking over the industry, it could not restrict output and raise price above OP over the long-run. To do so would be to invite entrance of other firms bringing the price back to OP. At this price there would be no monopoly profits and all inputs would be paid in accordance with their marginal contribution.® Something akin to monopoly ^Euler's theorem holds in this case. With production function P r f(a, b, ...... n), mP = f(ma, mb, ...... mn), because of linear homogeneity. Differentiating with respect to m, P z afa / bffc, / ...... / n f n . 99 Figure 9 Constant Costs to Scale SAC L ac LMC MR 100 could possibly prevail. If one firm or a few firms took over the industry, firm expansion would be limited by an ex ternal factor, market demand. Traditional competitive anal ysis assumes the firm size to be limited by internal dis- e conoraies. Despite the Indeterminateness of any one fiiun’s output in this case, the solution for the entire industry would be optimal by welfare standards. It even might be stated that the optimal solution would be maintained more easily in the constant cost case than in the conventional U shaped average cost situation, A potential competitor free to enter under the same cost conditions could enter with an extremely small scale of plant. If a firm operating with a large scale of plant, SACg in Figure 9, tried to apply monopoly power, It would be very easy for small firms to enter with a small scale of plant, e_.g. SAO-^ in Figure 9. Contrasted with the situation in the U shaped cost case, the optimum result would be maintained more easily. In the U shaped cost case, the minimum long-run average cost could well be at a very large capacity. Any smaller scale of plant would be at a cost disadvantage. With constant cost to scale, an extreme ly small plant could have all the cost advantages of a very large plant.^ • ^ C f . Abba P. Lerner, The Bconomlcs of Control, (New York: The Macmillan Co., 1946), p. 83. Lerner implicitly has assumed in his argument some barrier to entry. It Is the barrier to entry which makes monopoly profit possible. 101 It is of interest to note that the constant cost analy sis Is very applicable to arbitrage through time or space. If a discrepancy exists between prices plus transportation costs for two locations or between prices plus storage costs for two time periods, It Is customary to theorize the elim ination of the discrepancy. Because of the existence of constant costs to scale, however, it is impossible to theo retically determine the amount any individual shall engage in such arbitrage. It can only be stated that arbitrage will take place until the discrepancy is eliminated. Whether it is one arbitrager (monopolist) or many (competi tors), arbitrage always is expected to wipe out the profits arising from arbitrage, If free entrance exists. The arbi trage industry is expected to have a very stable result, in other words, though there Is no way to determine the equi librium point of any one arbitrager. B. Equilibrium of the Industry Focus of attention was placed on the possibility of competitive equilibrium of the firm in the previous section. It is now necessary to determine whether the nature of any industry cost pattern bars competitive equilibrium. Earlier in the study three industry cost conditions were derived. Three possible patterns of firm scale adjust ments were deduced also. Wine possible combinations of cost patterns exist, therefore. Each of the three Industry cost patterns can be combined with each of the three firm cost 102 patterns. As distinctions between firm and industry become impossible with decreasing cost to scale existing, three of such combinations are eliminated. As the focus of interest at this point is whether the industry cost pattern, in it self, bars competitive equilibrium, increasing costs to scale for all firms in the industry will be considered. With the firm cost patterns offering no impediment to opti mum equilibrium, it should then become evident whether any thing in the nature of external economies or diseconomies prevents a competitive equilibrium. Constant costs to scale with different industry cost patterns also will be consider ed to show the nature of industry equilibrium with firms having constant costs to scale. Conventional static assumptions of given technology re flected in production functions, given tastes and given in put supply functions will be used. In each cost case, how ever, the givens are defined to give the required cost pat terns. Change in tastes, which is really a dynamic element, will be used as an analytical device to show how static in dustry equilibrium is reached. 1. Increasing Industry Costs In analyzing increasing industry cost, Figure 10 can be used in illustrating the problem Involved. Assuming that demand curve DD Intersects the rising industry supply curve at total output 0M, the price which would prevail in the industry would be 0a. Every firm in the industry would see 103 Figure 10 Increasing Industry Costs SRS' LRS M C AC MC AC SRS' LRS SRS O u t p u t 104 that price- as given, i.*e.» a perfectly elastic demand curve at price 0 a, Each firm also would have a long-run average cost curve such that there would be a point of tangency with the horizontal demand curve at some output which would be the firm's long-run equilibrium output. The sum of all the firms’ equilibrium outputs would equal 0M. If the sum di verged, price would change from 0 a bringing entrances or exits until total output 0M at price 0a existed. The short- run Industry supply would be SRS which is the sum of the marginal costs for optimum sized plant of all the firms In the Industry. The firm illustrated, having a long-run average cost curve AC-j_, would have a long-run equilibrium output of 0 m. Other firms might have a greater or smaller long-run equi librium output depending on conditions peculiar to each firm. But they all would reach equilibrium^ARrMRrMCrAC) at price 0a. As mentioned earlier, any firm having minimum average costs greater than 0a would be driven out of the Industry because of the continual losses. Any firm having lower min imum average costs would be underpaying some input, probably the entrepreneur whose ability makes such low costs possi ble. Competitive bidding up of the services of the inptxt responsible for the low costs would raise the AG curve until minimum average cost equalled 0a. This is, of course, a case of opportunity costs. Assuming an Increase in the demand for the product of 105 the Industry resulting in a new industry demand curve, dd, cutting LRS at output ON, a new equilibrium price of Ob would exist. The demand increase would come about, by the previous discussion, because of a change in tastes such that consumers want more of the product of the Industry consid ered and less of the products of other industries. Price would rise above Oa to the price prevailing where SRS, the short-run supply curve Intersects the new demand curve. The increased price would result in profits to all firms in the industry and in the entrance of new firms into the industry. Just the opposite would be occurring in the industries sub ject to decreased demand. Industries having decreased demand would have a lower demand for Inputs and would be releasing them. The expan ding industry would have Increased demand for inputs and would be purchasing more of them. The absorbing industry would be hiring more inputs having relatively Inelastic sup ply than the contracting Industries were releasing and hiring less having elastic supply than the contracting In dustries were releasing. Costs in the expanding industry would rise, eliminating profits. Industry equilibrium would finally be attained when all firms would have minimum aver age cost of Ob and total Industry output would be ON at price Ob. Increasing industry costs with firms with constant costs to scale also can be illustrated with Figure 10, 106 neglecting MG, MCq, AG and ACq. With industry output of OM, the sum of the long-run costs for firms operating in the industry (whether one or many) would be ae_. The short-run industry supply which would be the sum of marginal costs of existing plants would be SRS. If industry output increased to ON, input prices would increase to raise the cost struc ture up to horizontal line bf, which would be the new sum of long-run costs. SRS', the sum of marginal Costs of ex isting plants, would be the short-run industry supply, LRS would be the points of equilibrium for the industry at dif ferent outputs. The equilibrium of every firm would be in determinate but industry equilibrium would be well defined, Of! output at price Oa and ON output at price Ob. To illustrate the method of achieving equilibrium, an increase In demand because of changing tastes can be used to facilitate explanation. Assuming industry supply LRS and demand DD, price would be Oa, industry output would be OIV!. The firms operating, whether one or many, would have plant inarginal cost curves totaling SRS. With the given produc tion functions and input prices, the long-run cost struc tures of all the firms (one or many) would total an. If de mand increased to dd, price would rise to the intersection of SRS and dd. Potential profits would result in expansion of existing firms and entrance of new firms,. As inputs shifted from the contracting industry to the expanding in dustry, the expanding industry would experience rising cost 107 structures because of the rise in prices of important in puts. When the industry output reached ON a new industry equilibrium would be reached with a price of Ob. Any fur ther entrance or expansion would raise costs and lower the price, resulting in readjustments back to the equilibrium point. If industry output were less than ON, costs would be less than Ob, prices would be greater than Ob. The re sulting profit would give incentive for expansion and en trance of new firms. Such industry expansion would continue until industry output ON with a price Ob would be reached. 2. Constant Industry Costs In an industry having constant industry costs, equilib rium price remains constant to the individual firm regard less of industry output changes. In Figure 11, the demand curve for the product cuts the industry supply curve LRS at OM, the resulting price to the firm being Oa. Assuming dis economies to scale, each firm woTald have a long-run cost pattern similar to that illustrated by AC and MC, Om being the firm's equilibrium output. If demand Increased so that the demand curve, dd, cut LRS at ON, the equilibrium price would remain Oa. The ad- — justment process which results in industry output totaling ON is of interest. As a result of an increase in the demand for industry output, the immediate effect would be an increase In price. The Increase would be to the price at the intersection of 108 Figure 11 Constant Industry Costs SRS nr AC a LRS SRS' O u t p u t o 109 SRS and dd. As a result of the price rise, firms in the in dustry would enjoy- pure profits. Other firms, attracted by the potential profits, would enter the industry. Price would fall as more firms entered. At the same time, other industries, suffering from decreased demand, would be con tracting. The expanding industry would have an increased demand for inputs while the contracting industries would have a decreased demand for inputs. Constant industry costs would be possible in two different ways. If the expanding industry purchased only Inputs having perfectly elastic sup ply, input prices to the expanding industry would remain constant despite input demand changes. If the expanding in dustry purchased inputs in the same proportions as contract ing Industries released inputs, the input prices would be unaffected by input demand changes. Input demands would fi nally be the same after the total adjustment was made as they were originally. Throughout the adjustment, costs to the firms in the expanding industry would remain the same. The cost structure of the illustrated firm would remain the same, i_.e_. AC and MC. As more and more firms entered, the price would approach Oa and when it finally reached Oa, in dustry equilibrium would be attained. Bach firm would have minimum average cost equal to Oa and the total industry out put would be ON. The new short-run industry supply would be SRS', which is the sum of the marginal costs of existing firms operating with optimum plant. If firms have constant costs to scale in a constant cost Industry, the firm equilibrium is indeterminate. The industry equilibrium, however, will be determinate. The sum of the long-run costs, both average and marginal, will coin cide with the long-run industry supply, LRS. If a market demand of DD exists, total output will equal OM with a price of Oa. The short-run industry siipply of SRS would be the sum of the marginal costs of all the firms, one or many, operating with some scale of plant. Oa would be the mini mum point on the short-run average cost curve of the exis ting plant for each firm. Any increase in output would de crease the price, causing firms to contract or leave the in dustry. Any decrease in output would raise the price, thus bringing profits. Assuming free entrance, the profits would result in an expansion in industry output by entrance of new firms, expansion of existing firms or a combination of both. While firm output and number of firms is indeterminate, total industry output and price is determinate. If demand Increased to dd, price would rise to the in tersection of dd and SRS. The resulting profits would bring entrance and expansion until the sum of the plant marginal costs at the new equilibrium would be SRS', the new short- run industry supply. The price, of course, would be Oa. Constant costs to scale with constant industry cost is typical of arbitrage through space with arbitrage operations representing a small portion of the total market for some Ill commodity. In Figure 11, it can be assumed that Oa is equal to the existing price at X plus transportation costs from X to Y. With a demand for the product at Y of DD, OM would be the total purchased for shipment at X for delivery at Y. If the amount shipped were less than OM, the price would rise, causing arbitrage. The equilibrium operations of any one arbitrager would be indeterminate, but the effect of the total operations could be predicted with certainty. Arbit rage would continue until OM was the total shipped at a price of Oa. An Increase in demand to dd would raise the price at Y. Shipment from X to Y would immediately In crease. The increased shipment would increase until ON was shipped at a price of Oa. Individual operations would not be determinate, though total operations would be both deter minate and, under given conditions, quite stable. Constant industry cost is no bar, therefore, to the achievement of an optimum. An optimum equilibrium, price equal to marginal cost, will be achieved with either in creasing costs to scale or constant costs bo scale. This is assuming, of course, that freedom of entrance exists. 3. Decreasing Industry Costs It Is often held that a decreasing cost industry cannot have a competitive equilibrium. It will be seen in the fol lowing discussion that such a view7 is not correct. Confu sion between an industry made up of firms having decreasing costs to scale and an Industry having decreasing cost as 112 total industry output increases because of entrance of new firms is the probable source for such a view. As has been shown previously, firms having decreasing costs to scale cannot find a competitive equilibrium. However, if firm costs are such that the firms can reach competitive equilib rium, the Industr;/ can also reach a stable equilibrium re gardless of the industry cost pattern. To this point it has been shown that increasing and constant industry costs offer no bar to competitive equilibrium. It now becomes necessary to show that decreasing industry costs and competitive equi librium are not incompatible. Figure 12 illustrates the Industry decreasing cost equilibrium situation. Industry output with the given sup ply curve, LRS, and demand curve, DD, would be OM at price Oa. An individual firm with diseconomies to scale would operate with minimum long-run average cost equal to Oa and some output, Om in the illustration. All firms in the In dustry would operate with minimum average cost of Oa, though not necessarily at equal outputs. The sum of all firm out puts would equal OM. Assuming an increase In demand, & new de:nand curve dd, the first result would be a higher price. Hie price would increase to the Intersection of dd and the short-run supply SRS, which Is the sum of the marginal cost curves of the optimum plant for all firms operating. As a result of the price Increase and the resultant profits to firms in the 113 Figure 12 Decreasing Industry Costs MC AC LRS SRS' O u t p u t M 114 Industry, new firms would enter. Tills entrance of new firms would result in a decrease in Input prices to all firms in the industry. Input prices would be decreased because the expanding indiistry would be hiring more inputs with perfectly elastic supply but hiring fewer Inputs having relatively inelastic supply than the contracting Industry was releasing. Some input prices woLild decrease, some would remain constant. The price per composite bundle of Inputs to the expanding industry would decrease. As a result, all firms In the in dustry would have a lower cost structure, I_.<3. minimum aver age cost would be lower. As output expanded as a result of more firms entering the industry, the price of the product would decrease. Be cause of the greater slope of the demand curve, the price would fall more rapidly than the costs. Eventually a new equilibrium point would be reached, Industry output of ON and price of Ob in Figure 12. All firms In the industry would have a new minimum average cost equal to the price. Each firm would adjust its own output to the given price, jL. Figure 12 also can be used to illustrate the situation prevailing with constant costs to scale and decreasing In dustry cost. With an industry output of OM, the short-run 115 industry supply would be SRS. SRS is the sum of the margin al costs for all firms operating with some scale of plant. The size of the individual plants is indeterminate as is the number of firms. The sum of the long-run marginal costs (equal to average cost) of all firms operating would be hor izontal line ae. The summation is to the point of actual operations for all firms. With an industry output of ON, a new short-run supply, SRST, would exist. The sum of the long-run marginal cost (also average cost) would be bf. The lower cost structure would result from lower input prices at the greater indus try OUtTJUt of ON. Assuming a demand of DD, the price of the product would be Oa and industry output of OM. A smaller output would re sult in profits and, therefore, expansion either by existing firms, new firms or a combination of both. Greater output would result in losses, for price would be lower than costs. With freedom of entrance, a price higher than the marginal cost (equal to average cost) could not be charged. To do so, would be to attract entrants who would enter with iden tical cost conditions. Any attempt to achieve monopoly profits, therefore, would be defeated. An optimum price (equal to marginal cost) would be the long-run equilibrium result. This would be true even if one firm produced the total output, if freedom of entrance were maintained. A demand increase, to dd, due to a change in 116 tastes would immediately raise price to the intersection of SRS and dd. Hie resulting profits would result in industry expansion. The expansion would continue until the total industry output of ON with a price of Ob was reached. The marginal and average costs for all firms in the industry would also equal Ob. A smaller output would raise price but the resulting profits would attract entrants, wiping out the profits. A larger output would cause losses and, therefore, industry contraction. The input prices would drop because of the release of inputs by the contracting industries. The expanding indus- try, purchasing a smaller proportion of inputs with rising supply price and a greater proportion of inputs with con stant supply price than the releasing industry, would have an over-all drop in input prices. The drop in input prices would affect not only existing firms but also potential en trants. Because of the general impact of falling input prices to all firms, no one firm could gain any competitive advantage. The new equilibrium would also result in an op timum pri ce . G. Summary and Conclusions Competitive equilibrium of the firm will result, no barriers to entrance existing, with increasing costs to scale. A combination of constant costs to scale and freedom of entrance will result in an optimum price (equal to mar ginal cost). While the equilibrium situation with constant 117 costs to scale can hardly be called a competitive result be cause of the possible existence of only one firm, neither can it be called a monopolistic result for a monopoly profit cannot be attained, if freedom of entrance exists. For the purposes of this study, the important matter is that under both increasing and constant costs to scale the optimum solution will be achieved, if freedom of entrance exists. If decreasing costs to scale exist, individual firms can benefit from expansion. Such expansion results finally in a monopolistic situation incompatible with the requirements for an optimum. If firms have internal cost conditions compatible with the optimum solution, i_.£. increasing or constant marginal costs, industry cost patterns furnish no barrier to optimum equilibrium. Such optimum equilibrium is possible with in creasing, constant or decreasing industry cost patterns. The foregoing discussion, it should be noticed, was based on static assumptions. Some mention should be made of elements of change which may bring monopoly. Innovation is the type of change commonly considered as making monopoly possible. The entrepreneur, as a result of an innovation, makes possible returns In excess of competitive returns. These returns coiild "be attributed to the fact that, for a time, his method is exclusively his own. Hence they might be called monopoly gains. Such gains would be expected 118 J. A. Schumpeter, "The Creative Response in Economic History," Journal of Sconomlc History, Vol, VII (November, 1947), p. 155 reprinted in Essays of J. A. Schumpeter, ed. by R. V. Clemence, (Cambridge: Add!son-Y/elsey Press, Inc., 1952), p. 222. to disappear after the Initial innovational advantage is lost. Eventually it Is to be expected that these innova' tional monopoly returns are passed on to the consumers in the form of lower prices. "But even here we meet the prac tice of innovators striving to keep their returns alive by means of patents and In other ways. The gains described above shade off Into gains from purposive restriction of competition and create difficulties of diagnosis that are sometimes insurmountable . "3-2 12Ibld., p. 222. Because of the unique and non-continuous character of Innovation, it is almost impossible to handle in a rigorous equilibrium analysis. The best that can be done Is to qual ify the discussion of this chapter by pointing out that monopoly may also result from a breaking up of a competitive equilibrium by Innovation. How long such monopoly can exist depends on the degree of freedom of entrance and the possi ble ability or inability of continued innovation keeping a firm In front of competitors. Thus, the preceding analysis must be modified by the dynamic considerations of innova- 119 tional monopolies. -^Policy related to such monopolies will be considered in Chapter V. PART TWO THE THEORY OF PUBLIC UTILITY PRICING In Part One of this study equilibrium conditions under the several cost patterns were considered from the viewpoint of economic welfare theory. In this part of the study the same basic consideration will be pursued In the direction of policy from the viewpoint of the theory. In this discussion the standards for an optimum derived from economic welfare theory will be the only standards considered. Non-econoraic considerations will be given attention in Part Three. Governmental entrance into the economic sphere in the form of setting prices in certain industries will be at the focus of attention. Before stating the principles of price- setting it will be necessary to consider the logic of such entrance. This necessarily involves the placing of those industries to be regulated Into one class. In other words, the first task will be to define the industries to be sub ject to the price-setting commonly considered to be a char acteristic of public utilities. The principal requirement of the definition will be consistency with the standards for an economic optimum developed In Chapter II. Once the defi nition of public utility Is derived, attention can be di rected to the theory of public utility price-setting. Pricing of public utility services will be considered In two broad categories. First, the problem of the long-run 120 121 price will be considered. The problem to be attacked will be that of the setting of the optimum long-run price. Then, several other problems can be given consideration. These are public utility pricing in the short-run, pricing prac tices with new and expanding industries, and pricing prac tices in contracting and dying industries. Attention should be called to the general nature of the theory of public ^ltility pricing in Part Two. The theory will be considered as being applicable to any industry placed in the public utility classification according to the definition derived in Chapter V. No attention will be given to the peculiarities of particular industries. Complete knowledge of demand and cost conditions will be assumed. Practical problems will be the focus of attention in Part Three. CHAPTER V THE CRITERIA FOR PUBLIC UTILITY CLASSIFICATION As has been stated above, the principal purpose of this study Is to Investigate pricing in public utility Industries with the aim of determining optimum pricing policies. In the previous chapters, the framework was developed. In Part One, the theory of production and cost was considered with a view of determining the nature of cost conditions which lead to the breakdown of optimal situations, i,.®,. those conditions in which optimal equilibrium does not re sult without interference of some type. Current welfare economics was the basis for the choosing of criteria for such an optimum. It is now necessary to consider the nature of those industries which are placed or warrant being placed in the public utility classification. It will then be pos sible to consider the problem of pricing in such Industries. Two approaches to the determination of a definition of public utility status will be made in this chapter. First, the conventional procedure of investigating the nature of those Industries placed in the public utility class by legal action will be repeated. This procedure is basically that of observing the industries placed in such a class and then attempting to find the common characteristics of such In dus tries. Next, a definition of public utility classification 122 123 will b© derived from a premise stating the goal which econ omic regulation presumably is attempting to achieve. This goal is the achievement of an economic optimum. In other words, the achievement of an economic optimum will be the assumed goal of economic regulation. All regulation will be considered with this goal in mind. In particular the def inition of public utility status will be deduced from this basic assumption. Public utility regulation, furthermore, will be required to be of such a nature as to contribute to the achievement of this goal. Methodologically this ap proach amounts to establishing a definition of class which is deducible from a value premise and placing individual cases within the class if they possess the characteristics stated in the definition of the class.This method, of "^The value premise chosen here is the desirability of an economic optimum. Whether other standards should or should not be chosen is not In question here. One could choose, if so minded, military preparedness, some type of social betterment, administrative feasibility or other such goals for a standard. course, differs radically from the first approach which at tempts to find common characteristics of individual cases which are in actual day to day life considered as belonging within the class. Once these two approaches are attempted, an evaluation can be made of each approach. The relationship or laclc of relationship between the legal standards and the postulated 124 economic standard can be explored. Then the limitations to the remainder of this study can be seen by the limit ations placed upon the economist, A. The Legal Criteria for Public Utility Regulation Before starting the examination of the court cases which deal with the legal classification of public utility, a note of caution should be inserted. The viewpoint pre sented is not one which has received emphasis in the liter ature. Furthermore, purely legal questions such as the legal nature of a contract, the limits of interstate com merce, the nature of common law, or other such points will be ignored. A relatively few basic court cases involve the defini tion of public utility industries and the correlative plac ing of specific industries under regulation. These cases are common to most public utility textbooks and case b o o k s . ^ p Typical of the textbook and case book treatment are: Emery Troxel, Economics of Public Utilities, (New York: Rinehart & G o 1947), Gh. 1; E. WZ Siemens’, Economics and Public Utili ties, (New York: Apple ton-Centure- Orof ts, 1950), Ch. 2; Francis X . Welch, Cases on Public Utility Regulation, (Washington; Public Utility Reports, 1946T* Chs . 