FILE COPYI DAVID F. CAVERSRS, DAVID AND AND } lt AND NELSONJAMES R. NELSON JOHlNS mr ca HOPKIN.S ~vie $6.OO Electric Power Regulation in Latin America BY DAVID F. CA1 VERS AND JAMES R. NELSON The economic developmeilt of Latin America depends to a grcat extent on a healthy ancd expanding electric power industry. For this reason, the WVorlcl Bank ancl the United Nations Economic Commissionl for Latin Amer-ica have collaborated in Llte pub- lication of this book, wvhichi was wr-ittell by LtVo clistinguished experts on the basis of a stLdly by the Harvard Law ScIhool of Lthe lawvs, pracLice, and adminiistrative systemiis govcrning the operation of electric power utilities in Latin America. Particular empihasis wvas placed in the study on the problem of Finanicinig the expansion of these utilities. (continutled On back flap) CN.., ELECTRIC POWER REGULATION IN LATIN AMERICA DAVID F. CAVERS Fessenden Professor of Law, Harvard University JAMES R. NELSON Charles E. Merrill Professor of Economics, Amherst College Electric Power Regukation in Latin America BASED ON a study conducted for the Harvard Law School by MARVIN S. FINK and STAFF PUBLISHED FOR The International Bank for Reconstruction and Development AND The United Nations Economic Commission for Latin America BY THE JOHNS HOPKINS PRESS, Baltimore @1959, The Johns Hopkins Press, Baltimore 18, Md. Distributed in Great Britain by Oxford University Press, London Printed in the U.S.A. by Franklin Printing Company, Primos, Pa. Library of Congress Catalog Card Number 59-14231 Foreword In presenting Electric Power Regulation in Latin America, the International Bank for Reconstruction and Development and the secretariat of the United Nations Economic Commission for Latin America have collaborated in the publication of an economic study for the first time. This collaboratipn is the re- sult of the common interest both institutions have in the economic development of the Latin American countries and their recognition of the importance to such development of an expanding electric power industry. In addition, the Interna- tional Bank has a particular interest in the problems of the Latin American power industry arising from its substantial investment in electric power expansion in the area. This led the Bank to request the Harvard Law School to sponsor a study of the Latin American laws and administrative systems governing the operation of electric utilities, with particular reference to the problem of financing their capital expansion. This book has been prepared on the basis of that study. Neither the International Bank nor the secretariat of the Commission for Latin America has reviewed the recommenda- tions of the book in detail. The recommendations are there- fore to be read as representing the views of the authors, and not necessarily those of the two institutions. However, both institutions believe that the recommendations, formulated as they were only after an extensive field investigation and a care- ful analysis of the findings of that investigation, deserve full and serious consideration. Moreover, both institutions strongly v vi Foreword endorse the over-all conclusion of the book that in the electric power industry in Latin America, whether public or private, the cost of the electricity services should be fully met by tariffs and that these should be adjusted to assure the proper func- tioning of the industry and the attraction of capital for expan- sion purposes. Many of the regulatory systems in Latin America were estab- lished under economic conditions which no longer exist. Clearly these systems might usefully be reconsidered in the light of the electric power industry's present need to accumu- late and attract large amounts of capital, and such a re-exami- nation is in fact already in progress in several Latin American countries. The Bank and the secretariat of the Commission hope that the present volume, although it sets forth in detail only one of several possible approaches, will prove useful to all those responsible for regulatory policy. Preface Soon after the problems of economic restoration and develop- ment emerged througlhout the world at the end of World War II, it became apparent that a pervasive problem would be the provision of adequate supplies of electric power. Rap- idly growing cities with rising standards of living, rapidly ex- panding industries with new technologies that depended on electricity, these quickly combined to absorb the surplus gen- erating capacity of the thirties and to call insistently for new construction. This situation, which was especially marked in Latin Amer- ica, led to the making of many engineering studies and many economic estimates of need and projections of future growth. The magnitude of the problem was made plain and, with this, its seriousness, especially for those capital-deficit nations which were striving to achieve more prosperous, more balanced econo- mies by attaining a greater degree of industrialization. To realize this aim, these countries needed more capital and more electric power, yet they soon were faced by the fact that no great industry requires more capital in proportion to its output than does electric power. It also became evident that many of the engineering and economic projections for the expansion of electric power were not going to be realized because the capital needed for this purpose would not be forthcoming. This disappointment seemed inescapable in Latin America. There, paradoxically in view of the mounting demand for electricity, the electric vii viii Preface power industry seemed a sick industry, its private companies depressed and unable to raise new capital, its government- owned enterprises sustained by subsidies at the expense of the taxpayer or, more commonly, by inflation-stimulating forms of governmental credit. Yet the electric power industry was supposedly protected from the vicissitudes of the market. Re- gardless of its ownership, it was a regulated industry. Could it be that the regulatory requirements and policies to which it was subject were in any substantial degree responsible for the industry's inability to rise to its post-war opportunity? This question was posed by an inquiry which the Interna- tional Bank for Reconstruction and Development addressed to the Law School of Harvard University in the fall of 1953. The Bank asked whether the School would be prepared to undertake a study of the regulation of the electric power in- dustry in Latin America, accepting as established the prospect of continuing growth in demand but examining whether the industry's apparent difficulties were inherent in its situation or might measurably be eased by changes in regulatory poli- cies and practices. The Bank undertook to provide funds to defray the cost of the proposed inquiry. The proposal was received by the Harvard Law School at a time when its program in International Legal Studies was already taking shape. That program had been born of the realization that, in the post-war world, the field of legal scho]- arship could no longer be confined to the traditional subjects of domestic law, coupled with Public and Private International Law. A need was felt for centers to study those