1-2. Munn v. Illinois Is considered the basic case,^ In this 394 U. S. 113 (1877). case, the standard for placing an industry under public 125 utility control was said to be that of being ’’affected with 4 the public interest." That is, "when .... one devotes his 4Ibid., 126. property to a use in which the public has an interest, he, in effect, grants to the public an Interest In that use, and must submit to be controlled by the public for the common good, to the extent of the interest he has thus created."® 5Ibid., 126. The above is familiar to all reader of textbooks. However, the emphasis on the existence of monopoly power also should be made clear. Chief Justice Waite speaking for the Court stated, "It is apparent that all the elevating facilities through which these productions of ’seven or eight great States of the West’ must pass on the way 'to four or five of the States on the sea-shore' may be a vir tual monopoly...... They stand .... in the very ’gateway of commerce' and take toll from all who pass."® 6Ibid., 131. Two points of an economic nature seem t@ have been made in this case. First, a monopoly existed which had pov/er to set prices higher than competitive prices, i_.<3. "take a 126 toll". Second, this monopoly existed in an important area, namely in handling "vast productions".^ These two points ^These criteria are also pointed out by John Bauer, Transforming Publlc Utility Regulation, (New York: Harper & Bros., 1.956), p. 4. will be seen to have been the essential standards until a definite break came in 1934. It might even be said that these two requirements could be placed in a necessary and sufficient ordering. The existence of monopoly was the nec essary condition and the importance of the industry the suf ficient condition. Public utility control could be applied, according to the courts, if monopoly existed in an important area of economic activity. It will not be contended that the Justices were or were not conscious of these standards. But it will be contended that these two fairly explicit cri teria crept into all the important cases until 1934, masque rading under the loose costume of "affected with a public interes t". Accompanying the Munn case in making up the Granger cases were two railroad cases.® Discussions in both cases ®Chicago, Burlington and Quincy Railroad Company v. Iowa, 94 U.S. 155 (1877); Peik v. Chicago and Northwestern Railway Company, 94 U.S. 164 (~1877y!! were brief with reference to the Munn case being relied on for the argument. The court did mention in the CB&Q case 127 that railroads are "given extraordinary powers, in order that they may the better serve the public in that capac ity."^ This is an apparent recognition of a peculiar char- 994 U.S. 161. acteristic of railroads, thus perhaps it may also mean a recognition of the existence of natural monopoly conditions. The next important case to arise Involved the setting of fire insurance rates.In this case the existence of ■*-%eman Alliance Insurance Go. v. Kansas, 233 U.S. 389 (1914). monopoly received considerable emphasis. The Court stated, We may venture to observe that the price of in surance is not fixed over the counters of the com panies by what Adam Smith calls the higgling of the market, but formed In the councils of the under writers, promulgated In schedules of practically controlling constancy which the applicant for in surance is powerless to oppose and which, there fore, has led to the assertion that the business of insurance is of monopolistic character and that it is Illusory to speak of a liberty of contract. It Is in the alternative presented of accepting the rates of the companies or refraining from In surance, business impelling if not compelling it, that we may discover the Inducement of the Kansas s tatute.H •L1Ibld., 416-417. It should be noted that not only was there a statement about the existence of monopoly but the importance of fire 128 insurance to business received attention. Business reqiiire- raents were such, according to the Court, that fire insurance might be considered compulsory, i_.e_. "business impelling if not compelling it." In considering wage regulation in the Kansas packing Industry the Court avoided deciding whether food preparation was a business clothed with the public interest saying, "We are relieved from considering and deciding definitely whether preparation of food should be put in the third class of quasi public business, noted above."I2 l^Wolff Packing Company v. Court of Industrial Rela tions of Kansas, 262 U.S. 522 (539) (1925*^ The reason for such relief was the non-existence of monopoly which, therefore, meant no need for the public utility type of regulation such as was found necessary In other situations, for, in the Court’s words, Never has regulation of food preparation been ex tended to fixing wages or the prices to the public as In the cases cited above, where fear of monopoly prompted, and was held to justify, regulation of rates. There is no monopoly in the preparation of foods. The prices charged by plaintiff In error are, it is conceded, fixed by competition through out the country at large. Food Is now produced in greater volume and variety than ever before. Given uninterrupted interstate commerce, the sources of the food supply In Kansas are countrywide, a short supply Is not likely, and the danger from local monopolistic control less than ever.^3 15Ibid., 538. 129 The Court was saying, in other words,' that since neither monopoly nor potential monopoly existed, it was un necessary to consider other aspects of the industry. The necessary condition of monopoly not having been met, it was immaterial whether the sufficient condition of having par ticular importance was met or not. In declaring unconstitutional a New York law regulating theater ticket brokers’ charges the Court again emphasized the importance of monopoly.^ Interpreting Lord Hale’s Tyson & Brother v. Banton, 273 U.S. 418 (1927). views the Court said, If the King or the subject have a public wharf, to which all persons must come, because it is the wharf only licensed by the King, or there is no other wharf in that port, arbitrary and excessive charges cannot be made. For it is then affected with a public interest..... A theatre or other place of entertainment does not meet this conception of Lord Hale’s aphorism or fall within the reasons of the decisions of this court based upon it.15 15Ibid., 439. A New Jersey law regulating employment agencies was de clared unconstitutional.-^ The Court while not mentioning 1SRibnik v. McBride, 277 U.S. 350 (1928). the element of monopoly, leaned heavily on the Tyson case, 130 It also pointed out the similarity of the services offered. The inference easily could be drawn that the Court was, thus, thinking of the similar absence of monopoly. In discussing Tennessee regulation of gas and oil prices, the Court while not emphasizing the element of mo nopoly did mention It saying, "There is nothing in the point that the act in question may be justified on the ground that the sale of gasoline in Tennessee is monopolized be cause .... an inspection of the pleadings and of the affi davits submitted to the lower court discloses an utter failure to show the existence of such monopoly.1,17 17Willi arns v. Standard Oil Company of Louisiana, 278 U.S. 240 (1928). Monopoly received great emphasis in a case ruling un constitutional Oklahoma regulation of Ice manufacturing and sale.18 The Court said, 18New State Ice Company v. Liebmann, 285 U.S. 262 (1932) The control here asserted does not protect against monopoly, but tends to foster it. The aim is not to encourage competition, but to prevent it .... We are not able to see anything peculiar in the business here in question which distinguishes it from ordinary manufacture and production. It is said to be recent; but it is the character of the business and not the date when it began that Is determinative. It is not the case of a natural monopoly, or of an enterprise in its nature depen dent upon the grant of public priviliges. The 131 peculiar requirement before us was evidently not imposed to prevent a practical monopoly of the business, since its tendency is quite the contrary.^9 19Ibid., 279. In the same case, the dissenting opinion pointed out, among other things, the monopolistic character of the ice business, or at least the legislature's view of such monop olistic character. "The business of supplying ice is not only a necessity, «... but the legislature could also con sider that it is one which lends itself peculiarly to monop oly. Characteristically the business is conducted in local plants with a market narrowly limited in area, and this for the reason that ice manufactured at a distance cannot ef- 20 fectively compete with a plant on the ground.” 20Ibid., 291-292. This New State Ice case represents a somewhat unusual situation for the Supreme Court. There was a controversy, not usually pointed out in commentaries, over the facts of the case. The fact in question was the existence of monop oly. The majority of the Court, as stated in the majority opinion, felt that no problem of monopoly existed save as the regulatory law's existence brought such monopoly into being. The dissenting opinion, among other things, expres sed the view that the ice industry was by nature monopol- 132 istic.^ The issue of regulation to a great extent hinged Justice Brandeis in the dissenting opinion, in gener al, expressed the view that any area is subject to price regulation if the legislature desires to place it under such regulation. It is, thus, doubly interesting that he empha sized the existence of monopoly. on the view of the Justices as to the existence of monopoly. The majority feeling no monopoly existed, regulation was declared invalid. The next major case dealing with the control of prices, milk in this instance, marks a drastic change in the view point of the courts.The Court, in effect, made control 2%ebbla v. New York, 291 U.S. 502 (1934). of price admissible in attempting to reach any goal a legis lature deemed attainable through such control saying, We may as well say at once that the dairy indus try is not, in the accepted sense of the phrase, a public utility...... But if, as must be conceded, the industry is subject to regulation in the public interest, what constitutional principle bars the state correcting existing maladjustments by legis lation touching prices? We think there is no such principle.23 23Ibid. , 531-532. In dealing with regulation of coal prices by the Na tional Bituminous Coal Commission, under the Bituminous Coal Act of 1937, the Court made even more explicit the possible 153 use of price regulation to meet any goal deemed, attainable. The Court stated, It was the judgment of Congress that price-fixing and the elimination of -unfair competitive practices were appropriate methods for prevention of financial ruin, low wages, poor working conditions, strikes and disruptions of the channels of trade which fol lowed in the wake of the demoralized price struc tures in this industry. If the strategic character of this industry in our economy and the chaotic conditions which have prevailed in it do not jus tify legislation, it is difficult to Imagine what would.24 04 Sunshine Anthracite Coal Company v. Adkins, 310 U.S. 381 (395), (1940'} . At this point, It is possible to summarize the fore going decisions. For many years, until the Nebbia decision in 1934, the basis for public utility regulation was the existence of monopoly in an important industry. Importance was judged by legislatures and the courts. It should also be pointed out that the essence of public utility control was considered to be control of prices. This is evident in decisions declaring regulation unconstitutional, as for ex ample in the Tyson case where the Court said, The authority to regulate the conduct of a busi ness or to require a license comes from a branch of the police power and which may be quite dis tinct from the powers to fix prices. The latter, ordinarily does not exist in respect of merely private property or business but' exists only where the business or the property involved has become affected with a public interest.25 134 25273 U.S. 430. After the Nebbia decision, public utility type of con trol, i.e.. regulation of prices, was applied with no objec tion from the Supreme Court to any type of goal, economic or other, which could be attained through such regulation. At the present time, therefore, there are apparently no simple, consistently followed legal criteria for industries placed under public utility regulation. B. An Economic Criterion for Public Utility Classification Before attempting to arrive at an economic definition of public utility, it will be necessary first to consider briefly economic regulation and control in general. It will be necessary also to consider the basic economic standards applicable to such control. These standards are assumptions or value judgments which must be made as to the fundamental justifications for such control. Two principal assumptions are basic to the following discussion of economic regulation. First, it will be as sumed for purposes of this dissertation, that it is desira ble to reach maximum economic efficiency. This may mean conflict with other values, es.£. support or subsidy on hu manitarian or nostalgic grounds for submerged or classes en gaged in uneconomic production. Arguments as to the impor tance of such views are recognized but cannot' be considered 135 In this portion of the study. Second, it will be assumed that the desirable economic order is basically of a private nature. In other words, governmental powers are not to be used to regulate or to operate with government ownership un less there Is a definite economic advantage to such govern mental activity. If any activity could be operated equally efficiently by government ownership or with government regu lation or by unregulated private ownership, then the unregu lated private ownership is the desired choice. Governmental regulation will be preferred similarly over governmental ownership. According to the standards of maximum economic welfare developed in Chapter II, perfect competition would result in the optimum being achieved. Government power would be used, therefore, to maintain the benefits of perfect compe tition. It was shown in Part I that, except for decreasing costs to scale, maintenance of free entry would assure the attainment of the optimum. Under decreasing marginal ^®There is an implicit assumption here that the means of achieving the goal is in Itself not undesirable by some standard. This point, frequently neglected, is basic to policy prescriptions. costs to scale a monopoly would result with free entrance. Governmental activity could hardly operate to enforce compe tition in such a situation. Placing industries having de creasing marginal costs to scale In the class of regulated 136 industries called public utilities would seem to be the so lution. These cost conditions could be at least the primary requirement of an industry labelled as a public utility. A difficulty still remains. Any kind of administration costs someone in the economy. For the economy as a whole, the cost is in the resources used. Administration of public utilities is no exception. To increase economic welfare ac cording to economic welfare standards, regulation would be required to result in gains to individuals sufficient to offset this extra use of resources. The compensation prin ciple can be used, then, to determine the secondary condi tions in the definition of a public utility. The secondary condition could be stated simply as requiring that the tax ing away of the gains resulting from regulation from each individual would cover or more than cover the costs of regu lation. The formal requirements for defining an industry fall ing into the public utility classification are (1) that the production process is such that decreasing marginal costs prevail; and (2) that if gains from public utility regula tion were taxed away they would cover or more than cover ex- 27 penses of such regulation. 27This follows from Reder, oje. cit., p. 54. In a later chapter these costs of regulation will be expanded to cover other payments which are necessary. The importance of requiring both primary and secondary 137 conditions Is indicated by the possibility of natural monop oly in the production of trivial products. It was pointed out in Chapter IV that the conventional U shaped long-run average cost curve has a corresponding marginal cost curve which decreases over a certain area of potential output. It is possible that a production process which would reach min imum average cost at a small scale of plant would still be a natural monopoly. This would result when demand, in the sense of the entire schedule or curve, would be so small that one plant could satisfy this demand while operating in the area of decreasing marginal costs. A real life example of such a result possibly might be in certain types of local industries.2® in such instances the cost of regulation ^Concrete grave vault construction In a small town would be limited by the market size. Transport costs would limit outside competition. might be more than the benefits which would arise from such regulation. Such monopolies co’ald be permitted to exist un regulated. The monopolist would be better off than if he didn’t produce the product and the consumers would be at least no worse off, as they are willing to purchase the pro duct. The existence of such unregulated monopolies would benefit someone while regulation or the elimination of such production would result in someone losing. However, mar ginal condition five (equality of marginal rate of substi tution and marginal rate of transformation) would be vio 138 lated. Another major point which must be made clear is the nature of public utility regulation. This is necessary be cause of the myriad controls placed on all types of activi ties for all types of purposes. Some of the purposes for which controls exist are safety, health of plants, animals- and humans, meeting of minimum education standards and pre vention of fraud. These are achieved by various licensing and inspection laws. Such controls, however, are not con sidered, for the purpose of this study, as being at the core of public utility regulation. Control of the price of the individual product will be considered as central to public utility regulation. Such regulation may or may not be accompanied by other require ments such as minimum standards of performance, product quality standards, health, sanitation and safety standards or specific output requirements. The existence of control over price, however, does not mean that public utility regu lation, as defined above, exists. Price control may be used for non-economic, not to mention uneconomic, goals, js.g,. subsidization of silver mining, whereas this study defines public utility control as being used to attain an economic optimum. In other words, the control over price will be considered as a necessity for those industries defined as public utilities in this study and placed under public util ity regulation. Other controls may or may not accompany the 139 control over price. However, the mere existence of control over price does not place an industry in the public utility category. For the purposes of this study, public utility regula tion and, therefore, regulation of public utility prices does not include anti-cyclical controls. Such a view is based on two assumptions. First, no generally satisfactory explanation is available regarding the inter-relationship of individual prices and general anti-cyclical activity such as monetary and fiscal control. Also, much of the actual price setting having anti-cyclical overtones has been to protect specific groups from extreme suffering due to price and income fluctuations, <3.g. agricultural parity prices, and thus has more of a social or political orientation than economic. The value premise upon which such control is based is that certain groups should be protected to some de gree from general economic fluctuations. Such price setting may lead, in fact, to greater problems from the optimum allocation standpoint. The static assumptions behind this analysis of policy must be emphasized. Both the welfare criteria and the theo ry of the firm upon which the foregoing discussion was based were entirely static--given preference functions, production functions and quantities of goods and services available. Under such static conditions the uneconomic resource alloca tion resulting from monopoly is simply demonstrated and un 140 exceptionable. "iTnfortunately, once we leave the delightful simplicities of static equilibrium neither the concept nor the appraisal of monopoly retains its clarity, as the re- markable intellectual confusion which surrounds the anti trust laws demonstrates."®® 29Kenneth E. Boulding, The Organ!zatlonal Revolution, (New York: Harper & Bros., 1953), p. 39. Of particular importance in bring qualifications into the analysis are the dynamic considerations involving inno vation and advance in scientific knowledge. Once these con siderations are brought into the analysis a case may be made for the existence or even the encouragement of some degree of monopoly. Restriction of freedom of entrance through patent protection as an encouragement to Innovation is one aspect of the case for monopoly. Scientific research re quiring large funds made available only by large, wealthy firms is the other aspect of this argument. However, the case is not completely clear, for monopoly may also act to restrict monopoly disrupting innovation. 30 30Tibor Scitovsky, Welfare and Competitlon, (Chicago: Richard D. Irwin, Inc., 1951), pp. 428-431. Scitovsky gives a thorough, though brief, discussion of the possibilities. The postulated economic definition of public utility, which rests upon a static theory of the firm and static wel fare assumptions may also require qualification. Economies 141 of scale which bar a competitive equilibrium may exist. However, benefits greater than those from public utility control may possibly exist if no regulation is imposed. This would be true if, without control, an oligopoly made up of firms vigorously competing in innovation would exist. The production functions would differ from those prevailing under competition but rather than monopoly restriction, con tinued innovational and scientific advance might result. Such considerations, of course, may be of overriding importance in individual cases. However, as Scitovsky states, nNo fully developed and generally accepted theory 'Z] exists on this difficult subject." As a result of this 31Ibld., p. 430. lack, the basis of the analysis to follow will be the static theory developed in the previous chapters. However, the possible necessity for individual exceptions stemming from dynamic considerations must be kept in mind when application is made to actual regulatory policy. C. Existing Public Utility Control in the Light of the Postulated Economic Definition Any regulation, economic or other, or for that matter lack of regulation, is based fundamentally on some value premise. This is implied in the discussion above. By the postulated economic definition of public utility, optimum 142 allocation was considered to be desirable. Under legal standards until 1934 monopoly exploitation was considered to be bad, or, conversely, tempering monopoly exploitation by control of prices was considered to be good. After 1934 the Supreme Court accepted any value judgment chosen by the legislature as a valid justification for public utility con trol of price. The only qualification according to the Nebbia decision was that the control not be ’’arbitrary, dis criminatory, or demonstrably irrelevant to the policy the legislature is free to adopt, and hence an unnecessary and unwarranted interference with individual liberty."ti 32 32291 U.S. 539. The earlier legal standards had the advantage that there were definite criteria which were looked for, namely monopoly and the importance of the product or service. The disadvantage of these standards was that they were concealed under the phrase "affected with a public interest". This is at best a vague phrase. As Justice Field pointed out in his dissent to the Munn decision, the public has a great inter est in residence buildings, textiles of all types, machinery construction, printing of books and periodicals and the manufacture of useful or ornamental utensils as well as the 33 grain elevators dealt with in the Munn case. ^ xt was al- 143 3394 U.S. 140-141. most sure to follow that the use of this vague phrase would lead to the opening of the public utility classification to all and sundry types of economic activity. With the opening of public utility status to any indus try the question arises as to what the economist can say about such action. One approach might be to attack as Horace Gray did, saying "Henceforth, the public utility status was to be the haven of refuge for all aspiring monop olists who found it too difficult, too costly, or too pre carious to secure and maintain monopoly by private action alone,"3^ On the other hand, the characteristics of each ^Horace Gray, "The Passing of the Public Utility Con cept," Journal of Land and Pub11c Utili ty Hoonomlcs, Vol. XVI (February, 1940), p. 9, industry placed in the public utility classification could be listed in the perpetual attempt to find the common char acteristics of such industries. With the first approach there is an implicit judgment that only economic goals are important. On that basis any regulation which uses the public utility approach to attain monopoly in order to stabilize or to support a sick Indus- try, e_.£. bituminous coal, would be attacked as being un desirable. The economist who does this In effect is saying that any norm, goal or value which conflicts with the econ 144 omic goal of optimum allocation is automatically to be ruled out as not worthy of being achieved. This is a rather sweeping value judgment. It is very questionable whether an economist is qualified to make such a judgment. The second approach of trying to find the characteris tics common to all industries placed in the public utility status seems to have two basic difficulties. If the conven tional attempts are any evidence, not much success can be expected to attend such attempts.Also, to the extent ^See, for example, J. M. Clark, Social Control of Buslne ss, (New York: McGraw-Hill Book Co., 1939 ) , Chs. XV- XVIII; Temporary National Economic Committee, Economic Stan dards of Government Price Control, (Monograph No. 32, 76th Congress, 3rd Session,”1941). that success is attained, there is a seeming justification of such regulation in all cases, even including those which show marked disadvantages. That is to say, the economist would not function as a critical analyst of the industries placed under regulation and the economic consequences of such classification. In general, It would seem, the economist's function is to point out the economic implications of any action of the social group. In particular with public utility regulation, he could point out two things. If the means, public utility regulation, are of such nature as to make possible the attainment of whatever end is sought by the use of these particular means is one question upon which light could be 145 shed. The economist also could point out the results upon the economy, upon the allocation of resources, of such action. By such analysis, the public, perhaps, could re ceive sufficient knowledge to judge whether the particular means used are adequate and also whether the ends gained are worth any possible loss of economic efficiency. The econo mist also would be able to stay out of an area in which he is not an expert, namely that of judging which end is of greatest importance. The economic definition postulated above, therefore, would not be used as an all inclusive requirement for plac ing any industry in the public utility classification. In stead, it would serve solely as the economic requirement for placing any industry under public utility regulation. There would be no ruling out the use of such regulation to meet other goals or criteria based on social, political, military or other grounds. However, the postulated definition and the framework of economic analysis surrounding it could be used to determine the economic implications of. the use of public utility type of regulation to meet other goals. The probability of the successful achievement of the non-econ- omic goals could also be judged by the use of these economic standards as points of reference. The function of.the economist as analyst of public utilities here suggested may be distasteful to some econom ists. It surely does limit free-wheeling judgments as to 146 the correctness of statutory provisions and gives the econ omist a more limited function in the entire area of public utilities. On the other hand, it is helpful in giving the economist such perspective that he is better able to resist the temptation of being dogmatic in areas in which he can hardly be considered as expert, or perhaps even adequate. The remainder of this study, which is to be devoted to pricing in public utilities, will assume these limitations placed upon the economist. It will be recognized that the pricing policies are to be considered from the standpoint of their effect on allocation of resources. The goal toward which such policies are to be directed is optimum alloca tion. Whether there are other more desirable goals of a non-economic nature will not be considered as within the scope of this study. CHAPTER VI THE DETERMINATION OP LONG-RUN PUBLIC UTILITY PRICES Up to this point, consideration has been given to situ ations in which competition breaks down, the nature of an economic optimum and the criteria for public utility classi fication. Having presented the economic basis for public utility regulation, It Is now possible to consider the na ture of price setting to achieve optimum economic welfare. Such price setting has two aspects, long-run and short-run. This chapter will be concerned with the problem of setting long-run prices. The next chapter will be concerned, along with other matters, with short-run pricing. Marginal-cost and average-cost pricing have been pro posed as the basic principles of public utility price set ting. Each of these will be considered as a basis for long- run pricing. Because of the nature of the marginal cost solution, an investigation of the nature and implications of required payments and compensations will be necessary. It will then be possible to evaluate the two proposals. A. The Possible Choices , Before proceeding to consider possible standards for long-run pricing, it is necessary to state explicitly the meaning of the concept of long-run. The long-run as used in this study means, as it does conventionally in economics, a period sufficiently long that all factors are variable or conversely that none are fixed. The specific problem be 147 148 comes one of setting a long-run price for a public utility service, given the long-run demand curve and the long-run cost functions for the production of the service. The cho sen price should result in the proper resource use. Two suggestions have been made as to pricing bases which result in proper use of resources. These bases are marginal cost and average cost. Each will be considered as a possible standard for public utility pricing.^ -'-The principal arguments for marginal-cost pricing can be found in H. Hotelling, "The General Welfare in Relation to Problems of Taxation and of Railway and Utility Rates," Econometrica, Vol. VI (July, 1938), pp. 242-269; and A. P. Lerner, "Statics and Dynamics in Socialist Economics," Econ- omic Journal, Vol. XLVII (June, 1937), pp. 253-270. The principal arguments for average' cost pricing can be found in R. H. Goase, "The Marginal Cost Controversy," Economica, Vol. XIII N.S. (November, 1946), pp. 169-182; and T. Wilson, "Price and Outlay Policy of State Enterprise," Economic Journal, Vol. LV (December, 1945), pp. 454-461. An excel lent summary of the debate is given by Nancy Ruggles, "The Welfare Basis of the Marginal Cost Pricing Principle," Re view of Economic Studies, Vol. XVII (1949-1950), pp. 24-46; id., ""Recent Developments in the Theory of Marginal Cost Pricing," ibid., pp. 107-126. 1. Marginal-Cost Pricing The case for marginal-cost pricing was presented in Chapter II. Under such a system of pricing an economic op timum results, i_.£. under the given conditions there would be no advantage to shifting resources to other lines of en deavor or for any consumer to shift the quantities of his purchases. However, with the existence of firms having the characteristics of public utilities as defined in the pre- 149 vious chapter1 some difficulties result in implementing mar ginal-cost pricing. It will be necessary to point out these difficulties and to consider possible solutions. Figure 13 illustrates cost and demand conditions with economies to scale prevailing. As a competitive solution is impossible under such cost conditions, It is just such a situation which would come under public utility control. Given these cost and demand data, the problem to be solved is one of the proper price. LAO and LMG are the long-run average and marginal cost curves respectively. SAG and SMC are respectively the short-run average and marginal cost curves for one of an infinite number of possible scales of plant. AR represents the given demand with MR being the marginal revenue curve. If marginal-cost pricing is to be followed in the long- run, then, In Figure 13, the price should be OA, the output should be OF and a scale of plant having short-run average cost curve SAG should be in operation over the long-run. The utility commission, therefore, should set the price at OA. Consumers would purchase a quantity of OS at this price. Any greater amount could only be sold at a lower pri ce. It should be noted that at the OE output the demand curve, AR, intersects the long-run marginal cost curve and the short-run marginal cost curve. Also, the long-run aver age cost is equal to the short-run average cost. More in- ov 0 Marginal-cost PricingMarginal-cost MR Figure 13 Figure £ LAC SAC 150 151 teresting, however, is the fact that the optimum plant for producing the long-run output OE would not be operating at full capacity at output OS. The output of the optimum plant would be considerably short of the minimum short-run average cost of this particular size of plant. This might lead to misinterpretation by a casual observer of such a plant in operation. On the one hand there possibly might be clamor for the utilization of "wasted" plant capacity. However, an output equal to that provided at minimum average cost by the plant represented by SAG would be more efficiently provided by a larger scale plant. The proper size plant would be the one having its average cost curve tangent to the long-run average cost curve at the required output. There also might be clamor for a smaller plant which could provide OE output at minimum average cost. This would result in higher aver age cost per unit, however. The plant represented by SAG provides OE output with the least possible use of resources. In the situation being considered, the biggest problem connected with marginal-cost pricing is the failure of total revenue to meet total cost. In Figure 15, the deficit per unit produced is equal to AB. Total revenue is deficient in covering total costs by an amount of AB times the output OE, or an amount equal to the area ABCD. If the firm is to exist over the long-run, it is necessary to provide the firm with such an amount. This would mean that the firm operat ing in such a situation would receive a subsidy equal to 152 ABCD. The problem becomes one of taxing certain individuals to pay this subsidy. A later section of this chapter will consider the implications of such taxing. One point of frequent confusion should be mentioned here. In articles on marginal-cost pricing the statement is made quite frequently that the amount of the subsidy is equal to the fixed costs.^ As Samuelson has pointed out, Q See Hotelling, op. cit., p. 257. 3 marginal cost cannot be considered as part of total cost. Samuelson, Foundations, op. cit., p. 241. It is merely the difference in total costs at two different outputs. Whether ABCD is equal to, less than or greater than fixed costs cannot be determined unless the breakdown between fixed and variable costs at output OE with the given size of plant is known. In general, it can have any of the three possible relationships to fixed costs. However, only if the minimum average variable costs for the given size plant occurred at output OE would ABCD equal fixed cost. This is true because minimum average variable costs are al ways equal to short-run marginal costs for any fixed size plant. The marginal cost curve intersects the average vari able cost curve from below at the output at which average variable costs are a minimum. 153 Discussions of marginal-cost pricing are rarely expli cit as to whether they are discussing long-run or short-run pricing. Hotelling, while he does consider advisability of new investment by relating it to consumers’ surplus, uses an already constructed bridge as his explanatory example.^ ^Hotelling, o£. cit., pp. 260-263. Other writers also emphasize short-run pricing. Lerner is an exception. His formulation, however, obviates a possible subsidy payment to make up losses.® ®Lerner, "Statics and Dynamics," ojc. ci t., pp. 261-263. Limitation of the discussion to the short-run problem is unfortunate. If marginal-cost pricing is to be used as a standard, then it should be expected to give the amounts of all inputs to be used over the long-run in the production of the optimum output. That includes, or perhaps even makes most important, the choice of plant size. Given the tech nical conditions and all input prices which determine the costs and given specific demand conditions any rule of pric ing will result In certain amounts of all Inputs to be used. The size of the investment, plant size, will be in cluded along with all other inputs. Thus, in Figure 13, a plant having short-run average cost curve SAG would be con structed. It would be the most efficient plant for produc- 154 ing the given output, OE, the ratio of its discounted marginal physical product to input price would equal the same ratio for any other input. One important difficulty remains. The use of marginal cost as the pricing standard assumes such a principle to be used in all firms and industries in the economy. This means either that perfect competition prevails or else that ap plied controls use marginal cost as the standard in all seg ments of the economy. If any imperfections exist, price Is greater than marginal cost, then marginal-cost pricing will not result In the Ideal output. This follows from the fifth marginal condition for an economic optimum which requires the marginal rate of transformation between any two products to be equal to the ratios of their marginal costs and of their prices. For this to be true in comparing the utility service and a product produced under any imper fect market condition, the utility service could not be priced at marginal cost. Instead, the public utility price should be greater than marginal cost by the same ratio as the product sold in the imperfect market Is priced above its marginal cost. Where various ratios between prices and ®A. M. Henderson, "The Pricing of Public Utility Under takings," Manchester School, Vol. XV (1947), p. 242. marginal costs exist (because of varying degrees of imper fection), Little's solution is that the utility price should 155 equal the weighted average of those ratios existing else- 17 where. ''’Little, Welfare Economics, op. clt., p. 180. On the face of it the above argument seems plausible enough. Strictly speaking, however, welfare economics can not make a "should be" statement as is done above. As Samuelson has pointed out, only "’twere better" statements can be made.® It would be better by the economic welfare O Samuelson, Foundations, op. cit., pp. 249-252. standards to set the utility price at some point above mar ginal costs when Imperfections exist. Better still would be for all products to be priced at marginal costs. If the welfare standards are to throw light on problems, they must be used as standards of evaluation. On the basis of the static welfare standards it can be stated that market imper fections interfere with the achievement of an economic opti mum. Of course, dynamic conditions may require some quali fication to such conclusions. Criticism on welfare grounds of imperfections existing in other markets, however, does not help a utility commis sion setting a public utility price. As the commission has no control over other markets, it probably would settle, if It does aim at the economic optimum as a goal, for a price 156 somewhat above marginal cost. Ideally it would do this and then bring the price closer to marginal cost as imperfec tions are eliminated, if they are at all, in other markets. It should be noted that a procedure of pricing above marginal cost because of these necessities does not Invali date marginal-cost pricing. Marginal cost is still the basis for price setting. The deviations from it because of imperfections elsewhere themselves use marginal cost as the pricing base. 2. Average-Cost Pricing The case for average-cost pricing is based on the pre mise that any consumer purchasing any commodity should bear the total costs of the production of the commodity.^ No ar- ®R. II. Goase, "Price and Output Policy of State Enter prise: A Comment," Economic Journal, Vol. LV (April, 1945), p. 112. gument is presented that average-cost pricing would maximize anything. There is, apparently, only the judgment that if resources are to be commanded by the consumers, then the consumers should pay the costs of these resources. Each consumer is to pay the same proportion of the total cost as he commands of total output. Figure 14 illustrates the long-run solution if pricing were to follow the average cost pricing principle. The price fixed by the commission would be set at OA at which 157 Figure 14 Average-Gost Pricing SMC LAC LMC MR QuaN TITf it 158 price consumers would purchase OB quantity of utility ser vices. Plant size would be that represented by short-run cost curves SAG and SMC, average and marginal costs respec tively. As in Figure 13, this plant size having its aver age cost curve tangent to the long-run average cost curve LAC at output OB would furnish the least expensive long-run production of OB output. The short-run marginal cost SMC would be equal to long-run marginal cost LMG at this output. A comparison can be made here of the differences in output if the firm were monopolistic, .i.e.. uncontrolled, if average-cost pricing were followed or if marginal-cost pric ing were followed. The monopoly output would be the smal lest, being set where MR equals MC, at OD. The price would be the amount consumers would pay over the long-run for such output. This price is given by the demand curve AR and would equal DE. Average-cost pricing would result in a lower price OA and an output OB. Marginal-cost pricing would res\;.lt in the largest output, 00, and the lowest price, equal to CF. The actual differences prevailing between pos sible prices and outputs would for any individual firm de pend upon the specific elasticities of the demand curve and cost curves. The order of magnitude would remain as stated. As with the long-run marginal-cost pricing solution, the long-run average cost solution gives the appearance of excess capacity. At output OB the most efficient size plant represented by SAC in Figure 14 would operate at an output 159 smaller than minimum average cost output. This might lead to calls for "more efficient use of existing plant capaci ty". A proponent of average-cost pricing seems to fall into such an error. Harry Norris states that for him the "golden rule" is to make adjustments until price equals average cost. However, "there are certain instances in which a loss is ad missible, namely where there is excess capacity which may as well be utilized if revenue will meet currently-disbursed costs (to the extent that it will, in fact)."-*-® •^Iiarry Norris, "State Enterprise and Output Policy and the Problem of Cost Imputation," Economlca, Vol. XIV N.S. (February, 1947), p. 61. In the situation represented in Figure 14, this propo sal to make use of excess capacity would result in output OG and a price of GH. This would be an ironical situation if accompanied by pretensions of efficiency In the use of re sources. Consumers would not be charged a price covering total costs, the basic average cost argument. The price would not be at marginal cost but a subsidy would still be needed. But worst of all, the output would be produced over the long-run by a scale of plant that was relatively inef ficient in the production of OG output. The result would be even more ironic if it should happen that the price required to operate at minimum short-run average cost would be less than the long-run marginal cost solution and with a conse quent greater output. This could occur if the demand were 160 quite Inelastic and the short-run average cost curve were rather shallow, !_•©.. If minimum short-run average cost would be reached at an output considerably greater than the point of tangency between short-run and long-run average cost. The previous discussion of the lack of explicit differ entiation between long-run and short-run in discussions of marginal-cost pricing has its counterpart in discussions of average-cost pricing. Goase in his discussion uses an ex ample which is quite obviously of a short-run nature.He Goase, "Marginal Cost Controversy," oje. cit., p. 171. postulates a central market which delivers by truck along radiating roads. Consumers would be charged a fixed trans portation charge and a marginal charge equal to the price of the good delivered. A fixed size delivery facility, i_.a_. the truck is assumed. It should be pointed out that unless each consumer was assessed a transportation charge equal to the amount of the transport facility for which he was re sponsible, the main rule of average-cost pricing would be violated. B. Payments Necessary with Marginal-Cost Pricing Under the cost conditions prevailing when a public utility exists according to the definition presented In the previous chapter, a loss exists if marginal-cost pricing is followed. This loss is equal to average costs minus mar- 161 ginal costs at the chosen output times the particular out put. In Figure 13 this is equal to ABOD. IF a public util ity Firm is to stay in operation over the long-run, this amount must be received. The only way in which this payment could be received would be through a subsidy payment. Such a subsidy must be paid by some individuals in some manner. It will be necessary to decide as to the individ uals who are to pay the subsidy and the manner oF the pay ment. The question as to whether suFFlcient payment can be made without violating economic welFare principles also must be met. This section will consider these matters. The problem oF the Individuals who are to pay will receive con sideration First. Then, the method oF collection will be considered. 1. Consumers’ Payment A major criticism oF marginal-cost pricing is based on the necessity oF taxing to pay the subsidy which would be required IF the Firm were to continue to operate. Goase stated that it would result in redistribution oF i n c o m e s . -^Coase, "Price and Output Policy oF State Enterprise: A Comment," o j d . cl t., p. 113. The community at large would be taxed For the beneFit oF purchasers oF certain products. IF such were the result oF marginal-cost pricing, it would be in violation oF the Fun damental principle oF welFare economics. That principle 162 defines a known increase in welfare as resulting only if one or more persons gain and no one loses. If some individuals gain at the expense of others, no welfare conclusion can be drawn. In considering the amount of the subsidy and the tax it is necessary, therefore, to tax so that no losses to Indi viduals are Incurred. The utility consumers who gain from the marginal-cost pricing must be the ones who are taxed to pay the subsidy. Furthermore, the tax must not be so large as to tax any one consumer more than he gains from pricing at marginal cost rather than at average cost. In other words, the maximum which any individual Is to be taxed equals the amount which can be taken from him with marginal- cost pricing In effect while leaving him as well off as if average-cost pricing were used. This amount which can be taken from a person to compensate for a price decrease is called the compensating variation by Hicks.^ Figure 15 Is -^J. R. Hicks, Value and Capital, Second Edition, (Oxford: Oxford University Press, 1946), pp. 40-41. useful in explaining this concept. In Figure 15 a portion of a consumer’s indifference map Is Illustrated. I q_ and I2 represent two indifference curves between Y, money, and X, the utility service. Assume the consumer has 0M]_ income per period of time to divide between quantities of the utility service and quantities of money Figure 15 Tax on Public Utility Consumer 164 (to be used for other commodities and for saving), If the price of the utility service Is that represented by price line P-^, the consumer will purchase x^ of utility service and retain X]_A of his income for other uses. This would put him tangent to the highest possible Indifference curve, 1-^. Given P-^ price, any change In the consumer’s expenditures would put him on a lower Indifference curve. Generalizing to all purchases, this amounts to maxi mizing utility, u, as measured by some arbitrary index func tion of utility, p(x-i , x2, xn ) where x^, is the quanti ty of the rth commodity, "subject to a given income. It is required that u increases as $ Increases and that total ex penditures be limited by the given Income. Thus, u - x2, is to be maximized sub ject to c = X]_P1 / xgpg / ••••/ xnPn where pr Is the price of the rth commodity. By the method of the Lagrangean mul tiplier, -m, u - gf(x^, Xg, * *. ., x^) — m(xjj_p]_ £ xgpg / ... xnpn ). “ = - mp-, = 0 m I — 3xq 1 1 Pi 3U , v ™ - ^2 axg- ^2 “ mP2 = 0 ” " p£ 4 = - mPn = 0 m = k m - fa. - £ 2 - i n Pi P 2 Pn ^r £r s z 1, w, ...., n) Pa ~ Ps r / s That Is, the marginal rate of substitution between any two commodities will be equal to the ratio of their prices, or the price line will.be tangent to the indifference curve, If the price were decreased to a level represented by 165 price line Pg, the consumer would purchase Xg quantity of the utility service. He would now be on a higher indiffer ence curve, Ig. Furthermore, if the consumer were charged price Pg, he could absorb a loss of income equal to MgM^ scad still be as well off as he was initially with the P^_ price. In other words, with the lower income of OMg and a price the consumer would be on indifference curve 1-^ as he would with the higher income of OM^ and a price line P]_. This loss of income Is called by Hicks the compensating variation, the amount of income loss which just compensates for the price decrease to keep the consumer on the same utility level. The compensating variation concept can be applied to the problem of the necessary tax and subsidy under marginal- cost pricing. Letting the price line P^ represent the price which would exist under average-cost pricing, the consumer would be on indifference curve 1^ while paying average costs. Letting price line Pg represent the lower price which would exist with marginal-cost pricing, the consumer would be on the higher indifference curve Ig. With mar ginal-cost pricing the consumer could, therefore, give up income equal to MgMp and still be as well off as he would be with average-cost pricing. The consumer could be taxed MgM-^ to pay the subsidy and still be as well off as he would be paying average costs. If the amount of the compensating variation for all utility consumers covers or more than 166 covers the' necessary subsidy, the tax could be levied, and the subsidy paid. No one in the economy would be worse off, as compared with average-cost pricing. To be completely accurate, however, the tax which each consumer would be required to bear would also include a pay ment for the costs of collecting the tax. The necessary re quirement would be amended, then. If the sum of the incomes which could be taxed away with marginal-cost pricing, leav ing consumers as well off as under average-cost pricing is equal to or greater than the necessary subsidy plus the costs of handling the tax and subsidy, marginal-cost pricing meets the welfare requirement. A question might arise as to whether the necessary pay ment should include regulation costs. The costs of regula tion, however, should be approximately the same whether mar ginal-cost pricing or average-cost pricing is used. The costs of regulation must be met In either situation. These costs, therefore, would have been taken care of previously. Sufficiency of the taxed away gains from marginal-cost pricing in meeting the required subsidy is another signifi cant question. If the gain in the Marshallian consumers' surplus resulting from marginal-cost pricing Is greater than the necessary subsidy, such payment can be made. In Figure 16, assuming OF to be the price with average-cost pricing and OA to be the price with marginal-cost pricing, the in crease in Marshallian consumers' surplus is equal to AFGH. 167 Figure 16 Increase in Consigners' Surplus F A 0 X 168 If AFG-H is greater than the necessary subsidy plus the cost of handling the tax and subsidy, marginal-cost pricing is possible without anyone being placed in a worse position. Such a statement about the Marshallian consumers' sur plus can be made because of its relation to the compensating variation. In Figure 15 the individual consumer's increase in consumer's suirplus is equal to BG. This is true because BG Is the income the Individual could give up while purchas ing xg °£ ^he utility service at the marginal-cost price and still be as well off as under average-cost pricing. BG, it will be noticed, is less than MgM^, the compensating vari ation. BG could be equal to MgM^ if the slopes of 1-^ and Ig were equal at Xg output. This would be the case if the utility surface followed Marshall's assumption of the con stancy of the marginal utility of money. In any case, the Marshallian consumer's surplus for each Individual could never be greater than the compensating variation. 15 Con- -*-®For a more detailed examination of the relationship between the compensating variation and consumer's surplus, see A. M. Henderson, "Consumer's Surplus and the Compensat ing Variation," Review of Economic Studies, Vol. VIII (Feb ruary, 1941), pp. 117-121. sumers' surplus for all consumers together, therefore, could not exceed the sum of all compensating variations. The simplest way to determine If the consumers could make the payment would be to compare the Increased consum ers' surplus with the necessary payment. It is unfortunate 169 that no generalization can be made as to whether this sur plus gain is sufficient or not. The situation would depend on the relative elasticities of the average cost curve and the demand curve. It would be necessary to measure the areas and make direct comparison in each case. It must be remembered, furthermore, that the consumers’ surplus in the market demand curve does not tell how much each individual can be taxed. For that it is still necessary to consult the individual utility surfaces. One additional complication must be mentioned. It Is quite possible that the demand curve for the public utility service would shift after the tax is collected. If that did occur, the original calculation of the required subsidy would be incorrect. If consumers had utility surfaces simi lar to the one illustrated in Figure 15, there would be an income effect on the market demand curve. If, however, the indifference curves had the same slope at every quantity, i_.e_. on any ordinate erected from the abscissa, the demand curve would not shift. In other words, if the utility sur faces followed from Marshall's assumption of constancy of the marginal utility of money, the problem of the shift in the demand curve would not exist. For purposes of simpli city it would be best to assume the constancy of the mar ginal utility of money. However, in cases where this is not true quite a complication arises as to the proper amount of the subsidy. The complication would result from the circu 170 larity in which the collection of the tax to pay the subsidy would itself affect the size of the needed subsidy. 2. Method of Tax Payment The tax which collects the subsidy can be levied through excise taxes, income taxes, lump sum taxes or taxing away pure rent. Each of these will have a different econ omic effect. If an optimum pricing system Is the goal which is to be achieved, the particular tax which is levied to help achieve the goal should not in itself result in viola tion of that goal, i.e^. the way In which the tax is collec ted should not result In violation of the welfare principles. If an excise tax Is to be used to collect the necessary amount, It could be placed upon either the public utility service Itself or upon other commodities. If the excise were placed upon the public utility service, it would amount to average-cost pricing. The price plus the tax would even tually end up at the price which would be charged using average-cost pricing. The amount of the service which would be purchased would also be the same amount as under average- cost pricing. Collection of the subsidy in this manner would be self defeating. Of course, an excise could be placed on other commodi ties. This would result in violation of the marginal-cost pricing principle in these other areas of production. The price of the taxed commodities would be higher than marginal cost. The marginal rates of substitution between products 171 would not equal the marginal rates of transformation between products. The fifth marginal condition of maximum welfare would be violated. An Income tax, if used to finance the subsidy, would in effect place an excise on effort. It would disturb the •^■^Coase, "Marginal Cost Controversy," o]D. cit., p. 179. adjustment between work and leisure. The marginal rate of substitution for any input between Income for work and ren dering direct services to itself (leisure) would not be equal to the marginal rate of transformation between the in put's time spent at aiding production and the amount pro duced. The sixth marginal condition of maximum economic welfare would be violated, therefore. The third possibility is to apply a lump sum tax on the utility users. This tax would be in the nature of a tax on the individual as a consumer of utility services. The con sumer would be required to pay a lump sum tax not larger than the compensatory variation and a charge equal to mar ginal cost for every unit of the service purchased. This would be an ideal solution. There would be no Interference with prices of other products, there would be no redistri bution of incomes and effort would not be taxed. A. M. Hen derson is apparently the first to recognize this, stating, "If then a method can be devised for charging a fixed annual 172 sum to each user, adjusted so that no one has to pay more than his consumer's surplus, this will be the ideal method of charging. The application of this method would, of •'•'^Henderson, "The Pricing of Public Utility Undertak ings," o]D. cit. , p. 231. course, mean the application of a two-part tariff. This has been suggested by several writers, notably Coase.-*-8 Howev- ^8 Coase, "Marginal Cost Controversy," ojo. cit., pp. 173-174. er, Henderson pointed out that the initial payment should not exceed the gains to the consumer from the use of mar ginal-cost pricing. A tax on the consumers' gains from marginal-cost pric ing is, of course, a tax on consumers' rent. There is a possibility of taxing pure rent to obtain the necessary funds, as Hotelling pointed out in his well-known article.^ ■^Hotelling, oje. ci t. , p. 256. Henderson, however, points out that if there is a general policy of taxing away pure rent, such rents would have been taxed away previously for other purposes.There remains 2 8 Henderson, "The Pricing of Public Utility Undertak ings," ojo. ci t., p. 229. the possibility, nevertheless, that pricing of the utility service at marginal cost instead of at average cost would increase rent for some individuals. It would seem such rent could be taxed away to subsidize the utility. An example might be the increased rent to property surrounding a train terminal as a result of the increased use of train service because of the lower train fares. This follows from the logic of taxing the consumers according to the gains ac cruing to them from the lower price. Such taxing of rents, however, probably also Implies other things. In an economy which regularly taxed away pure rent, it would be necessary to argue that all such taxes should be used In some payment for the cause of the rent. This would be awkward in such cases as rents resulting from an Increased population.Or, in an economy where such rents were not regularly taxedaway it would be necessary to argue for the taxing away of all such re turns. Because of the difficulties connected with other taxes, the lump sum tax on utility users seems to be the most sat isfactory from the welfare viewpoint. The tax could be In the form of an annual installation, service or inspection charge if necessary. Such obscuring may be necessary be cause the application of the tax itself as a pure charge for being a consumer may result In the loss of the expected gains in the very process of attaining them because of the unhappiness with the tax. The variation in the tax as be- 174 between individuals may still cause difficulty in this re spect. This is a special case of the general problem of losing potential welfare gains because the method of a- chievement is so distasteful. G. Concisions In comparing average- and marginal-cost pricing as principles of public utility price setting, no writer has pointed out the dependence of the case for average-cost pricing upon a fundamental value judgment--consumers should pay the full costs of production. The value Judgment ap parently stems from the theoretical pure competition solu tion in which consumers do pay the full cost of production in the long-run. Competition, however, is not preferred for its own sake. Because price equalling marginal cost results in optimum allocation and because under perfect competition price does equal marginal cost, perfect competition is con sidered to be the ideal market situation. Where competition cannot prevail, it is not the competitive market and its results which are appealed to for guidance in achieving maximum economic welfare. Rather, reference must be mads to the principles of economic welfare which define that opti mum. These principles, as stated in the seven marginal con ditions of maximum economic welfare, make marginal-cost pricing the key guide to an economic optimum. Perfect com petition results in marginal-cost pricing. Therefore, per- 175 feet competition contributes to the ideal output under stated assumptions. If a market in which competition is im possible is to contribute to the achievement of maximum economic welfare, then price should be set equal to marginal cost. Evaluating average-cost pricing in the light of the welfare principles, it can be stated that under such pricing the output of the product is restricted. It would be better, from the view of achieving maximum economic welfare, for output to be expanded to the output at which price equals marginal cost, if the subsidy can be paid by the consumers. However, it should be pointed out that welfare econ omics can not argue with an ethical or value oriented view that each consumer should pay his proportion of the full costs of production as measured by average cost. Welfare economics is limited to one statement--if it is true that each consumer should pay an amount equal to average cost per unit, then maximum economic welfare can not be achieved if any areas in the economy are characterized by conditions in which perfect competition cannot exist, if natural mo nopoly conditions exist. Putting this in another way, the case for marginal-cost 21 pricing is not a dogmatic one. Rather, it presents a ^ C f . B. P. Beckwith, Mar gi nal - Co s t Price-Output Con trol, (Hew York: Columbia University Press, 1955"). This is essentially a polemic favoring marginal-cost pricing which neglects economic welfare theory. 4. 176 method of determining whether welfare is being maximized from on point of view. Whether this particular point of view, the economic, should or should not be taken or prefer- ed is not within the power of welfare economics to decide. Thus, it may be thought desirable on some humanitarian grounds to price certain services or products below marginal cost to make these services or products available to low in come groups at a low price. The principles of welfare econ omics can be used to point out the effects on real income distribution and on allocation of resources. These prin ciples can not be used to judge whether this low price poli cy should or should not be put into effect. They can only furnish a basis for judgment as to whether the losses in welfare looked at from the economic point of view are offset or compensated for by the gains in welfare looked at from another, e_.g,. humanitarian, point of view. Summing up this comparison between average- and mar ginal-cost pricing, average-cost pricing is not consistent with the maximization of economic welfare. Arguments for average-cost pricing which claim it results in maximum econ omic welfare are incorrect. If it is economic welfare which is to be maximized, then marginal-cost pricing furnishes the simple rule by which this may be achieved. If the value judgment that each consumer should pay his proportional share of the total costs of output is deemed correct, then average-cost pricing should be used. One or the other of 177 these pricing principles may be more convenient on the basis of practical considerations. These practical considerations will be given attention in a later chapter. CHAPTER VTI THE'SHORT-RUN AND OTHER PROBLEMS OP PUBLIC UTILITY PRICING The previous chapter contained a discussion of long-run pricing. Pricing practices with a stable long-run demand curve and given long-run cost conditions in the firm were considered. The assumption of such stability made for rela tive simplicity of analysis as compared with the analysis of other problems. Other problems are short-run pricing, pric ing in new and expanding industries and pricing in dying and contracting industries. Given a fixed plant and fluctuations in demand, a prob lem of some complexity is encountered. This problem involv es pricing in the face of cyclical, seasonal and random changes in demand. The implications of average- and mar ginal-cost standards of pricing under such conditions will be given attention. In other words, short-run public utili ty price setting will be one subject of analysis In this chapter. More complex still is the problem of pricing in new and expanding Industries. Consideration must be given to the basis for judging when a new Industry should begin opera tions and to the nature of pricing standards in such a situ ation. The desirable rate of expansion of the plant and the means of insuring such expansion must also be examined. Dying and contracting industries pose a further prob lem. Here the question arises as to the pricing practices 178 to be followed so as to contribute to maximum economic wel fare. The problem of the proper time to cease production also must be analyzed in such a discussion. In this chapter it will be assumed that demand and cost conditions are known and measurable. This, of course, as sumes away any conditions of uncertainty. It also means that certain practical considerations will be Ignored. These practical considerations will be examined in Part III of this study. A. Short-Run Public Utility Pricing A certain sized plant may be constructed by a public utility firm to satisfy the long-run demand of public util ity consumers. But, with this given sized plant, demand may be subject to fluctuations in the short-run. These fluctu ations may be of a cyclical, seasonal or random nature. The seasonal fluctuations may be recurring changes with actual seasons, through a month, a week or even a day. The ques tion of pricing practices with the existence of such fluc tuations must be considered so that the results of pricing with either marginal- or average-cost pricing bases can be seen. 1. Marginal-Cost Pricing in the Short-Run Figure 17 illustrates the conditions which a public utility firm and commission might face in the short-run. With long-run demand curve Intersecting the long-run mar ginal cost curve LMC, a plant would be constructed, assuming 180 Figure 17 Short-Run Marginal-Cost Pricing s*c LAC LMC Ou T P U T 0 181 marginal-cost pricing, which had short-run cost curves SAC and SMC, average and marginal respectively. The long-run price of OB would exist. In any short period of time, how ever, demand would be subject to change. Such short-run shifts in demand away from the long-run demand could be rep resented by demand curves D2 and D3 . Demand changes such as those represented by demand curves Dg and D3 may have any one of several sources. They may result from cycles with Dg representing the top of the cycle and Dg the bottom. Such changes also may result from any of the seasonals with D3 and Dg representing the great est and smallest demands respectively during the year, month, week or day. Another possible source of shifting de mand is the random or episodic fluctuation. The resulting pricing problem is one of using variable inputs optimally with some inputs, i_.e_. plant, fixed in amount. Several pos sible policies could be followed in the short-run. These possibilities should be examined. Strict adherence to marginal-cost pricing would result in price changing as demand changed. The resulting output would always be the output at which the average revenue equaled marginal cost. At the lowest demand the price would be OA while with the greatest demand the price would be QC. As price varied in this manner one marginal condition of maximum economic welfare could not be met. The marginal rate of substitution between the capital input (plant) and 182 other' inputs of the public utility firm would not be equal to these same marginal rates of substitution for other firms in the economy. This is just another way of saying that this one particular input could not be adjusted optimally. The constraint is placed upon any system by the very defi nition of the short-run. Marginal-cost pricing and its con sequent output would represent maximization of economic wel fare given the constraint. A question might arise as to covering variable costs at the lowest level of demand. Assuming Dg as the lowest de mand at which price OA would be charged according to the marginal-cost pricing principle, it would be quite possible for the price to be less than average variable cost. The short-run problem is just a variation of the long-run prob lem. Under marginal-cost pricing in the long-run, the price does not cover total costs. But total costs are variable costs in the long-run. The relationship between price and average variable cost in the short-run should not be the criterion for shutting down any more than the relationship between price and total costs is the criterion for continued operation in the long-run. For shutting down in the short-run, the consumersT sur plus should be consulted. If the sum representing the con sumers' surplus plus the amount paid for the service pur chased is equal to or greater than the total variable costs, the plant should stay in operation. In other words, if the 183 total amount consumers would be willing to pay rather than do without the service is greater than the costs of the needed variable inputs, the plant should continue to operate. Each consumer should be taxed the amotint of his consumer's surplus to pay for continuing operation. Such a situation, however, must be temporary. The persistence of the problem over any length of time would lead to considerations dealt with in the discussion of dying and contracting industries. The much discussed zero price would prevail at such low levels of demand. If the bottom of the marginal cost curve were along the abscissa, marginal cost would be zero.-'- With ■^A negative marginal cost is an impossibility. the demand curve intersecting the marginal cost curve at such an output, marginal-cost pricing would result in a zero price. This occurs when the low level of demand results in the existence of a great deal of unused capacity. Indivisi bility of Inputs is also necessary. Such indivisibility im plies the ability of the very same inputs being able to pro duce several different outputs, or a range of outputs. The inability to transform the product or service into anything g else is another interpretation of zero marginal cost. Such 2 Little, Welfare Economics, op. cit., p. 189. a view means that zero marginal cost is characterized by in- 184 puts being committed to production in such a way that they cannot be varied. This explains the popularity of examples such as bridges in discussions of zero marginal cost. Considerable price fluctuation might be taking place if marginal-cost pricing were followed rigidly. The pricing system, furthermore, would be quite complex. The price at any given time would depend on the time of the day, day of the week, season of the year and the portion of whatever cycle happens to be operating. Under such a pricing system little regularity of pricing would exist. Consumers, for example, could not depend on the exact price prevailing on every Tuesday at four o'clock. Specific practical problems which would develop from such a situation would be very great. In general, the uncertainty added to consumers as a result of fluctuating prices could be quite disturbing. The consumer would have difficulty In planning expenditures be cause of the uncertainty over the price at any future time. The hardship Involved in determining or estimating the price which is going to prevail might be so heavy as to more than overcome the expected gains from marginal-cost pricing. A question might even arise as to the ability to publish prices in view of the lack of predictability of the time series movements which have been mentioned. Even assuming some predictability the advance publication of prices would be subject to the difficulties Involved in a prediction's in influencing its own data. 185 An alternative course of action to be followed would be to keep output constant at the long-run optimal output. The prices would fluctuate violently if such a practice were followed. At times the price might even become negative. Under such a practice the marginal rates of transformation between inputs and the product would equal the respective ratios of input prices. However, the marginal rates of sub stitution between products for consumers would not equal the marginal rates of transformation between these products. This would be true because the marginal rates of transfor mation between the products would equal the ratios of the products1 marginal costs while the marginal rates of substi tution would be equal to the price ratios, with prices not equal to marginal costs. Prom an economic welfare stand point it would be preferable to set the prices at marginal costs at all times. Another possible policy which would violate the econ omic welfare principles and might even bring about serious difficulties would be to maintain price at the long-run mar ginal cost. Because of the dr as ticoutput changes all mar ginal conditions relating to marginal rates of transfor mation would be violated. Serious difficulty might be en countered when demand was great. This difficulty would take the form of excessive strain on equipment. In Figure 17, if the long-run price of OB prevailed at all times, BG output would be purchased at peak demand. BG output would have a 186 marginal cost represented by point H (OB plus GH). With a large increase in demand the marginal cost might even ap proach Infinity. A breakdown would result. Examples might be the collapse of a bridge under excessive weight or the breakdown of electrical generation equipment because of ex cessive load. To sum up, ideally, from an economic welfare viewpoint, the price would vary as demand changed, price always equal ling marginal cost. Such a policy might be quite disturb ing, however, particularly if there were large and frequent demand changes. The alternatives, maintaining the long-run output or maintaining the long-run price, would both violate the marginal conditions of economic welfare and might be even more disturbing. 2. Average-Gost Pricing in the Short-Run For some reason there is an absence of discussions as to the results of using average-cost pricing as a basis for setting public utility prices in the short-run. A consider ation of the problems of such pricing would fill an existing gap. Figure 18 represents the situation which would face the firm and the commission in the short-run using average- cost pricing as the pricing principle. With a long-run de mand represented by demand curve the plant having short- run average and marginal cost curves SAG and SMC respective ly would be constructed. A price of OB and an output of BE would prevail over the long-run. However, short-run flue- 'Se 188 tuatlons In demand would occur. Such short-run fluctuations having the same origins mentioned in the discussion of mar ginal-cost pricing are represented by demand curves Dg and Do- Rigid adherence to average-cost pricing would result in OG price, with CP output, when minimum demand Dg existed. At peak demand D- the price would be set at OA, with AD out put. With any amount of the expected curvature in the aver age cost curve the result is somewhat unusual. If demand decreases, price goes up. If demand increases, price goes down until at minimum average cost it begins to rise. Of course, if the average cost curve were horizontal through out, the price would not vary. Because of the existence of fixed costs in the short-run, a completely horizontal aver age cost curve is quite improbable. A proponent of average- cost pricing, therefore, must accept an increase in price when demand decreases. With a violent cyclical downturn, demand might fall by a great amount. If there were considerable slope to the average cost curve, the price might be quite high in a deep depression. It would be possible for the decreased demand to result in the demand curve not intersecting the average cost curve at any point. Following the basic rule of aver age-cost pricing, 3^.e_. consumers should pay the full costs of production, price discrimination would be required. The public utility would sub-divide consumers into separate mar 189 kets so no trading between markets could take place. A mon- rz opoly price would then be charged in each market.. If the ®The rule would be that the marginal revenue in each market v/otild equal the marginal cost at the total output. See, A. G. Pigou, The Economics of Welfare, (London: Mac millan Sc Co., 192071 p. 267, n. 1; Joan Robinson, Economics of Imperfec t Competitlon, op. ci t., p. 181. total returns from maximum possible discrimination are in sufficient to meet the full costs, then presumably the firm would close down in the short-run* However, a deviation from the general rule of meeting total costs coiild be per mitted. Such a deviation would require the meeting of total variable costs only, following conventional theorizing on short-run problems. Whether prices would be subject to greater fluctuation under marginal-cost pricing than under average-cost pricing would depend on the relative slopes of the marginal and average cost curves. If the cost curves were of the type commonly presented, e_.£. Figures 17 and 18, marginal-cost pricing would result in greater price fluctuations. But a short-r\.in linear total cost function would have constant marginal cost and decreasing average cost.* With such a ^This is a type of short-run total cost function found in empirical studies. See, U. S. Steel Corporations, TWEC Papers, (Hew York: U. S. Steel Corp., 1940). 190 short-run cost pattern, marginal-cost pricing would result in a stable price while average-cost pricing would result in a fluctuating price. Any criticism applying to marginal- cost pricing on grounds of instability of price also would apply to average-cost pricing. The only doubt is as to the degree, 1 _»e_. which is more subject to such criticism because of the amount of fluctuation. Two other possible policies can be followed using aver age-cost pricing as the basic standard. The long-run output can be maintained at all times, i_.e_. the output at which the long-run demand curve intersects the long-run average cost. Following such a policy, the price would move in the same direction as demand. The amount of price fluctuation would depend upon the amount of the demand change and the price elasticity of demand. Price averaging out to equal the long-run average cost would be expected. This solution can be criticized on the basis of the great price fluctuations. Price might be negative at times as with the similar solu tion using long-run marginal cost as the base. The possible very large profits or losses might lead to difficulties mentioned in the third alternative which follows. Keeping price equal to the long-run average cost, i_.e. where the long-run demand curve intersects long-run average cost is the other possibility. With any negative slope to the short-run average cost curve, this would result in los ses with low demand and possible profits with high demand. 191 With a U shaped short-run average-cost curve extremely high demands also might lead to losses. With any protracted depression, some uneasiness about sustained losses is almost sure to result. Pressure would probably arise to raise price to equal the short-run average cost. Similarly, if any large profits accrued during any protracted cyclical peaks, there might be public clamor for price decreases. If average-cost pricing is to be used as a price base, maintaining prf.ce equal to the long-run average cost has some advantages. The prf ce would remain stable. Consumers could plan expenditures with more certainty. Output would not decrease as much as when the short-run average-cost pricing formula was used during periods of low demand. How ever, if the demand was quite elastic and with a large in crease in demand cyclically, a heavy strain might be put on productive capacity. Dangers of breakdown similar to those mentioned in the discussion of marginal-cost pricing would exist. 3. Summary Objections to marginal-cost pricing on the ground that fluctuating prices upset consumers furnish no support for average-cost pricing. Price also would fluctuate with average-cost pricing. The possibility exists of even great er fluctuations with average-cost pricing. Of course, there is the possibility of keeping price stable at the long-run average price. In such case, however, there must be a wil 192 lingness to risk overloading and possible breakdown. It would seem that any possible pricing policy requires price fluctuations as demand changes. Most markets have changes in prices, except oligopolistic markets in which fear of starting price wars exists. Consumers must adjust and make plans on the basis-" of possible price changes In most of their purchases. Price changes or embarrassment from over supply or excess demand are the usual conditions where changes occur In demand or supply conditions. Such a result in the case of public utility prices should be no great sur prise . The big question with respect to price changes is the possibility of chaotic conditions resulting from the magni tude of the changes. This is, of course, a matter of empir ical fact. It depends on the actual amount of change in demand in each particular instance. Generalized judgments based on intuition are not much help. Some idea of the amount of change In price probably could be obtained by in vestigating data from Individual public utility firms. De mand fluctuations might be so regular as to make possible a simple pattern of price changes. '/'/here this does occur un due uncertainty from price changes would not result. B. New and Expanding Industries Two problems of considerable interest are those of put ting productive services into operation and the handling of expanding industries. The first of these problems deals 193 with the standards applicable in judging when it is worth while to construct productive equipment whUch supplies some product or service which is demanded. The second deals with the problem of price setting if the public utility has a secular increase in demand. These problems will be consid ered from the view of both marginal-cost pricing and aver age-cost pricing. In the case of new industries, a criterion as to when a new plant should be biailt is the point in question. If average-cost pricing is used as the basic principle, then a plant should be constructed when the consumers are able to pay the full costs of the output with the price set at aver age cost. In other words, when the demand curve Intersects the long-run average cost curve, the plant which has its short-run average cost curve tangent to the long-run average cost curve at the intersection should be constructed. If the average-cost standard is not held with great tenacity, then the criterion can be modified. A plant can be con structed when consumers are able to pay with discrimination existing. As soon as the total costs can be exacted from the consumers by subdividing the market into several markets and using discriminatory pricing, the plant should be con- 5 strueted. ^D. H. Wallace proposed such a solution though he dis cussed a situation in which the long-run average cost curve was U shaped. It Is interesting to note that Wallace seemed to put emphasis on the role of marginal cost to attain best 194 utilization as early as 1934. See, D. H. Wallace, ’'Joint and Overhead Cost and Railway Rate Policy," Quarterly Jour nal of Sconomics, Vol. XLVTII (August, 1934), pp. 583-619. Pigou suggests a similar policy but points oxit the possibil ity of a subsidy. See A. G. Pigou, ojc. cit., pp. 273-279. The solution to the problem of the construction of a new plant is not very different if marginal-cost pricing is used as the basic principle. Under such conditions, a plant Is to be constructed if the sum of the individual consumer surpluses which result plus the payments made for the pro duct are equal to or greater than the total costs.® Hotel- ®Hendarson, "The Pricing of Public Utility Undertak ings," ojc. cit., p. 231. ling generalized this view to Include surpluses which would arise to all producers and consumers of all commodities who would be affected by the construction of the plant. The ^Hotelling, o]c. ci t., p. 247. Hotelling version would be more accurate. The price which would be charged would be determined by the intersection of the demand curve and the long-run marginal cost curve. This intersection also would determine the size of plant. The deficiency in income would be met by lump sum taxes on the surpluses accruing. In other words, It would be necessary to investigate demand and supply conditions not only of the public utility service but also of all direct and indirect 195 effects upon consumption and production of all other pro ducts . This criterion, it should be noted, does not justify the prediction of future demand and the setting of promo tional rates. The surpluses arising are those related to existing demand and supply conditions. Where development and increased demand in the future are expected, conditions are too dynamic to receive much help from the standards based upon the static assumptions of welfare economics. The summing up of the expected surpluses of the future might be suggested. Difficulty arises, however, from guesses or in tuitions being highly variable, differing drastically from person to person. Ex ante, one man’s guess is about as good as another's.® ^This discussion touches on the role of the entrepre neur, an area in which authors use adjectives such as crea tive, energizing, shaping and bursting. As an example, see, J. A. Schumpeter, "The Creative Response in Economic His tory, M ojc. cit. The problem of a public utility which faces a growing demand also should be considered in this section. Basical ly the problem seems quite simple. As demand increases, the firm can also increase and the price would change according ly. Some difficulties arise, however, using either average- or marginal-cost pricing. If average-cost pricing were followed, a long-run in crease in demand would put pressure on existing plant. This 196 increase would become evident not only through pressure on existing plant but also because of the effect on price. If a short-run policy of pricing at the short-run average cost were followed, the eventual result from any sizeable in crease in demand would be a rather high price. This would occur because of average operations at an output greater than minimum average cost. Many variable inputs would be added bzzt, because of the limited plant facilities, output would not be increasing very rapidly. A. simple solution if the price were quite high would be to determine whether, at the existing price and with a larger plant, payment could be made for the new plant plus the undepreciated portion of the old plant minus scrap value. If it can, the new plant should be built. If it cannot, operations should stay at the old level until more of the old plant is depreciated, i_.e_. , the undepreciated portion is decreased. The scale of the new plant should be that capable of producing the new long-run output most cheaply. When the old plant deprecia tion can be paid off, price can drop to the long-run average cost. Figure 19 illustrates this solution. With an original long-run demand D, a plant having short-run average cost curve SAG would be built. If long-run demand moved to D 1 and short-run pricing set price where demand equaled average cost, the long-run price would move to GA. If the demand Increase were greater, the price would be greater. If less, 197 Figure 19 Average-Cost Criterion for Increased Investment SAC S^C' 198 price would be lower. To meet the new long-run demand a plant ’with short-run average cost curve SAC1 would be con structed. With the new plant and a price of OA, a per unit return BC in excess of the average cost of output would be received. If BC could cover the undepreciated old plant minus scrap value, the new plant should be constructed. When the old plant is paid off, price could be lowered to OM which would set output at the intersection of long-run aver age cost and demand. If marginal-cost pricing is used as the pricing prin ciple, the price will rise as a result of the increased de mand. The new long-run price will move toward the inter section of the short-run marginal cost and the demand curve. The plant will feel considerable pressure from the high rate of output. Profits might be quite high for a time. To build the new plant, the increased consumers' surplus which would result from building the new plant must be considered. If the increased consumers' surplus is equal to or greater than the new deficit which must be paid plus paying off the undepreciated portion of the old plant minus scrap value, the new plant should be constructed. If it is not, the old plant should continue In operation, thus reducing the unde preciated amount, until the consumers' surplus is sufficient to cover both the deficit and payment on the old plant. The new subsidy which is collected by lump sum taxation would be equal to the deficit between operating costs and revenues 199 plus depreciation of the old plant. When the old plant is depreciated, the lump sum tax would be lowered to cover the deficit between operating costs and total revenues. One big difficulty arises in the discussion of changes In the size of the plant. Public utility management may resist increases in plant size. Particularly is this true If marginal-cost pricing is used. With demand remaining high and prices set to equal short-run marginal cost, re turns to the firm would be quite high. Management might do a great deal to conceal these high returns, including incur ring unnecessary expenses. The management would see only the competitive return upon construction of the new plant whereas the smaller plant would give something in the nature of a monopoly return. As Troxel points out, it is difficult Q to overcome the tendency of a monopoly to underinvest. ^Emery Troxel, "Limitations of the Incremental Cost Pattern of Pricing," Journal of Land and Public Utility Economics, Vol. IXX (February, 1943J, pp. 28-39. A variation on the problem of new and expanding indus tries is the problem of technological improvement which re sults in new and lower cost structures. The rule to be fol lowed is similar to the rule for expanding industries. The gain from putting the new equipment Into operation should cover the cost of using new equipment plus writing off the undepreciated old equipment.^ The big problem here is aim- 200 For a good textbook discussion of* the problem, see Emery Troxel, Economics of Public Utilities, (New York: Rinehart & Co.^ 1947), pp. 359-363. ilar to the problem of expansion. It would be difficult to convince a privately managed public utility firm that any advantages are available from the Installation of the new equipment. After all, the rate of return would remain the same. There would be no increase in the dividend rate or the returns to management. G. Dying and Contracting Industries In any economy, development of new products results in decreases in the demand for old products. In the area dis cussed in this study, the problem of a public utility facing large decreases in demand arises. The industry Is contract ing or dying. The regulatory problem in such a situation would not be so much the prevention of monopoly returns as it would be to see that sufficient service Is provided to insure economic benefits. If average-cost pricing were used, the firm could be permitted to use monopoly pricing. If the price charged were greater than average variable costs, the firm would stay in operation. When price no longer covered average variable costs, the firm would shut down its operations. This follows the general rule of average-cost pricing that consumers should pay for the costs of production. As the full costs cannot be met, the old Investment is forgotten. 201 If revenues more than covered variable costs, investors would recotip some of their sunk funds. When this is no longer possible, the firm would operate so long as the vari able costs were covered. The firm also could be permitted to use discriminatory pricing so long as the total revenue did not exceed total costs. With such discriminatory pric ing, the firm would be able to stay in operation for a greater length of time. The loss on Investment would not be so great. If marginal-cost pricing is to be followed, the price should be set at the short-run marginal cost. If the reve nue plus the consumers' surplus is greater than variable costs, the firm should stay in operation. The consumers' surplus should be taxed away by lump sum taxation to pay the subsidy to the public utility firm. The entire consumers' surplus should be taxed away so long as the total of re ceipts from sale plus the subsidy from the taxes Is equal to or less than total costs. When variable costs are not met by receipts from sale plus consumers' surplus, the firm should close down. It may be possible, of course, for a much smaller scale of plant to be built which will pay Its way. In this case the standards for the construction of a brand new plant should be used. For an investor to build such a plant when the industry is decaying and regulation results in a compet itive return at best is rather difficult to conceive. How- 202 ever, it is possible for the product or service to have be come so insignificant that consumers would not be able to pay for the costs of regulation. In such case monopoly could exist unregulated. The product or service would be supplied to those who are willing to pay and the monopolist would be gaining returns. By the welfare standards this would be better than not having output. D* Conclusions Marginal-cost pricing furnishes an adequate welfare criterion for short-run pricing. The pricing mechanism would allocate resources so as to contribute to maximum economic welfare. Difficulties such as breakdown or under utilization which might occur if some inflexible pricing practice were used, would be prevented. However, if de mand fluctuations were at all frequent, large and irregular, a considerable amount of disorganization might result. Average-cost pricing offers the same possible difficulty. With average-cost pricing, however, resources are not allo cated so as to maximize welfare. It does result in each consumer paying for the total costs of his purchases (defin ed as average cost times amount), if this is insisted upon. In the consideration of the construction of new plant or expansion of plant the limitations of the static assump tions become evident. By laying down rules for measurement of consumers' surpluses much of the spontaneity and imagin ation of investment in unregulated markets is lost. Instead 203 of guesses being made and acted upon with the attendant pos sible gains or losses, perfect knowledge is required. Under the rules set forth in this study there would be no incen tive for the entrepreneur to take risks or to innovate. There would be only the danger of losses on the one hand and the regulating away of any gains on the other. This seems to be one of the penalties of the application of rules for investment, particularly if these rules are based on the static assumptions of given utility and production functions. PART THREE THEORY AND PRACTICE IN PUBLIC UTILITY REGULATION In the preceding discussion certain practical consider ations were Ignored. Because of the implications of the material considered In this study, a discussion of actual practice in public utility regulation would seem desirable. The main purpose of this study was to consider the theory in a rigorous manner. However, the relationship of the theory to practice and the possible practical use of the theory are too important to be neglected. Actual public utility regulation will be considered from two viewpoints. These viewpoints will be the apparent goals of regulation as evidenced by legislation regarding regulation of transportation and the general standards of pricing applied in public utility regulation. The actual goals will be compared with the economic standard of optimum utilization of resources. The pricing standards will be In vestigated in an attempt to determine the standards which are used in general public utility price setting. Practical use of the theoretical tools considered in this study will be the other topic to be examined. Possible difficulties in the applicability of marginal-cost or aver age-cost pricing in actual public utility pricing will be the point of greatest concern. The goals of actual regulation, pricing standards used in practice and the applicability of the tools of economic 204 •205 theory having been considered, some conclusions can be drawn and some suggestions as to feasible policy can be made. These conclusions and suggestions should make possible some evaluation of actual public utility regulation and also shed light on the place of economic theory in public utility regulati on. CHAPTER VIII FEDERAL TRANSPORTATION REGULATION POLICY A study of the regulation of the several regulated in dustries in 48 States plus Federal regulation is too impos ing a task to be attempted within the limitations of this study. A compromise may be reached by considering Federal regulation of transportation by the Interstate Commerce Com mission. Such a limitation will not lead to deceptive re sults if care is taken in generalizing from the one in stance. The discussion to follow will consist of an inves tigation of: (1) goals of regulation as indicated by Federal legislation dealing with regulation of transportation; and (2) standards for pricing in Federal regulation. A. The Goals of Federal Transportation Regtilation Discussion of the definition of public utilities in Chapter V contrasted the legal definition of public utility with a proposed economic definition. The legal standard, as pointed out In Chapter V, is quite loose at present. It is difficult, therefore, to understand what is being attempted through public utility regulation. In trying to ascertain the goals of public utility regulation it Is necessary to investigate actual regulation. Within the constitutional limitations as interpreted by the courts, the legislature sets the standards of regulatory action. A consideration of the pertinent legislation should give an insight into the goals of public utility regulation. 206 207 Before proceeding on this analysis, one point should be emphasized. Only goals which are included in the legisla tion will be considered. Statements in hearings or debates as to opinions of individuals regarding the goals or aims of the legislation will not be given consideration. Inferences will be drawn only from the language of the legislation. Furthermore, the concern of the study is on general policy rather than individual regulatory provisions. The individ ual provisions will be entertained only when they are espe cially pertinent to general policy. For purposes of this study the general statements of policy or purpose, when in cluded in the legislation, is more significant than the in dividual provisions. Federal legislation dealing with regulation of railways started with the Interstate Commerce Act of 1887.^- The act •*•24 Statutes at Large, pp. 379-387. had no general statement as to goals or general policy. The title stated the act to be one "to regulate Commerce". To find the goals to be achieved, the individual provisions must be considered. A provision in Section 1 states that, "All charges.... shall be reasonable and just; and every un just and unreasonable charge....is prohibited and declared to be unlawful." Justice and reasonableness of railway charges were undefined. In juxtaposition to provisions 208 which followed in the act, this principle of rate-making could be interpreted as indicating opposition to exception ally high monopoly rates. Charges were desired which did not exact an exorbitant monopoly tribute from rail users. Section 2, 3, 4, and 5 seemed to be aimed at eliminat ing monopolistic elements in railwajr charges, and section 6 might be interpreted as being similarly directed. Section 2 prohibited special rates or rebates. This prohibition was apparently directed at undue discrimination, which is char acteristic of monopoly. Sections 3 and 4 seemed similarly directed. Section 3 prohibited undue preference to any par ti cular persons, localities, firms, or traffic. Section 4 contained a prohibition against a particular type of dis crimination, namely charging more, in the aggregate, for a short haxal than for a long haul of like kind, in the same direction and similar circumstances, the shorter being in cluded within the longer distance. Exceptions to the pro hibition could be granted by the Commission. This is the well known long- and short-haul clause. Section 5 prohib ited pooling of earnings in the apparent attempt to elimi nate collusion between rival carriers which were expected, presumably, to compete. Section 6, which apparently was aimed at eliminating the secretiveness necessary for certain types of discrimination, provided for the publication and display of rates, the filing of rates with the Interstate Commerce Commission and advance notice of 10 days for rate 209 advances. The remainder of idle act dealt with penalties, organizational provisions, regulatory procedures and other such details which do not lend themselves readily to inter pretation as to the goals of the legislation, or at least would not change the interpretation which can be made from the sections already summarized. Before presenting an interpretation of this basic act, it would be well to mention pertinent provisions of a series 2 of amending acts. The Elkins Act of 1903 added one fea- ^32 Statutes at Large, pp. 847-849. ture which is pertinent to this study. With the passage of the Elkins Act it became unlawful Mto offer, grant, or give or to solicit, accept or receive any rebate, conces sion, or discrimination". This provision seems to have been aimed principally at those who possessed monopsonistic power and who used this power to gain special consideration from the rail carriers. Discrimination and rebating were already covered in the Act of 1887. 3 The Hepburn Act of 1906 which, because it strengthened 334 Statutes at Large, part I, pp. 584-595. the power of the Commission, was a major amending act had, for the purpose of this study, two fairly important provi sions. Power to set, upon complaint, the just and reason- 210 able rate or the maximum rate which could be charged was given to the Commission. This provision, when considered from the viewpoint of the indications of general policy, represented a strengthening of the earlier provision prohib iting unjust and unreasonable rates. The other important provision was the well known commodities clause which stated that: It shall be unlawful for any railroad company to transport.... any article or commodity, other than timber and the manufactured products thereof, manu factured, mined, or produced by It, or under its authority, or which it may own in whole, or in part, or in which it may have any Interest direct or indirect except such articles or commodities as may be necessary and intended for its use in the conduct of its business as a common carrier The clause was consistent with the general policy em bodied in the series of acts under consideration. It at tempted to reduce or eliminate monopolistic discrimination favoring railway owned commodities, particularly coal. The Mann-Elkins Act of 1910^ has importance in this ^36 Statutes at Large, part I, pp. 539-557. study principally in its granting of increased power to the Commission, especially In restrengthening the long- and short-haul clause. However, one feature should be mentioned specifically, namely prohibiting an increase in railroad rates following a decrease to meet water competition. SUch a prohibition obviously was aimed at the elimination of com petition by attrition against less powerful rival modes of 211 transportation by large and powerful railroads. While other transportation legislation enacted during the 1887-1920 period contains no provisions pertinent to this study, the Clayton Act of 19145 has some significance. ^38 Statutes at Large, part I, pp. 730-740. The other principal legislatTon was the Valuation Act of 1913, 37 Statutes at Large, part I, pp. 701-703, which dealt with the calculation of rail property valuation, and the Panama Canal Act of 1912, 37 Statutes at Large, part I, pp. 560-569, which will be considered Tn the discussion of water carrier regulation. The Clayton Act barred the acquisition, in general, by one corporation of the securities of other corporations where such acquisition would lessen competition between the rele vant corporations or restrain commerce. Common carriers were exempted from this provision only in the extension of their own lines or where branch or feeder lines were acquir ed. Even in these situations the necessary condition was the original absence of competition. The several acts considered to this point constitute a unit when considered from the viewpoint taken in this analy sis. It is now possible to summarize and interpret these acts to point up the policies or goals which apparently were basic to the Interstate Commerce Act and its amendments. While no explicit statement of policy was made In these acts, the basic aim seemed to be the tempering of monopolis tic results, especially the possible unduly discriminatory 212 features of monopoly conditions. Hie means of achieving this goal was the enforcement of competition. This is indi cated not so much by the pricing provisions dealing with just and reasonable rates. Rather, the prohibition of pool ing, long-and short-haul clause discouragement of destruc tive competition and the barriers to acquisition of control of other firms in the Clayton Act were the means expected to enforce competition. Such a conclusion is no revelation. The same conclusion was stated many years ago by E. J. James and P. H. Dixon.® Both writers pointed out the low proba- ®E. J. James, "Federal Regulation of Railways," Publi cations of the Amerlcan Economic Association, Vol. II (July, 1887), pp"I 246-295; F. H. Dixon, "The Interstate Com merce Act as Amended," Quarterly Journal of Economics, Vol. XXI (November, 1906), pp. 22-51. bility of success of attempted enforcement of competition in the railroad industry. Dixon stated, "The theory clung to so tenaciously by the people at large, and given weighty sanction by the decisions of our highest court, that compe tition can be relied upon to give shippers reasonable rates Is utterly Impracticable when applied to the railroad Indus- ry try." If it Is assumed that decreasing costs to scale ^Ibid., p. 50. exist in the railway industry, then the conclusion is in agreement with the theory In Part One of this study. 213 Legislation pertaining to r a i l w a y regulation following 1920 cannot be considered as a unit as was the previous leg islation. The provisions Included in the Transportation Act of 1920, The Hoah-Smith Resolution of 1925, The Emergency Transportation Act of 1933 will be considered in that order. Because the Transportation Act of 1940 was concerned with motor carriers and water carriers as well as railways, it will be examined after motor and water carrier legislation i s summarized. The Transportation Act of 1920 represents a definite S41 Statutes at Large, part I, pp. 456-499. turning point with respect to the objectives of railway regulation. This redirection revolved on two pivotal pro visions of the act. First, the rule of rate-making was modified. The modification, along with the recapture clause, represented an attempt to establish the credit of the railroads without unduly burdening railway users.® Sec- ®E. J. Rich, "The Transportation Act of 1920," Amerlcan Economlc Review, Vol. X (September, 1920), p. 527. ond, provisions with respect to consolidations represented at least a partial movement away from the enforcement of competition.^® 214 10T. G. Bigham and M. J. Roberts, Transportation, Principles and Problems, (New York: McGraw-Hill Book Co., 1952), pp. 239. By the rule of rate-making established by the Act of 1920 the Interstate Commerce Commission was to initiate, modify, establish or adjust such rates so that carriers as a whole(or as a whole in each of such rate groups or territories as the Commis sion may from time to time designate) will, under honest, efficient and economical management and reasonable expenditures for maintenance of way, structures and equipment, earn an aggregate annual net railway operating Income equal, as nearly as may be, to a fair return upon the aggregate value of the railway property of such carriers held for and used in the service of transportation. As the earning of a fair return on the value of proper ty by carriers as a whole, or in groups, does not ensure such fair return for each rail carrier, Congress added a provision known as the recapture clause to help the weaker railways. Under the clause any carrier earning over a 6% return on the value of its property would retain half of the excess in a reserve fund. Ike remainder of the excess over Q% would be turned over to the Interstate Commerce Commis sion. The funds thus recaptured by the Commission would be used for lending to weak roads and for the purchase of equipment to be leased to carriers providing the rentals returned 6% on the cost of the equipment. Taken together, the rule of rate-making and the re capture clause indicate a reorientation of policy away from the railway user toward the railways themselves. Particu 215 larly, the reorientation took the form of concern for the financial viability of the carriers. Such concern was not in great evidence in previous legislation. The other principal provision of the Transportation Act of 1920 related to consolidations. The interstate Commerce Commission was required to set up a plan for the consolida tion of the railways of the United States into a limited number of systems. Under the plan, existing routes were to be maintained wherever practicable. The Commission also was expected to arrange these several systems so the cost of transportation would be the same in each system, "so far as practicable". Uniform rates then could be charged and "substantially the same rate of return upon the value of their respective properties" would be earned by each of the railroads. Ironically, the plan was required to preserve competition among railways "as fully as possible". Any pro posed consolidation of rail carriers required conformity with the Commission's plan. Acquisition of control of one carrier by another by stock purchase or lease, with no fusion of identity occurring, required evidence of promotion of the public Interest. A fairly abrupt transition in the policy goals of rail way transportation regulation was evidenced in the foregoing provisions of the Act of 1920. Rate-making was redirected from concentration on the existence of just and reasonable rates to consideration of the financial needs of the car- 216 riers. The attitude toward combination as represented by the provisions with respect to permissive acquisition of control, though limited by the qualification as to preserva tion of competition, seemed to have undergone some change through the years. Experience with regulation and the changing of conditions led to emphasis upon different prob lems and resulted in changed views as to the goals of regulation. In 1925 Congress passed an interesting joint resolu tion, the Hoch-Smith Resolution.^ This resolution is in- -'--*-43 Statutes at Large, part I, pp. 801-802. teres ting because it is the first explicit statement of basic policy goals to be pursued by the Commission in set ting railway rates. It is also of interest because of the very specific type of promotion which was considered desir able . The Hoch-Smith Resolution contained three paragraphs, each containing one provision. First, the Interstate Com merce Commission in adjusting freight rates was required to follow the "true policy" of rate-making which was to consid er conditions prevailing in the industries concerned, to the extent legally possible, so "that commodities may freely move". Second, the Commission was required to make a thor ough investigation of the freight rate structure to deter- 217 mine if any undue burdens were imposed upon or undue advan tage was being received by any localities, parts of the country, or commodities. Changes of rates were required, in accordance with law, if such undue burdens or advantages were found. In making such changes regard had to be given to comparative market values of the various commodities over a period of years as related to "a natural and proper devel opment of the country as a whole, and to the maintenance of an adequate system of transportation". Third, "in view of the existing depression in agriculture", the Commission was required to make lawful adjustments in freight rates so as to promote the free movement by carriers of the depression- affected products of agriculture. Such rates on agricultur al products were to be set at the lowest possible level "compatible with the maintenance of adequate transportation service". The resolution imposed quite a burden on the Interstate Commerce Commission. The provision requiring an over-all study of complicated railway rate structure, composed of an astronomically large number of rates, set up a task upon which the Commission seemed not especially eager to em- 17 bark. " The provisions relating to .regard for the condi- •^E. 0. Malott, The Hoch-Smith Resolution, (Washington, D.C.: 1942), pp. 2S-28. tions of industries and relative market values of commodi- 218 ties and giving agricultural products special consideration led to several difficulties. As conditions and market prices changed, rates would be subject to considerable fluc tuation. Because of the requirement of the Transportation Act of 1920 that a fair return upon the value of railway- proper ty be earned if possible, a rate reduction to a de pressed commodity would probably "be met by a demand from the railroads for an increase of rates on other commodi ties."^'-’ Furthermore, depressions in industries other than 15Ibid., p. 112. agriculture, even transportation itself, were possible, and actually became pervasive after 1929. Possible difficulty in administration of contradictory mandates is well illustrated by the inconsistency between the Act of 1920 and the Hoch-Smith Resolution. The former was concerned with the well-being of the railroads. The latter was concerned xvith depressed industries, particularly agriculture.-^ The Supreme Court resolved the conflict. It -*-^For a discussion of specific difficulties which arose, see Malott, o£. cit. nullified the directive power of the Hoch-Smith Resolution by interpreting the resolution as meaning the existing law, i.e.. the Transportation Act of 1920, should be given 219 effect.15 15Ann Arbor R.R. Go. v. United States, 281 IT. S. 658 (1930). The Emergency Railroad Transportation Act of 193315 il- ^■®43 Statutes at Large, part I, pp. 211-221. lustrates the effect of changing conditions resulting in em phasis on particular aspects of regulation. The purposes of the act, according to the statement of purpose, 1 *7' were: 17Ibid., pp. 212-213. (1) to encourage and promote or require action on the part of the carriers to avoid unnecessary duplication, control allowances, services and charges, and to avoid other waste and preventable expense; (2) to promote financial reorgani zations to reduce fixed charges; and (3) to provide for the immediate study of other means of Improving conditions in transportation and the preparation of plans for such im provement. These objectives were limited by a very important qualification. Improvements which resulted in a decrease in employment, except for normal attrition, or which would worsen a worker's compensatory position were not to be ef fected. Such a limitation obviously would hamper any effec 220 tive reorganization which followed the desires expressed in the statement of purpose. There is one rather interesting point to the statement of purpose and the provisions for effecting these stated purposes. All three purposes stated above were aimed at providing for the improvement of the railways. Power and responsibility for effecting such improvements were placed largely in the hands of the newly created office of Federal Coordinator of Transportation. In essence, the apparent in tention of Congress was to improve the conditions of pri vately owned and managed firms by pressures, compulsive and persuasive, placed upon these firms by a government offi cial. The question arises as to the justification for the continued private ownership and management if such guidance by a government official is necessary. The provisions of the Emergency Act contained an implicit criticism of private railway management’s initiative and ability to cope with changing circums tances . l®The reports of the Coordinator contained more explic it criticism. See, Federal Coordinator of Transportation, Freight Traffic Report, (Washington, D. C.:1935); idem, Passenger Traffic Report, (Washington, D. C.: 193571 idem., Report on Freight Car Pooling, (Washington, D. C.: 1935); idem., Traffic Organization Report, (Washington, D. C.: 1935). Before proceeding to a consideration of the next major Act regulating railways, the Transportation Act of 1940, the 221 purposes of this study would be served best by reviewing the apparent policy with respect to regulation of other trans port media. Of particular significance are water and motor carrier regulation. Pipe lines moving oil were placed under Interstate Commerce Commission regulation by the Hepburn Act of 1906. Regulation of pipe lines followed general railway policy.19 19Regulation of air carriers is ommitted. There is a different regulatory body, the Civil Aeronautics Board, and the concern is more with promotion than regulation. Water transport was subjected to Interstate Commerce Commission regulation under the Act of 1887 if the water transportation was part of a joint water and railway move ment under common "control, management, or arrangement, for a continuous carriage or shipment.Joint rail and water ^924 Statutes at Large, p. 379. rates were subject to Commission review as to reasonable ness. Such review was of importance to both shippers and water carriers. Provisions to enforce competition between the railways and water carriers awaited the passage of the Panama Canal Act of 1912.91 Two paragraphs of the Act are pertinent to ^ ‘*'37 Statutes at Large, part I, pp. 560-569. 222 the basic goals of legislation prior to 1920. Railways were not "to own, lease, operate, control or have any interest whatsoever (by stock ownership or otherwise, either direct ly, indirectly, through any holding company, or by stock holders or directors in common, or in any other manner)" in competing water carriers operating through the Panama Canal or elsewhere. The other relevant provision of the act re ferred to water carriers which were controlled by railways operating in waters other than the Panama Canal. Extension of the service of such carriers was to be permitted only if it "would neither exclude, prevent, nor reduce competition on the route by water". Such water transport was to be sub ject to regulation in the same manner as the controlling railway. Both provisions were ,aimed at enforcing competi tion, the first through preventing railroads from swallowing up potential water carrier competition, the second through prevention of railroad control of new routes which might be competitive. The act thus followed the general pattern of transport legislation of the period. The first legislation applying to general water trans- O p portation was the Shipping Act of 1916. The regulatory ^®39 Statutes at Large , part I, pp. 728-738. provisions followed the pattern of railway regulation in prohibiting unreasonable rates and practices, discriminatory devices and unfair methods of driving out competition. Of particular* interest is the statement of purpose included in the title. The act was for the "purpose of encouraging, de veloping and creating a naval auxiliary and naval reserve and a merchant marine to meet the requirements of the com merce of the United States with its Territories and posses sions and with foreign countries; to regulate carriers by water engaged in the foreign and interstate commerce of the United States; and for other purposes." Promotion bulked large In the purposes of the act. No mention, however, was made regarding the purposes of the regulatory provisions. The individual regulatory provisions of the act were direct ed at the lessening of monopolistic devices, particularly discrimination. In contrast to railway regulation, pooling, rate fixing and other such agreements, not found to be un just or unfair, were to be permitted. Apparently the aim was to encourage some degree of monopoly power in transport ation by water carriers. In the Transportation Act of 1920 it was "declared to be the policy of Congress to promote, encourage, and develop water transportation, service, and facilities." To imple- 23 41 Statutes at Large, part I, p. 499. ment this policy the Secretary of War was directed to make studies of water traffic oriented to the promotion of the traffi c. 224 The Intercoastal Shipping Act of 1933^represented an ^ 4 7 Statutes at Large, part I, pp. 425-427. ------VUBL------extension of the provisions of the Shipping Act of 1916. No change in basic policy toward carrier regulation was eviden ced. Other legislation dealing with water carriers included the Merchant Marine Acts of 1920, 1928 and 1936. These three acts were primarily promotional rather than regula tory. Maintenance of a merchant marine for defense pur poses seemed to be a primary aim of the legislation. Con trols vested In the United States Maritime Commission, later exercised by the Federal Maritime Board, seemed to be di rected at competition which might be ruinous to individual lines. 25 D. P. Locklin, Economics of Transportation, (Chicago: Richard D. Irwin, Inc., 1954), pp. 778-779; Bigham and Roberts, oje. cit., pp. 260-261. The provisions of the Panama Canal Act of 1912 were for the prevention of railway domination of water transporta tion. The Shipping Act of 1916, i/rfaile subjecting water rates to regulation, was directed toward helping establish water carriers strong enough to compete with the railways. As a result, pooling, rate fixing and other such agreements were permitted. Federal motor carrier regulation began in 1935 with the 225 26 passage of the Motor Carrier Act of 1935. Of principal Statutes at Large, part I, pp. 543-567. importance in the act is the declaration of policy: It is hereby declared to be the policy of Congress to regulate transportation by motor carriers in such manner as to recognize and preserve the inherent ad vantages of, and foster sound economic conditions in, such transportation and among such carriers in the public interest; promote adequate, economical and efficient service by motor carriers, and reason able charges therefore, without unjust discrimina tions, undue preferences or advantages and unfair or destructive competitive practices; improve the relations between, and coordinate transportation by and regulation of, motor carriers and other car riers; develop and preserve a highway transporta tion system properly adapted to the needs of the commerce of the United States and of the national defense; and cooperate with the several States and the duly authorized officials thereof and with any organization of motor carriers in the administra tion and enforcement of this part. While the generality of the policy declaration makes precise interpretation difficult, the recognition of the im pact of motor transport on the other regulated transport agencies comes out quite clearly. Recognition of the ef fects of motor carrier operations upon other carriers, par ticularly the railways, Is so apparent that the temptation exists to charge that the motivation behind the passage of the act was the protection of the railways from the compe tition of motor carriers. Examining Individual provisions of the Motor Carrier Act of 1935 does not allay the temptation.26 226 26 It Is interesting to note that there was no clamor on the part of shippers for such regulation. Nor was the trucking industry united in its attitude towards regulation. See, J. C. Nelson, "The Motor Carrier Act of 1935," Journal of Poli ti cal Economy, Vol. XLIV (August, 1936), pp. 464-504. Entrance into interstate motor transport was subject to licensing by the Interstate Commerce Commission. Financial responsibility and consistency with the public interest was to be required of both contract and common carriers upon their applications for permits or certificates of conven ience and necessity.^ The Commission was empowered to set ^For common carriers to receive a certificate of con venience and necessity the service must be required by pub lic convenience. To receive a permit, contract carriers1 service must be consistent with the public interest. maximum, minimum, or specific rates of common motor carriers and the minimum charge of contract carriers. Contract car riers were subject to a requirement of 30 days notice pre ceding a rate decrease. Also contract carriers were not permitted to charge a rate lower than that filed with the Commission. These provisions of the Motor Carrier Acto of 1935 seem to have been aimed at the elimination of the entry of rate cutting; contract carriers. Such a limitation on entry should result, of course, in the lessening of competition in transport. Shortly after the passage of the Motor Carrier Act, J. G. Melson stated, 227 These controls will dovibtless render it consider ably more difficult for truck drivers and others to establish themselves in the trucking business by the familiar device of cutting rates, thereby ef fecting a diversion of the traffic from existing motor and rail carriers. Whether such a result will as predicted by the sponsors of this legis lation, react to the public or shipper good in the long run remains to be seen although, if stabili zation 'is actually accomplished, carriers having well-financed and well-established businesses prior to the effective date of this law should stand to gain from the elimination of competition from the irresponsible ’wildcatter* operators. ® CO Ibid., p. 474. Having reviewed apparent policy goals of regulatory legislation over transportation which was enacted prior to 1940, it is possible to consider the Transportation Act of 1940^ with some understanding of the background and experi- ^ 5 4 Statutes at Large, part I, pp. 898-956. ence which had a bearing on its enactment. For the purposes of this study, the most significant feature of the Transpor tation Act of 1940 is the statement of a national transpor tation policy. Ho previous legislation explicitly stated a national transportation policy for all carriers. As the Transportation Act of 1940 placed water carriers under the Interstate Commerce Commission regulation, the statement of policy applied to railroads, motor carriers, water carriers and oil pipe lines. The policy statement modeled after the 228 Motor Carrier Act of 1955, states, It is hereby declared to be the national trans portation policy of the Congress to provide for fair and impartial regulation of all modes of transportation subject to the provisions of this Act, so administered as to recognize and preserve the inherent advantages of each; to promote safe, adequate, economical, and effic5.ent service and foster sound economic conditions in transportation and among the several carriers; to encourage the establishment and maintenance of reasonable charges for transportation services, without unjust dis criminations, undue preferences or advantages, or unfair or destructive competitive practices; to cooperate with the several States and the duly authorized officials thereof; and to encourage fair wages and equitable working conditions;--all to the end of developing, coordinating, and pre serving a national transportation system by water, highway, and rail, as well as other means, adequate to meet the needs of the commerce of the United States, of the Postal Service, and of the national defense. All of the provisions of this Act shall be administered and enforced with a view to car rying out the above declaration of policy. The comprehensiveness of the mandate places a task of great magnitude before the Interstate Commerce Commission. Fulfillment of the individual stipulations of the statement appears to offer great difficulty. The difficulty arises because of the generality of the statement which makes defi nition of particular provisions difficult and because of possible conflict, and even contradiction, between indivi dual provisions of the directive. As Dewey points out, fair and impartial regulation of all modes of transportation so administered as to recognize and preserve the inherent advantages of each may mean that water and motor carriers are not to be subject to the same 229 *zr\ pattern of regulation as rail carriers. If the economic Ralph L. Dewey, "Transportation Act of 1940," .Ameri can Economic Review, Vol. XXXI (March, 1941), pp. 18-20. characteristics are different, then the nature of control also should differ. Assuming that motor and water carriers are not natural monopolies, as defined In Chapter IV, regu lation to "preserve the inherent advantages" of the various transport media would include only guarantees of safety and responsibility. Railroads, assuming conditions of natural monopoly, would be subject to more thorough-going regulation particularly in restraints against discrimination and use of the strong powers to ruin motor and water competition. To the unsophisticated, however, motor and water carrier Incur sion into rail traffic, because of "inherent advantages", would appear to constitute as much "unfair or destructive competitive practice" as railway use of temporary rate-cut ting below variable costs to ruin actual or potential motor or water competition. Several possible contradictions appeared in tho policy proclamation of the Transportation Act of 1940. Encourage ment of fair wages may do other than foster sound economic conditions In transportation, if fair wages are interpreted to mean some type of "just wage" not related to the produc tivity of the workers. The provision for developing and preserving a transportation system adequate to meet the 230 needs of national defense also may be contradictory to "pre serving the inherent advantages" of the various modes of transportation. Maintenance of a transportation system ade quate for national defense requirements may lead to a prob lem similar to import tariff protection of national defense industries. Capacity much greater than required to meet the demand of the transport users would be supported by these same users for the presumptive benefit of the entire nation. An argument could be made, as in the protective tariff situ ation, for a subsidy from general tax revenues, charged to national defense, to support the excess capacity being main tained. A comparison then could be made readily between the expense to the taxpayer and the worth of the contribution of the industry to national defense. Having surveyed the basic law of transportation regu lation by the Interstate Commerce Commission from the view point of apparent goals behind the regulation, it is pos sible now to summarize the legislation. The summary will be useful in contrasting the economic standard of optimum allo cation of resources to the goals set up by Congress for the guidance of the Commission. The most evident characteristic of the legislation governing regulation of transportation by the Interstate Commerce Commission is the changing emphasis of the goals. Prior to 1920 the emphasis was on the protection of the rail user from discriminatory practices on the part of the rail 231 ways. The methods used to bring sueh protection into effect were provisions of a negative nature, i_.e_. prohibitions, 31 31R. L. Dewey, The Long and Short Haul Principle of Rate Regulation, (Columbus: The Ohio "State University Press, 1^ . ) ' , p." 12.-- and the attempted enforcement of competition, a method not to be recommended considering the nature of the industry. The Act of 1920 was more concerned with the railways than the rail user. Rates were to be set to yield a fair return, without, however, guaranteeing it, to the carriers. Provi- vision for setting up a plan of consolidation also represen ted a movement from emphasis upon competition. Promotion of a specific industry (agriculture) other than the industry being regulated was the basic theme of the Hoch-Smith Resolution. This represented the greatest devi ation from the general regulatory patterns of transportation legislation. The Emergency Railroad Transportation Act of 1933 was a partial incursion into private management's domain. Rec ommendations and c Quip ui.1 s i_ ons placed upon rail management by a government official were expected to improve rail trans portation. Congress seemed to be emphasizing the need for improved efficiency of operations to maintain rail carrier solvency in the face of the depressed economic conditions prevailing in the 1930's. 23 2 Concern over inter-carrier relationships dominated both the Motor Carrier Act of 1935 and the Transportation Act of 1940. Such concern was largely depression oriented. Com petition for the relatively meager traffic available put a strain on the financial strength of the carriers, particu larly of the railroads. Congress, at the same time, seemed to be sensitive to the nation’s need for adequate transport ation facilities to meet the requirements of possible pros perous times and war. Railroads are the backbone of the nation's transportation system in such periods of large movements of passengers and goods. The standards under which the Interstate Commerce Com mission must function are vague, variable and frequently in consistent as compared to the relative simplicity of the economic standard for regulation, optimum utilization of re sources. The Commission has difficulty reconciling the many individual goals. Such complexity is to be expected In a changing, complex world. Legislatures are subjected to pressures from many groups. General conditions and problems are continually changing and the situation confronting regu lated industries also changes. The result to be expected Is variation in emphasis on possible goals, Inclusions of new goals and inconsistency in the aims of regulation. B. Federal Public Utility Price Setting Policy In setting public utility prices, the decisions and 255 procedures of State and federal regulatory bodies are sub ject to judicial review. The result, through the years, has been that the Supreme Court of the United States sets the basic standards for the general level of rates. The rele vant cases must be examined, therefore, if the basic stan dard for rate setting is to be found. Although there are a great number of cases dealing with proper rats setting, relatively few cases are important to the line of inquiry being pursued.^ Two of these cases are ®^If an attempt to define "evolutionary" or "develop mental" patterns were the basis of this study, then all cases would, of necessity, be considered. The point of in terest here, however, is the basic standard of rate setting. the well known Smyth v. Ames and the Hope Natural Gas cases.The former was the basis for the whole series of ^ Smyth v. Ames, 169 U.S. 466 (1898). Federal Power Jommiaslon et. al. v. Hope Natural Gas Co.., 320 U.S. 591 (194i")". subsequent cases. The latter is considered as the ruling case at the present time. For an understanding of the stan dards applied by the courts, It is necessary to translate the language used by the Court into the language of econ omics . Examination of the two cases gives evidence of the Court1s seeming desire for rates, on the average, meeting average costs. If it is assumed that a public utility pays 254 competitive prices for labor, raw materials and other such currently used inputs, the problem remains of determining the proper total interest payment on the capital investment. The Court encountered this problem rather early in regula tory history. The solution basically involves opportunity cost determination. The problem breaks down into the deter mination of the proper rate of interest and the amount or base upon which the interest payment is to be made. In Smyth v. Ames the Court said, "What the company is entitled to is a fair return upon the value of that which it employs for the public convenience."^ 34169 U.S. 547. Largely ignoring the rate of return, the Court, through Justice Harlan, held "that the basis of all calculations as to the reasonableness of rates to be charged by a corpora tion maintaining a highway under legislative sanction must be the fair value of the property being used by it for the convenience of the public."33 The so-called rule or formula 35Ibid., 546. for the determination of fair value required that consider ation be given to original cost, amount expended in improve ments, market value of stocks and bonds, present as compared with original construction cost, probable earning power of 235 the property under the particular rates, and the operating expenses. This could be interpreted as stating possible bases for making comparisons with alternatives to find the value of the assets. Justice Harlan was aware of the difficulties in deter mining the appropriate base upon which the interest payment should be made. The criteria seem to have been stated as rough guides. As an indication of Justice Harlan’s aware ness that he was not presenting a formula is his statement, "The corporation may not be required to use its property for the benefit of the public without receiving just compensa tion for the services rendered by it. How such compensation may be ascertained, and what are the necessary elements in such an inquiry will always be an embarrassing question. 56Ibid., 546. Recognition of the possible over-inflation of the base upon which an interest return should be paid is Indicated by the following statement: If a railroad corporation has bonded its property for an amount that exceeds its fair value, or if its capitalization is largely fictitious, it may not impose upon the public the burden of such in creased rates as may be required for the purpose of realising profits upon such excessive valuation or fictitious capitalization, and the apparent value of the property and franchises used by the corporation, as represented by its stock, bonds and obligations is not alone to be considered when de- termining the rates that may be reasonably charged. 236 37Ibid., 544-545. The language of the Smyth case is not clear, especially in the insertion of irrelevant and contradictory elements among the bases to be considered. The lack of economic so phistication, it might be held, made for awkward presenta tion of the view that equivalent of competitive costs should be met by the revenues. Such a lack definitely is not pres ent in the Hope decision. Imphasis on meeting average cost comparable to competitive situations is more easily inferred from the Hope case. The Hope Natural G-as case is noted primarily for its statement, "Under the statutory standard of 'just and rea sonable' it is the result reached not the method employed rz q which is controlling." The controlling result or basic r z p 320 U.S. 602. standard is also significant for this study. The majority opinion stated, The rate-making process under the Act, i_.e_. the fixing of 'just and reasonable’ rates, involves a balancing of the investor and consumer interests. Thus, we stated in the Natural Gas Pipeline Go. case that 'regulation does not insure that the business shall produce net revenues.' But such considerations aside, the investor interest has a legitimate concern with the financial integrity of the company whose rates are being regulated. Prom the investor or company point of view it is impor tant that there be enough revenue not only for operating expenses but also for the capital costs of the business. These include service on the 237 debt and dividends on the stock. By that standard the return to the equity owner should be commen surate with returns on Investments In other enter prises having corresponding risks. I’hat return, moreover, should be sufficient to assure confi dence in the financial Integrity of the enterprise so as to maintain its credit and to attract capital.39 59Ibid., 603. Interpretation of the opinion is relatively easy as compared with the statements in the Smyth case. Operating expenses, determined In the market by the interaction of competing buyers and sellers, had to be met. But also, it was necessary for the investors to receive the equivalent of a competitive return. This is indicated by the statement, "the return to the equity owner should be commensurate with returns on investments In other enterprises having corres ponding risks." The language of this statement is strongly reminiscent of economics textbook discussions of opportunity cost determination. It says, in effect, that the investor should receive an amount equal to what he would have re ceived had he invested the same amount in a competitive in dustry having equivalent risks.-0 4L°This U n e of reasoning as to the two cases is contra ry to a common view that the Hope case overthrew the rule Smyth v. Ames. On this see, Martin G. Glaeser, "The United States Supreme Court Redeems Itself," Journal of Land and Public TJtili ty Economics, Vol. XVIII (May^ 1942), pp. 146-154. 238 Although ths Smyth and Hope cases are of greatest im portance to this study, several other cases made relevant points. In upholding the constitutionality of the recapture clause, the Court emphasized the absence of any right to more than the fair return on a fair value. Chief Justice Taft, speaking for the Court, stated, The carrier owning and operating a railroad however strong financially, however economical in its facilities, or favorably situated as to traffic, is not entitled as of constitutional right to more than a fair net operating income upon the value of its properties which are being devoted to transportation. By investment in a business dedicated to the public service the owner must recognize that, as compared with in vestment in private business, he can not expect either high or speculative dividends but that his obligation limits him to only fair or reas onable prof It. 41Dayton-G-oos e Creek Railway Company v. United States, 263 U.S. 456 (481) (1924). On the other hand, there is no guarantee of a return meeting all costs. Considering the problem of a street railway which was facing losses, the Court stated, "The due process clause has been applied to prevent governmental de struction of existing economic values. It has not and can not be applied to insure values or to restore values that have been lost by the operation of economic forces."^ 42 Market Street Railway Co., v. Railroad Commission of California, 3*24 U.S. 548 (5677^(1945). 239 Because of difficulties in computation of the costs of individual services offered by a public utility, it is not necessary for each rate or price to equal average cost. The Court, on this question, stated, Vjhere rates as a whole are under consideration, there is a possibility of deciding, with more or less certainty, whether the total earnings afford a reasonable return. But whether the carrier earned dividends or not sheds little light on the question as to whether the rate on a particular article is reasonable. ° "^Inters tate Commerce Commission v. Union Pacific R.R. Co. , 22lS~U. S . 541 T549J (19121. Summarizing these decisions, the requirements of due process are met if a commission sets rates to make available to the firm an amount equal to operating expenses plus the proper Interest return. The proper interest return is equal to the amount the Investment would earn under competitive conditions. A firm is not entitled to more than this com petitive return, even if the firm’s Income is reduced to the competitive return by seizure of part of the funds. How ever, a firm Is not guaranteed such a return in face of ad verse economic conditions. Furthermore, it Is not necessary that each Individual rate equal average cost but rather that individual rates tend to cluster about the average cost as a central tendency. If the foregoing analysis was correct, two points of interest stand out. The judgment of the proponents of aver- 240 age-cost pricing that consumers should pay the full costs of services supplied them is shared in a general way by the Supreme Court. Considerations of optimal output are not of primary importance. The other point refers to the perennial argument as to the proper computation of the rate base. By this analysis, discussions as to original cost, reproduction cost, prudent investment or other such concepts represent hypotheses as to the corroct method of calculating the com petitive return which is basic to calculation of the full costs of production. CHAPTER IX USE OF THE THEORETICAL CONCEPTS IN ACTUAL PRICE SETTING Practical considerations were neglected in the earlier examination of the theory of public utility pricing. The possibility of practical application of the theory should be considered for use as at least a partial judgment on the practical value of the theory. Several problems are encountered in the practical ap plication of the theoretical tools. The first problem is the measurement of costs and demand. Another problem arises in attempting to insure the efficiency of output, i.e. pro ducing any given output at the lowest possible cost. A third involves the collection of funds to pay the subsidy required with marginal-cost pricing. Public utility price setting aimed at achieving an optimum In an economy in which Imperfections exist is the next problem. The advisability of new Investment in practice Is the fifth problem area. Each of these problems will be considered in the order stated. A. Measurement of Cost and Demand Functions Precise knowledge of the demand and cost functions was assumed in the discussion of the theory of pricing. In ac tual practice many difficulties exist in the computation of either demand or cost functions. Statistical attempts to determine costs exhibit a low degree of success. This Is 241 242 true of the determination of any of the coat concepts used in economic theory. ■^For a summary of cost studies see, Committee on Price Determination, Cost Behavior and Price Policy, (New York: National Bureau of1 Economic Research, 1943). Many writers have criticized marginal-cost pricing on o the basis of the difficulty of computation. Several points ^For a summary of computational difficulties, see Robert W. Harbeson, "A Critique of Marginal Cost Pricing,rl Land Economics, Vol. XXXI (February, 1955), pp. 54-74. receive special attention. The nature of available data is not compatible with the determination of marginal cost as defined in theory, jL.e^ increase in total cost as output in- creases by one unit. Because of the large discrete ^The mathematical definition requires output Increases to be inf ini te sinal. This would require a continuous total cost curve. changes in output commonly found in practice, it would be necessary to average marginal cost over a considerable range, i_.£. divide the increase in total costs by the amount of the changed output. Another difficulty is added if joint products are pro duced. In fact, marginal cost of the individual items is indeterminate in such a situation. Only the marginal cost 245 of tli© combination of products can be calculated. However, no pricing problem results, assuming marginal cost of the combination can be computed. If the sum of the prices equals the marginal cost of the combination of products, the proper solution is reached. Though marginal condition five (equality of marginal rates of transformation and marginal rates of substitution) is not met, the welfare optimum is realized subject to the constraint of production In fixed proportions for the joint products. Pour other difficulties are somewhat related as their practical determination requires considerable conjecture on the part of the person doing the computation. The first Is the well known difficulty Involved in the allocation of com mon costs. Common costs are usually allocated by some stan dard formula. However, dispute over the proper formula is chronic, as witness controversies over multipie-purpose fed eral projects which also produce electrical power. Deter mination of user costs, i_.e_. wearing out of equipment from operation as distinct from depreciation not related to out put, adds another difficulty in marginal-cost determin ation.^ Long-run costs, including marginal cost, are of a ^For the difficulties involved In measurement, see W# A. Lewis, "Fixed Costs," Economica, Vol. XIII N.S, (No vember, 1946), p. 232. rather conjectural nature. To vary plant size or to find 244 comparable plants of different sizes is almost impossible. Precise knowledge of the long-run cost function is extremely unlikely. The last difficulty of computation arises because of possible difference between private and social costs, marginal or others. External economies or diseconomies cause these discrepancies. External economies or diseconom ies are often difficult to recognize, effect of smoke upon health. To calculate them even when recognized becomes almost impossible. A more basic difficulty in the determination of mar ginal cost arises from the very nature of the definition of the short-run. As W. A. Lewis has pointed out, marginal cost as a single quantity does not exist.® The simple 5Ibid., p. 233. choice between short-run and long-run marginal cost does not even exist. There are a great number of marginal costs de pending upon the time involved and the commitment of resour ces. By this Lewis means that as of any given time there are a whole series of short-run marginal costs. Which one is applicable to the particular situation depends upon the distance into the future one is looking, or, more correctly, to what extent resources are committed. For example, a railway1s marginal cost is one thing when allocating resour ces a month in advance. It is another when the plans are 245 for a week in advance and rolling stock and men are moving toward or are situated in another location and still some thing else an hour before the train is to run and it be comes difficult to adjust the train makeup without serious dislocation. In general, the marginal cost curve becomes more inelastic the shorter the time there is to make adjust ments, i_.the greater the commitment of available resour ces. While the objections mentioned above are directed at marginal-cost pricing, they apply to average-cost pricing to a similar degree. Objection to marginal-cost pricing on any of the above mentioned practical considerations are not ar guments in favor of average-cost pricing. All points of difficulty in the computation of marginal cost could be di rected at the computation of average cost just as easily. An average cost function Is computed no more easily than a marginal cost function. Because of their mathematical re lationship, the computation of one implies the derivation of the other. If the average cost function could be computed, the marginal cost function could be readily derived. Lewis' point regarding the multiplicity of short-run marginal cost curves applies eqiially to short-run average cost. There are any number of short-run average cost curves depending on the time involved and the commitment of resources. Average cost has one big advantage over marginal cost from the standpoint of computation. Average cost at one particular output Is 246 much easier to estimate than the corresponding marginal cost. Average cost, thus represents a standard which is more readily verified than marginal cost. Statistical attempts at the derivation of demand en counter as much difficulty as the determination of cost functions. Most of the difficulties arise because the g For a good discussion of the meaning of statistical demand curves, see E. J. Working, "What Do Statistical 'Demand Curves' Show?" Quarterly Journal of Economics, Vol. XLI (1926-27), pp. 212-235, reprinted in Readings in Price Theory, ed. by K. E. Boulding and George J. Stigler, (Chicago: Richard D. Irwin, Inc., 1952), pp. 97-115. data must be collected over a period of time. During such a period changes are taking place in incomes, consumer tas tes and the price level. It is very difficult to adjust the data for such changes, though it is just these variables which are assumed to be constant in the theory of demand. It would seem that a vague guess, at best, as to the demand for public utility service would be held by either the man agement or the commission. Some experimentation in price setting, if it were possible, probably would be necessary to estimate demand conditions. The problem of demand computa tion is common to any principle of price setting, i_.e_. it is just as difficult to determine demand for marginal-cost pricing, average-cost pricing or any other standard for pricing. Very great difficulties arise in particularizing the 247 theory to specific industries. Xn the theory output was stated in abstract homogeneous units. In practical appli cations these uxnits must be defined in concrete terms. Dif ficulty arises because consumers do not purchase a homo geneous product in most public utility markets. A few il lustrations will demonstrate this point. Electricity is not purchased in homogeneous kilowatt- hours. Rather, it is a certain quantity of power delivered at a specific location at a specific time. Each consumer is actually purchasing a different product. Any pricing prin ciple would require a price at the plant for energy pro duced plus amounts for transmission expenses, installation, servicing, metering and any other attendant expenses. This would require the computation of s everal marginal or average costs (depending on the standard) for each of the thousands of consumers. With marginal-cost pricing a bill to any con sumer would equal marginal cost of producing the electricity at the plant at the specific time it was being produced plus anoints equal to the marginal costs of transmission on main lines, installation in the specific location, servicing re quired, metering and record keeping, and many other specific expenses which are probably known to experts in the elec trical industry. The same would be true if average-cost pricing were used as the pricing standard. Very similarly, gas is not purchased in b.t.u.’s or a volume measure. Rather, it is a certain quantity of gas 248 at a specific location at a specific time which is pur chased. As with electricity, a tremendous amount of cost information would be needed. Transportation is not purchased as pure ton-miles. Each load transported involves separate handling, loading, storing and other such expenses. A price would be equal to a charge for pure transport plus amounts for the other indi vidual expenses. If marginal-cost pricing were used the charge would be eqtial to marginal cost of the pure transport plus the marginal costs of the additional elements. As each public utility Industry exhibits its own unique problems, investigation of each individual industry fur nishes material for a study in Itself. Individual studies of particular industries are probably desirable in fur nishing information as to the feasibility of different pos sible pricing standards in the particular industries. Each such study would require detailed examination of the nature of the product and the technology of the particular Industry to determine the separate possible costs involved. *7 ^Proposals as to the nature of the particular pricing methods in the specific industries based on anything other than Intimate knowledge of the specific industry may err In over -simplifying. For a brief discussion of marginal-cost pricing as ap plied to individual industries, see B. P. Beckwith, op. cit., pp. 226-261. 249 B. Guaranteeing the Efficiency of Operation A major problem which commissions must face is the as surance of the most efficient operation of plant. Whether marginal cost or average cost is used as a pricing base, the computation must represent the most efficient possible oper ation. Otherwise not much is to be gained from regulation. Under either pricing principle the firm will cover costs of operation. With such a guarantee, the temptation to permit inefficiency would be rather great. Theoretically, the test for efficient production of any output is quite simple. Equality between the ratios of the physical marginal product and the price of the input for all Q inputs is the necessary condition. If this ratio is higher O A. P. Lerner, Economics of Control, op. cit., pp. 119- 122. A proof of this principle may be found in Ch. II of thi s s tudy. higher for any one input than any others, the firm can pro duce the output more cheaply by expanding the use of the one input and contracting the use of the others. Conversely, If this ratio is less for one input than for the others, the use of this input should be contracted and the use of the others expanded. A difficulty arises because of possible qualitative differences In one particular input, plant and equipment. The cost curves used In considering the setting of output 250 assumed the use of the most efficient plant and equipment available. However, less efficient plant and equipment could be used to produce the same output and the proportion ality rule could be met. The short-run cost curve would lie well above the long-run cost curve at all points. Above the long-run cost curve, which is composed of a series of short- run cost curves representing the use of the most efficient plant and equipment, are a great number of possible short- run cost curves all representing use of less efficient plant and equipment. At some point on each of the curves repre senting less efficient plant and equipment the ratio rule could be satisfied. Hoxvever, the same output could be pro duced more cheaply by qualitatively superior plant and equipment. The problem to be faced by a commission can be broken into two parts. First, the commission must be sure the most efficient available equipment is being used. Next, there must be assurance of the most efficient use of this equip ment, i,.e.. meeting the proportionality rule. The implemen tation of the latter requirement furnishes no great diffi culty, assuming marginal productivity can be measured with some accuracy. Choice of the most efficient plant furnishes great dif ficulty. Every situation differs in some respect. ^hat is an efficient plant in one location is not necessarily, and probably is not, an efficient plant in another location. As 251 a result, a commission would have difficulty in determining the degree of efficiency of any one plant. Pilot plant studies or "yardstick” operations would be of little help. In any comparison, the lack of comparability would become the significant point. Commissions, as a result, would be required to accept the testimony of .engineers as to the ef ficiency of any plant. When conflicting testimony would be presented, the commission would face a dilemma. While the implementation of the proportionality rule is not difficult assuming measurement of marginal productivity, such an assumption Itself avoids many difficulties. To measure the marginal productivities, all inputs must be variable In infinitesimal amounts. Such a variability Is difficult to arrange even under ideal conditions. C. Collection of the Tax and Payment of the Subsidy One problem exists which is peculiar to marginal-cost pricing. Difficulty arises In determining the gain to each consumer resulting from marginal-cost pricing as opposed to average-cost pricing. such determination Is necessary if a tax Is to be placed to yield the needed subsidy without sub sidizing public utility consumers at the expense of others. Because of the subjective nature of the gain from mar ginal-cost pricing, questioning each Individual consumer Q would seem to be the only way to determine such gain. The 252 ^Henderson, "The Pricing of Public Utility Undertak ings," op. ci t., pp. 231-232. practicality of the necessary individual inquiry is dubious at best. Each consumer would be required to state the lump sum he would be willing to pay to assure a low price rather than paying a higher price. In practice, pragmatic expedients would probably be used. One of these would be the use of objective standards such as value of property, income or size of buildings. The tax would then cease being a Tump sum tax. It vvould become a tax on property, Income or building. The basic principles of marginal-cost pricing woLild be violated. A basic criticism of the marginal-cost principle has been that the precedent of subsidizing firms not meeting full costs is undesirable.^^ This argument states that, ■^T. Wilson, op., clt., p. 461. although a subsidy may be defensible in a particular indus try, granting such subsidies would result in a clamor for subsidies In any area where losses occurred. Inhere would be danger of Indefensible subsidies becoming ubiquitous. D. Market Imperfections and Marginal-Cost Pricing Existence of imperfections, i_.p. monopoly elements, in other markets requires deviation from marginal-cost pricing in regulated Industries according to the analysis of Chapter 253 VI. The correct rule then Is to set the price where the ratio of price to marginal cost in the regulated Industry is equal to the weighted average of the ratios of prices to marginal costs throughout the economy. This rule follows from the fifth marginal condition for an economic optimum. According to the fifth marginal condition, the marginal rate of transformation between any two products for the community must equal the marginal rate of substitution between these products for any two consumers consuming the products. According to this argument, the regulated Industry would be overproducing If marginal-cost pricing is used, im perfections existing elsewhere. Resources used to produce the public iitllity service could provide greater satisfac tion to consumers in other lines of production. Instead of setting price equal to marginal cost, the public utility commission should set price higher, i..e.. where the ratio of price to marginal cost equals the average ratio of price to marginal cost throughout the economy. Public utility commissions faced with market imperfec tions in the economy and following the average price to mar ginal-cost ratio policy would be faced with very great problems. The calculation of marginal cost In the regu lated plant would furnish all the difficulties mentioned above. -&ut the commission would also be required to know the marginal costs and prices in all firms in the economy to compute the weighted average of prices to marginals costs. 254 Someone, at any rate, would have this huge task of statis tical computation and design of many experiments to obtain the basic data. Pricing so the ratio of price to marginal cost equals the same ratio prevailing on the average throughout the economy has one possible bright feature. The possible salu tary result would obtain if the average price to marginal cost ratio were greater than the ratio of average cost to marginal cost in the regulated plant. If such a situation prevailed, the price of the public utility service would be higher than the average cost. Many advantages would be present. Taking existing market imperfections as given, the ideal output would be produced. The price would more than cover average costs. A profit would accrue, making public utility owners and management happy. There would be no necessity of paying a subsidy, thus eliminating all the problems of computing individual gains from marginal-cost pricing. If the average price to marginal cost ratio for all In dus tries.were equal to the ratio of average cost to marginal cost in the regulated industry, the result would be more convenient. By the equal ratio rule, the regulated price would equal average cost. Wo subsidy would be necessary. At the same time, the regulated industry would be receiving no pure profits. Furthermore, oiitput would be ideal, taking existing market perfections as given. 255 E. Standards Tor Investment According to Chapter VII, the standard for new invest ment under marginal-cost pricing is that the consumers1 surplus plus expenditure on the product be greater than the costs incurred in providing the new service. The use of the consumers' surplus standard would require the measurement of such a surplus in every instance of possible investment. Such measurement would be extremely difficult and perhaps Impossible. It would require asking all potential consumers the maximum they would be willing to pay per period of time rather than do without a particular service which Is not available to them at the time of questioning. The question able accuracy of responses to such questioning should be ob vious. Il1he end result of the attendant Inaccuracy would be considerable second guessing as to the actual gains to be derived from new Investment. Because of the vagueness surrounding the gains from new investment, there might be some tendency toward overinvest ment. This would be true in governmental projects. As the pay and prestige of a manager of a government project Is positively correlated to the size of the project and number of employees, the temptation for empire building might bias errors on the overestimation side. Even with privately owned and operated public utility ventures, there might be some tendency toward overinvestment. Public utility commissions, entranced by economies of scale 256 and the potential power of regiilation of mammoth enter prises, might be prone to encourage overinvestment. Public utility managers seeing possible subsidies from government might be encouraged to overinvest rather easily. However, this temptation for public utility managers to overinvest would be offset by the counterbalancing temptation to under invest to make hidden monopoly returns. A profitability standard, i_.e_. meeting total costs, would make a more unambiguous, verifiable ex post standard for new investment.^ In deciding on any new investment, ^Harbeson, ojo. cl t., p. 67. the decision to proceed would follow an affirmative answer to the question, "Can It pay its way?". However, the ac ceptance of the profitability standard means the sacrifice of optimum use of resources for the convenience of the more practi cal. Combination of the profitability standard with some degree of freedom of decision by the public utility manage ment offers advantages. Management could be permitted to make decisions as to product promotion, expansion of plant or introduction of new techniques. The responsibility for the decisions could rest with management. If losses were incurred, the stockholders would bear the loss and manage ment presumably would feel the wrath of the owners. If 257 some profits were made, the public utility could be allowed to retain them subject to gradual reductions in the profits with rate decreases. The New Jersey or Washington plans or some similar device could be used to slowly diminish prof its.12 12.For a description of the New Jersey and Washington plans, see Troxel, Economics of Public Utilities, op. cit., pp. 406-417. Under any such system the regulatory commission must take care that the public utility management does not at tempt to insure profitable operation through underinvest ment. The possibility of any practical, simple, unambiguous method of preventing underinvestment where the private uti lity management has personal discretion seems, however, about as remote as devising a simple method for computing consumers' surplus. F. Summary on Practical Application of the Theoretical Tools The difficulties involved in the application of mar ginal-cost pricing are rather imposing. Measurement of mar ginal cost furnishes great difficulty. Guaranteeing of ef ficient operation of the plant represents a problem of great magnitude. Measurement of the compensating variation, i_.e_. gains from marginal-cost pricing as opposed to average-cost pricing, is impractical. Determination of consumers' sur- 258 plus for Investment is close to impossible. If techniques were developed for coping with the problems, the expense in volved might overwhelm any advantages. Average-cost pricing with any high degree of precision does not come off much better in practical application. Al most all of the difficulties encountered in computation of marginal cost are also encountered in the computation of average cost. However, because no subsidy is necessary un der average-cost pricing, the problems of tax collection methods and feasible size of the tax are avoided. If no great Insistence on precision exists, a trial and error ap proach to average-cost pricing might be used. The goal to be achieved with such an approach would be to attain the largest possible plant which could pay Its long-run total costs. Caution should be used, however, to avoid having a smaller plant. Any scale of plant smaller than this largest scale could pay its way even more easily. There would be great temptation, therefore, for public utility management to maintain a smaller scale of plant. It would be the func tion of the public utility commission to guard against such action. If a recommendation were to be called for, It would be In favor of aver age-cost pricing. All the practical advan tages are in Its favor. By a trial-and-error approach, the results would be amenable to ex post test. This would be true of standards for pricing and the standards for new in 259 vestment. There would he no necessity for levying a special tax. A basic policy of average-cost pricing would eliminate the most outrageous monopoly restrictions. Furthermore, as suming the existence of many imperfections in the economy, the optimum price would be greater than marginal costs of producing the optimum output. On the assumption of equal Ignorance, average-cost pricing might lead to the optimum with as great likelihood as marginal-cost pricing. Special attention must be given to encouragement of in novation. The recommendation for average-cost pricing must be modified to provide for such encouragement. It would seem that incentives to innovate would be greater if public utility firms were given assurance that any profits arising from innovation would not Immediately disappear. Choice of commissioners would be very helpful in this respect. Hven more helpful would be some statutory assurance. A plan similar to the Hew Jersey or Washington plans would assure public utility firm of realization of retiirns due to inno vation. SUch a plan would also make possible the ultimate passing of innovational benefits to consumers. An approxi mation to one of the economically desirable aspects of com petition is achieved in this manner. As a final word, it might be worth risking some monopoly restriction to gain the greater advantages arising from innovation. A ‘similar case is made in considering anti-trust and patent policies. CHAPTER X SUMMARY, CONCLUSIONS AND RECOMMENDATIONS Contemporary economic welfare theory has two principal areas of analysis. One is the area of efficiency in produc tion and consumption. The guiding rule, assuming given technology, resources, tastes and income distribution, re quires the meeting of seven marginal conditions for the achievement of the economic welfare optimum. These marginal conditions are met if prices equal marginal cost throughout the economy. Prices eqxial marginal cost under perfect com petition. The existence of perfect competition, therefore, results in the achievement of the optimum. The other area of economic welfare analysis covers reorganizations in which changes occur in the total output. Any one state can be said to be an improvement over another if those who gain can and do pay sufficient compensation to those who lose. No statement can be made as to optimal income distribution ex cept on the basis of some arbitrary value judgment. With the maintenance of freedom of entry, the optimal condition of price equal to marginal cost occurs with- any firm or industry cost structure except decreasing long-run marginal costs to scale for the firm. Natural monopoly oc curs in situations In which decreasing long-run marginal costs exist at scales of plant small enough to be compatible with perfect competition. The profit possibilities foreseen by each firm induces expansion of plant size. Eventually 260 261 the size of the plant brings monopoly power— either pure monopoly or some type of oligopoly solution. Using optimum utilization of resources as the basic standard, a public utility is defined as an industry in which natural monopoly exists and in which the individual gains from regulation are sufficient to pay for the regula tion. To achieve maximum economic welfare, marginal-cost pricing is the proper standard, if payment could be made sufficient to overcome the losses incurred. The use of the long-run marginal cost curve as the basis for judgment on new investment and contraction of plant leads to maximum welfare. Average-cost pricing is consistent with the value judgment that consumers should pay for the full costs of production. Average-cost pricing is not consistent with maximum welfare. Practical application of the marginal-cost pricing standard encounters severe difficulties. Existing statisti cal and accounting techniques make the computation of demand and costs quite Impractical, if not impossible. Efficiency of operation of plant is difficult to determine. Collection of the tax used to finance the required subsidy without vio lating the welfare standards approaches impossibility. With the existence of imperfect markets, the rule is not price equal to marginal cost. Instead, the price should equal the weighted average of all the ratios of price to marginal cost in the economy. The use of marginal costs and consumer’s 262 surpluses for the determination of new investments lacks the elements of daring, initiative and imagination basic to major innovational investment. The inclusions of goals other than the economic goal of optimum utilization of resources also limits the applicabil ity of a rigid marginal-cost pricing standard. Even if com putational difficulties did not exist, a severe limitation to the application of marginal-cost pricing would result from the inclusion of non-economic and uneconomic goals in the norms for public utility pricing. The goals included in the legislative norms for public utility regulation may be inconsistent not only with the economic standard but also with each other. Two sets of conclusions arise from the analysis of the theory of optimum public utility pricing. One refers to general recommendations for actual regulation. The other set is concerned with the economist as an analyst of public utility regulation. In making recommendations for the regulation•of public utilities, suggestions as to rate policy and the functions of public utility commissions will be considered. In agree ment with the recommendations in Chapter IX, a general pro- gram of average-cost pricing Is suggested. Some qualifica tions to this general policy must be made. As mentioned in Chapter IX, computation of average cost of services provided to each individual is very difficult. The policy to be fol- 263 lowed would be to have patterns of prices which, vary for different types of consumers who are responsible for differ ent cost levels. On a trial and error basis some fair ap proximation to true average costs could be achieved. An other qualification involves the meeting of non-economic goals. The discussion of transportation regulation legisla tion emphasized the variety of goals which have been inser ted in regulation policy. These included such diverse goals as economic efficiency, contribution to national defense, promotion of particular modes of transport especially water carriers, promotion of depressed Industries and contribution to an adequate postal system. To the extent non-economic goals are included, variation from the general average-cost pricing would be required. Promotion of some particular interest might require a lower pattern of prices to promote the specific Interest, lower rates to a particular region to aid in its industrial development. Prices higher than average cost might be required to support capacity in excess of normal needs but required for emergencies, ©_.g_. support of capacity available for defense purposes. The particular deviations v/ould be dependent on the number and nature of the non-economic goals. A major reason for deviation from average-cost pricing also seems justified for the encouragement of innovation. Encouragement of innovation through introduction of new techniques, new services, extension of markets and more 264 dependable provision of service would seem to be a very de sirable postulate of public utility regulation. To the ex tent that provision for temporary retention of profits from innovation adds to innovational incentives, a justifiable deviation from average-cost pricing exists. The gains from innovation could be retained by the firms temporarily, sub ject to slow decrease as the advantages are passed to con sumers. A plan similar to the New Jersey or Washington plans might furnish a more certain and stable guarantee. Some of the advantages of average-cost pri-Ging were mentioned in Chapter IX. Assuming Imperfections in the economy, the optimum price would be greater than marginal cost. If it is not known whether marginal- or average-cost pricing would more closely approach the optimum, one could on the basis of insufficient reason assume equal probability of greater contribution to the optimum on the part of either marginal- or average-cost pricing. It is certain that aver age-cost pricing would eliminate the more extreme monopoly restrictions. The convenience arises for several reasons. No subsidy and its attendant tax are required. Average cost is more amenable to ex post verification. Average-cost pricing seems to be in accord with general standards of jus tice, as witness the Supreme Court cases dealing with pric ing. finally average cost is a concept which is easily un derstood, as compared with marginal cost, by regulatory com missions and the public. 265 The personnel or facilities available for regulation have received no attention in this stucfer. It is just this, however, which is likely to assure greatest success of pub lic utility regulation. A set of intelligent commissioners having intimate knowledge of the regulated industries, cog nizant of the relative importance of the various possible non-economic goals and having the facilities of the know ledge and continuing study of an adequate staff probably is of the greatest Importance in successful regulation. This is particularly true with respect to special consideration of innovation. A commission and staff with intimate know ledge of the regulated Industries, continually studying the changing conditions and sensitive to the peculiar circum stances of particular situations would be able to distin guish returns in excess of average costs due to innovation and exceptional efficiency from extra returns resulting from restriction. Encouragement of the former and the discour agement of the latter would be the required policy. This emphasis on the commission point up the fact that most economic policy is more of an art than it is an exact science subject to slide rule or electronic computer manip ulation of data according to exact, simple rules or for mulas. Most economic policy, Including public utility regu lation deals with problems Intimately related to personal passion and prejudice and the greatest of economically im portant intangibles, incentives. As in most economic poli 266 cy, the success of public utility regulation depends upon a high degree of tact, statesmanship and Intuitive awareness of circumstances not readily subject to empirical verifica- ti on . The question now arises as to the place of the econo mist in the analysis of public utility regulation. The ana lytical implements available to the economist enable him to make judgments of the economic implications of any regula tory policy or action. Action can be judged as to the eco nomic gains or losses. This information is available, then, to those concerned. Whoever may be concerned, whether the public, the commission or the public utility firm, can then decide as to the relative effect of economic gains or losses as compared with any other gain or loss of a non-eco nomic nature. In this the economist can be quite valuable in shedding light on the implications of a possible decision by making more information available. Further contribtition to public utility regulation can be made by the economist through his analysis of the feasibility of certain means contributing to the achievement of chosen ends. As the implementation of the economic goal of optimal utilization of resources rests on the value judgment that such optimal utilization is desirable as compared with other possible goals, preparing for or fighting a war or sub sidization of fig-growers for the esthetic contribution of fig orchards, the economist cannot demonstrate the supremacy 267 of the economic goal. The economist, however, can Indicate possible economic losses which would occur with the achieve ment of some non-economic goal which has economic aspects. The final decision as to the priority of the economic or non-economic goals rests with whatever machinery is avail able for determination of the ranking priority of the goals. Such machinery usually depends on the methods available for political expression. Even in the case of conflicting or contradictory goals, the economist cannot demonstrate any case for consistency as a fundamental goal. Such contradiction or Inconsistency can be pointed out by the economist. The effect of such incon sistency can be demonstrated. But the economist cannot set up consistency as an inalterable, fundamental goal any more than any other goal, be it economic, political or religious. As a matter of fact, it might be argued that the right to inconsistency is a fundamental human right and that the most Interesting aspect of life involves the compromise or reso lution of conflicting goals. These limitations of marginal-cost pricing and of the economist, however, do not cast marginal-cost pricing into oblivion as a standard for the optimum. Nor does it deprive the economist of his functions. The economist can and must use marginal-cost pricing as the standard for economic eval uation. The losses arising from imperfections or existence of non-economic goals in markets can still be pointed out 268 and marginal cost is basic to the estimation of these losses. Information is then available for judgment whether present losses due to the imperfections are worth potential gains which may be achieved. The worth of the non-economic goals also can be balanced against the economic losses. To the extent that economists, accountants and statisticians can devise methods for the more accurate measurement of mar ginal cost, they will contribute to the achievement of not only economic goals but also non-economic goals. A series of studies of specific public utility industries with the aim of determining the feasibility of marginal-cost pricing, or pricing at the weighted average of the ratio of price to marginal cost, in these particular industries would seem to be a contribution to public utility regulation. Studies of Innovation in public utility industries and the conditions conducive to such innovation would also be helpful in pro viding guides to the effective regulation of public util ities. i APPENDIX A Derivation of Marginal Conditions of Maximum Economic Welfare**- ^"The derivation follows Oskar Lange, ojo. cit., pp. 224- 228. It is assumed that each individual possesses some util ity function TJ1 I u M x ^ , xj-j, . x^) where the superscript i_ stands for the i_th individual and the x ’ s include all commodities (inputs and products) which the individual pos sesses. The individual’s utility is measured by some arbi trary index. It also is assumed that the social welfare function is monotonically increasing with respect to indi vidual utilities. Each individual (firm) will have a transformation func tion fi (y^, y^, ...., y^) = 0 , where the superscript i_ stands for the ith individual and the s Include all the commodities (inputs and products) which the individual (firm) transforms. For the economy as a whole the amount of the rth com modity possessed must equal the amount transformed and simi larly for all commodities, i_.e_. m m 22 x - r - H J r . (p = 1* 2 > . ...J n). 1=1 1=1 For each individual, however, this need not be true as exchange may take place or gifts may be made. 269 270 Xt will be necessary to maximize irMx^, x|, x^) (i Z 1, 2, m) subject to constraints u k (xk , x]|, . x^) z constant, (k z 1, 2 , ....,i- 1 , 7 2 * . y^) = 0 , and i/l, ...., m) m m 2 1 - X y^ = 1 > 2 , •.. ., n) . izl 1 = 1 With the use of Lagrangean multipliers, the maximi zation problem becomes m m n m m ZX.ui / z y±f / x vr( z 4 = Z y£) izl 1 i - 1 r=l 1 = 1 r izl where - 1 for all i_* s successively. Differentiating and eliminating the Lagrangean multi pliers gives three sets of equations. In the equations only the marginal utilities for each individual are compared. Between individuals it is the ratios of marginal utilities or marginal rates of substitution which are compared. 5 uA 8 u^ axi ax^ (r and s = 1 , 2 , n) (1 ) " 9^r (i and k z 1 , 2 , ...., m) ax 1 ax£ d f1? 8fk t 1 ^ k ayf r / s (2 ) - 5 5 ? S i S 554 Syg Prom (1) 1. The marginal rates of substitution between any two products (r and s) are the same for any two individuals (i and k) consuming both. Prom (2) 2. The marginal rate of transformation be twee-* any two products (r and s) must be the same for any firms (i and k) producing both. Prom (2) with ya an input and yr a product 3. The marginal rate of transformation between any input (y3 ) and any product (yr ) must be the same for any two firms (i and k) using the input and producing the product. Prom (2) with yr and ys as inputs 4. The marginal technical rate of substitution between any two inputs (yr and ys ) must be the same for any two firms (i and k) using both to produce the same product. Prom (3) 5. The marginal rate of substitution between any pair of products (r and s) for any person (i) consuming both must be equal to the marginal rate of transformation (for the community as represented by k) between them. 272 From (3) where yr is a product and yg is an input 6. The marginal rate of substitution between the rate of* remuneration (measured in r) and the time spent aiding in the production of* a product must be equal for any input owner (i) to the marginal rate of transformation between such input time and the product. From (1), (2) and (3) when utility functions and trans formation functions are assumed to contain intertemporal relations, commodities yr and ys are assumed to refer to the same commodity at different moments of time (t^ and t2 ) • 7. From (3): The marginal rates of substitution of a com modity between two moments of time (ys at t]_ and yr at tg) must be equal to the marginal rate of transformation of the commodity between the two moments of time. . From (1): The marginal rate of substitution of a com modity between two moments of time (ys at t^_ and yr at tg) must be the same for any two individuals (i and k). From (2): The marginal rates of transformation of a commodity between two moments of time (ys at t^ and yr at tg) must be the same for any two firms (i and k). SELECTED BIBLIOGRAPHY I . Books Allen, R. G. D., Mathematical Analysis for Economists, (London: Macmillan & Co., 1949) . Arrow, K. J. , Social Choice and Individual Values, (Hew York: John Wiley and Sonsj 1951). Bauer, John, Transforming Public Utility Regulation, (New York: Harper & BrosTT 1950T7 Baumol, W. J #, Welfare Economics and the Theory of the State, (Cambridge: Harvard University Press, 19527^ Beckwith, B. P., Marginal-Cost Price-Output Control, (New York: Columbia University Press, 19557^ Bigham, T. C. and M. J. Roberts, Transportation, Prin- ciples and Problems, (New York: McGraw-Hill Book Co., 1952). Boulding, Kenneth E., Economic Analysis, (New York: Harper & Bros., 1948"JT ______, The Organizational Revolution, (New York: Harper & Bros.,1953). Chamberlin, Edward H., The Theory of Monopolistic Compe- titlon, (Cambridge: Harvard University Press, 1948). Clark, J. M., Social Control of Business, (New York: McGraw-Hill Book Co., 1939). ______, Studies in the Theory of Overhead Costs, (Chicago: Universi ty of Chicago Press'^ 1923). Clemens, E. W., Economics and Public Utilities, (New York: Appleton-Century-Crofts, 1950). Committee on Price Determination, Cost Behavior and Price Determination, (New York: National Bureau of Economic Research, 1943). Dewey, R. L., The Long and Short Haul Principle of Rate Regulation, '(Columbus: The Ohio State University Press, 1935). 273 274 Douglas, Paul H., The Theory of Wages, (New York: The millan Go., 193477 Harrod, R. P., Toward A Dynamo, c Bconomi cs, (London: Macmillan & Co., 1949). Hicks, J. R., The Theory of Wages, (London: Macmillan & Co., 1932). ______, Value and Capltal, Second Edition, (Oxford: Oxford University Press, 1946). Knight, P. H., Risk, Uncertainty and Profit, (New York: Houghton Mifflin CoT~, 1921) . Lerner, A. P., The Economics of Control, (New York: The Macmillan Co., 1946). Little, I. M. D., A Critique of Welfare Economics, (Oxford: Oxford University Press, 195077 Locklin, D. P., Economics of Transportation, (Chicago: Richard D. Irwin, Inc., 1954). Marshall, Alfred, Principles of Economics, (New York: The Macmillan Co77 1949 ) . Pigou, A. C., The Economics of Welfare, (London: Macmillan & Co., 1920). Reder, M. W., Studies in the Theory of Welfare Economics, (New York: Columbia University Press, 1947). Robinson, E. A. G., The Structure of Competitive Indus- try, (Cambridge: Cambridge University Press, 1950). Robinson, Joan, The Economics of Imperfect Competition, (London: Macmillan & Co., 1933). Samuelson, Paul A., Foundations of Economic Analysis, (Cambridge: Harvarl University Press, 194877 Schumpeter, J. A., History of Economic Analysis, (New York: Oxford University Press, 1954). Scitovsky, Tibor, Welfare and Competition, (Chicago: Richard D. Irwin, Inc., 1951). Smith, Adam, The Wealth of Nations, ed. by Edwin Cannan, (New York: Random House, 1937). 275 Stigler, George J., The Theory of Price, (New York: The Macmillan Co., 1947'). Triffin, Robert, Monopolistic Competition and General Equilibrium Theory, (Cambridge: Harvard University Press, 1940). Troxel, Emery, Economics of Publlc Utllities, (New York: Rinehart & Go~, 1947) . U. S. Steel Corporation, TNEC Papers, (New York: U. S. Steel Corp., 1940). Welch, Francis X., Cases on Public Utility Regulation, (Washington: Public Utility Reports, 1946). II. Journal Articles Alchian, A. A., "The Meaning of Utility Measurement," American Economic Review, Vol. XLIII (March, 1953), pp. 26-50. Baumol, W. J., "Economic Theory and the Political Scientist," World Politics, Vol. VI (January, 1954), pp. 266-277. Bergson, A., "A Reformulation of Certain Aspects of Wel fare Economics," Quarterly Journal of Economics, Vol. LII (February, 1938), pp. 310-334. ______, "Socialist Economics," A Survey of Contem porary Economics, Volume I, ed. by H. S. Ellis^ (Phila delphia: The Blakiston Co., 1949), pp. 412-448. Boulding, K. E«, "Welfare Economics," A Survey of Contem porary Economics, Volume II, ed. by B. F. Haley, (Chicago: Richard D. Irwin, Inc., 1952), pp. 1-38. Cassels, John W., "On the Law of Variable Proportions," Explorations in Economics, (New York: McGraw-Hill Book Co., 1936), pp. 223-236, reprinted in Readings in the Theory of Income Distribution, ed. by William Fellner and B. F. Haley, (London: George Allen & Unwin Ltd., 1950), pp. 103-118. Coase, R. H., "The Marginal Cost Controversy," Economioa, Vol. XIII N.S. (November, 1946), pp. 169-182. 276 ______, "The Nature of the Firm," Economica, Vol. IV U.S. (T937), pp. 386-405, reprinted in Readings in Price theory, ed. by George J. Stigler and Kenneth E. Bouldihg, (Chicago: Richard D. Irwin, Inc., 1952), pp. 331-351. ______, "Price and Output Policy of State Enter prise: A Comment," Economic Journal, Vol. LV (April, 1945), pp. 112-113. Dewey, Ralph L., "Transportation Act of 1940," American Economic Review, Vol. XXXI (March, 1941), pp. 15-26. Dixon, F. H., "The Interstate Commerce Act as Amended," Quarterly Journal of Economics, Vol. XXI (November, 1906'), pp. "22-51. Ezekial, M., "The Cobweb Theorem," Quarterly Journal of Economics, Vol. LII (1937-38), pp. 255-280. Frisch, R., "The Dupuit Taxation Theorem," Econometrica, Vol. VII (April, 1939), pp. 145-150. Glaeser, Martin G., "The United States Supreme Court Re deems Itself," Journal of Land and Public Utility Eco nomi cs, Vol. XVIII (May, 1942), pp. 146-154. Gray, Horace, "The Passing of the Public Utility Con cept," Journal of Land and Public Utili ty Economics, Vol. XVT (February, 1940), pp. 8-20. Harbeson, R. W., "A Critique of Marginal Cost Pricing," Land Economi cs, Vol. XXXI (February, 1955), pp. 54-74. Henderson, A. M., "Consumer's Surplus and the Compensat ing Variation," Review of Economic Studies, Vol. VIII (February, 194-1), pp. 117-121. ______, "The Pricing of Public Utility Under- takings," Manchester School, Vol. XV (1947), pp. 223- 250. Hicks, J. R., "Foundations of Welfare Economics," Econom ic Journal, Vol. XLIXZ (December, 1939), pp. 696-712. Hotelling, H., "The General Welfare in Relation to Prob lems of Taxation and of Railway and Utility Rates," Sconometrica, Vol. VI (July, 1938), pp. 242-269. 277 James, E. J., "Federal Regulation of Railways," Fubllca- tlons of the American Economic Association, Vol. II (July, 18877, p p . "246-295. » Kaldor, N ., "The Equilibrium of the Firm," Economic Journal, Vol. XLIV (March, 1934), pp. 60-76. ______, "Welfare Propositions in Economics," Economic Journal, Vol. XLIX (September, 1939), pp. 549-555. Lange, 0., "The Foundations of Welfare Economics," Econometrlca, Vol. X (1942), pp. 215-228. Lerner, A. P., "Statics and Dynamics in Socialist Eco nomics," Economic Journal, Vol. XLVII (June, 1937), pp. 253-2707 Lewis, W. A., "Fixed Costs," Economica, Vol. XIII N.S. (November, 1946), pp. 231-258. Little, I. M. D., "A Reformulation of the Theory of Con sumers* Behavior," Oxford Economic Papers, New Series, No. 1 (January, 1949), pp. 90-99. Long, Clarence D., "The Labor Force and Economic Change," Insights Into Labor Issues, ed. by Richard A. Lester and Joseph Shis ter, (N ew York: The Macmillan Co., 1948), p p . .329-355. Nelson, J. C„, "The Motor Carrier Act of 1935," Journal of Polltlcal Economy, Vol. XLIV (August, 1936), pp. 464-504. Norris, Harry, "State Enterprise and Output Policy and the Problem of Cost Imputation," Economica, Vol. XIV N.S. (February, 1947), pp. 54-62. Rich, E. J., "The Transportation Act of 1920," American Economic Review, Vol. X (September, 1920), pp. 507-527. Robinson, Joan, "Rising Supply Price," Economica, Vol. XIII N.S. (1941), pp. 1-8, reprinted in Readings in Price Theory, ed. by Kenneth E. Boulding and George J. Stigler, Cchicago: Richard D. Irwin, Inc., 1952), pp. 233-241. Ruggles, Nancy, "Recent Developments in the Theory of Marginal Cost Pricing," Review of Economic Studies, Vol. XVII (1949-1950), pp. 107-126. 278 ______, "The Welfare Basis of the Marginal Cost Pricing Principle," Review of Economic Studies, Vol. XVII (1949-1950), pp. 2CT46. Samuelson, Paul A., "Comment," A_ Survey of Contemporary Economics, Volume II, ed. by B. F. Haley, (Chicago: Richard D. Irwin, Inc., 1952), pp. 36-38. Schumpeter, J. A., "The Analysis of Economic Change," Review of Economic Statistics, Vol. XVII (May, 1935), pp. 2-10, reprinted in Readings in Business Cycle Theory, ed. by Gottfried Ilaberler, (London: George Allen & Unwin Ltd., 1950), pp. 1-19. Schumpeter, J. A., "The Creative Response In Economic History," Journal of Economic I-Ilstory, Vol. VII (Novem ber, 1947), pp. 149-159, reprinted in Essays of j. A. Schumpeter, ed. by Richard V. Glemence"^ (Cambridge: Addison-Wesley Press, 1951), pp. 216-226. Scitovsky, T., "A Note on Welfare Propositions in Eco nomics," Review of Economic Studies, Vol. IX (November, 1941), pp. 77-88. ______, "A Reconsideration of the Theory of Tariffs,'" Review of Economic Studies, Vol. IX (1942), pp. 89-110. Troxel, Emery, "Limitations of the Incremental Cost Pat tern of Pricing," Journal of Land and Public TJtliity Economics, Vol. iXX (February, 1943), pp. 28-39. Viner, Jacob, "Cost Curves and Supply Curves," Zeit- schrift fur N a ti onaloko nomle, Vol. Ill (1931), pp. 23- 46, reprinted in Readings in Price Theory, ed. by Ken neth E. Boulding and George J. Stigler, (Chicago; Richard D. Irwin, Inc., 1952), pp. 198-226. Wallace, D. H., "Joint and Overhead Cost and Railway Rate Policy," Quarterly Journal of Economics, Vol. XLVIII (August, 1934")”, pp. 583-619. Wilson, T., "Price and Outlay Policy of State Enter prise," Economic Journal, Vol. LV (December, 1945), pp. 454-461. 279 Working, E. J., "What Do Statistical ’Demand Curves' Show?,f Quarterly Journal of Economics, Vol. XLI (1926-1927), pp. 212-235, reprinted in Readings In Price Theory, ed. by Kenneth E. Boulding and George J. Stigler(Chicago: Richard D. Irwin, Inc., 1952), pp. 97-115. III. Government Publications Federal Coordinator of Transportation, Freight Traffic Report, (Washington: 1935). Federal Coordinator of Transportation, Passenger Traffic Report, (Washington: 1935). Federal Coordinator of Transportation, Railway Traffic Organlzatlon Report, (Washington: 1935). Federal Coordinator of Transportation, Report on Freight 1 Gar Pooling, (Washington: 1934). Malott, E. 0., The Hoch-Smith Resolution, Agricultural History Series no. 4, TWashington: U. S. Bureau of Agricultural Economics, 1942). Temporary National Economic Committee, Economic Stan dards of Government Prlce Control, (Monograph No. 32, 76th Congress, 3rd Session^ 1941). 280 IV. Legislation "Clayton Act of 1914,” 38 Statutes at Large, part I, pp. 701-703. "Elkins Act of 1903,” 32 Statutes at Large, pp. 847-849. ’’Emergency Railroad Transportation Act of 1933,” 43 Statutes at Large, part I, pp. 211-221. "Hepburn Act of 1903,” 34 Statutes at Large, part I, pp. 584-595. "Hoch-Smith Resolution," 43 Statutes at Large, part I, pp. 801-802. * "Intercoastal Shipping Act of 1933,” 47 Statutes at Large, part I, pp. 425-427. "Interstate Commerce Act of 1887," 24 Statutes at Large, pp. 379-387. "Mann-Elkins Act of 1910," 36 Statutes at Large, part I, pp. 539-557. "Motor Carrier Act of 1935," 49 Statutes at Large, part I, pp. 543-567. "Panama Canal Act of 1912,” 37 Statutes at Large, part I, pp. 560-569. "The Shipping Act of 1916," 39 Statutes at Large, part I, pp. 728-738. "The Transportation Act of 1920," 41 Statutes at Large, part I, pp. 456-499. "Transportation Act of 1940," 54 Statutes at Large, part I, pp. 898-956. "Valuation Act of 1913," 37 Statutes at Large, part I, pp. 701-703. V. Supreme Court Cases Ann Arbor R, R. Co. v. United Staces, 281 U. S. 658 (1930). Chicago, Burlinton and Quincy Railroad Company v. Iowa, 94 U. S. 155 (187777 !' 281 Dayton-Goose Creek Railway Company v. United States, 263 U. S. 456 (1924J, Federal Power Commission et. al. v. Hope Natural Gas Co., 320 U. S. 591TT944"). German Alliance Insurance Co. v. Kansas, 233 U. S. 389 TT9T47T Interstate Commerce Commission v. Union Pacific R. R. ~~To77“ 222 tJTS.' 5Ti“(T§T£T: ' “ Market Street Railway Co. v. Railroad Commission of UaXTfornia, 324 U. S. 548 (1945). Munn v. Illinois, 94 U. S. 113 (1877). Nebbla v. New York, 291.U. S. 502 (1934). New State Ice Comp any v. Llebmann, 285 U. S. 262 (1932). Peik v. Chicago and Northwestern Railway Company, 94 U. S. 164 (1877T Ribnik v. McBride, 277 U. S. 350 (1928). Smyth v. Ames, 169 U. S. 466 (1898). Sunshine Anthracite Coal Company v. Adkins, 310 U. S. 381"" (1940JI Tyson & Brother v. Banton, 273 U. S. 418 (1927). Williams v. Standard Oil Oompan.y of Louisiana, &7Q U. S. 240 (1928T7” Wolff Packing Company v. Court of Industrial Relations of ' Kansas, 2 6 2 U. S. 522 (1923). AUTOBIOGRAPHY I, Thomas Anton Martinsek, was born in Cleveland, Ohio, July 15, 1918. I received my secondary school education in the public schools of Maple Heights, Ohio. My undergraduate training was obtained at Cleveland College, Western Reserve University, from which I received the degree Bachelor of Arts in 1948. Prom The Ohio State University, I received the degree Master of Arts in 1949. In January, 1950, I received an appointment as Graduate Assistant In the Depart ment of Economics at The Ohio State University. I held various positions from Graduate Assistant to Instructor at The Ohio State University until June, 1955, and have been an Instructor at Worth Carolina State College from September, 1955, while completing the requirements for the degree Doctor of Philosophy.