common prob- lems with significant legal aspects which the processes of eco- nomic, social and political readjustment were everywhere bringing to the fore. From the study of such problems, in the light of the differing laws and experiences of representative countries, it was hoped that bodies of expert knowledge relat- ing to them could be created and made available, not only to other centers of study, but also to persons charged with the responsibility of shaping public and business policies. Preface ix The World Bank's proposal for a study of the regulation of the electric power industry in Latin America afforded an op- portunity to embark on the study of a problem of the sort that the Law School had envisaged. Indeed, it was peculiarly ap- propriate for it lent itself to another objective of International Legal Studies, as these had been conceived at Harvard: the co- operation of legal scholars with scholars in the other social disciplines. To deal adequately with the problem the Bank presented, it was clear that no appraisal of simply the legal mechanisms of regulation would suffice. It would be essential to examine the respective regulatory systems in the light of the economic needs they confronted and the economic conse- quences their policies and practices produced. In view of these considerations, the Harvard Law School undertook to carry out the study which, it was hoped, could be completed in two years or so. Early in 1954, it enlisted the services of Marvin S. Fink, LL.B. Harvard, 1939, to direct the study. Mr. Fink was then a member of the legal staff of the United States Securities and Exchange Commission. He had engaged in extensive work on problems arising under the Pub- lic Utility Holding Company Act and was therefore well ac- quainted with regulatory and financial problems of the electric industry in the United States. To assist Mr. Fink on a full-time basis, we retained a member of the Department of Public Util- ities and Transportation of the New York University School of Commerce, Radford L. Schantz. Mr. Schantz had been trained in engineering as well as economics. We were fortunate in securing the services, on a part-time basis, of Professor James R. Nelson of the Department of Economics of Amherst College, who had earlier made an extensive study of the British electric industry and who had served in war-time on economic missions in Latin America. As Associate Dean then engaged in the development of International Legal Studies and as a teacher of Public Utility Law, I undertook the general super- vision of the study on behalf of the Harvard Law Faculty. Preliminary studies made it clear that it would not be prac- x Preface ticable to cover all countries of Latin America. An adequate base was thought to be provided by Brazil, Chile, Colombia, Costa Rica, and Mexico. After a thorough study of the materi- als available in the United States, Mr. Fink arranged to spend some time in each country for the purpose of securing addi- tional needed data and to obtain a first-hand acquaintance with the situation existing there. Accordingly, accompanied by Mr. Schantz, he spent a number of weeks in Brazil and Chile in the fall of 1954. Early in 1955, accompanied by Professor Nelson, Mr. Fink visited Mexico and then went on alone to Costa Rica and Colombia. In each country, Mr. Fink's experience was the same: he was welcomed cordially by officials of the government and public power authorities and by executives of the investor-owned power companies, whether their ownership was domestic or foreign. Mr. Fink was able to engage in candid discussions of policies and practices and their consequences. He was given access to and obtained large quantities of unpublished data and secured an illuminating view of the practical working problems of both regulation and management. Subsequent to these journeys, Mr. Fink was able to supplement the informa- tion thus obtained by correspondence and by visits to the home offices in the United States and Canada of major holding com- panies with Latin American subsidiaries. The work of Mr. Fink and his associates yielded so much relevant material that they were faced by an embarrassment of riches. To have reported all the pertinent developments and data fully for each of the countries studied would have re- sulted in five volumes of reports, not one. Moreover, a pattern of common problems and difficulties was seen to be emerging which was highly suggestive of corrective measures that might be employed in most, perhaps all, of the countries studied. Accordingly, it was decided to prepare a summary report on each country and an over-all report directed to the central problems that "the country studies" had revealed, drawing upon the latter for basic supporting material. Preface xi Even the undertaking thus circumscribed proved to be be- yond the resources available to us. Before the exhaustion of the financial support which had been contributed by the Bank, Mr. Fink had completed the five country studies, had formu- lated, in broad outline, the corrective measures which he be- lieved were indicated by his findings, and had developed in early draft a report presenting these measures in their legal and economic context. Since, upon the termination of his service as director of the study, Mr. Fink accepted a position on the legal staff of a company with large interests in the Latin Amer- ican electric industry, it was agreed that he would be disso- ciated completely from the further development of the report. Fortunately Professor Nelson of Amherst College agreed to con- tinue his association with the study, and together Professor Nelson and I have prepared the present volume. As the time elapsed since the completion of Mr. Fink's study in the fall of 1956 suggests, this undertaking by two pro- fessors with full-time responsibilities in addition to it has not been easy to dispatch. A much more extensive presentation of the economic setting of the problem, the financing difficulties of the industry, and the prevailing rate levels and structures was completed by Professor Nelson in 1957 while, at the same time, I completed a brief exposition of the regulatory systems and a draft of the second part of the report, revising and de- veloping in more specific terms the suggestions outlined by Mr. Fink. The full draft report was circulated to a number of informed readers for criticism. Many helpful comments were received on specific points in both parts of the report, from which we have sought to profit in subsequent revisions. How- ever, we also encountered a consensus that, for maximum use- fulness, the volume should be materially shorter, especially the first part depicting the economic background, a work that had been rendered less important by the recent publication of that excellent study, Energy in Latin America, by the United Nations Economic Commission for Latin America. Professor Nelson undertook to rewrite Part I of the report xii Preface in substantially shorter compass, and, in the course of thor- ouglhly reworking his material for this purpose, he was able to bring the data on many points closer to the current figures. We have concluded regretfully that it will not be practicable to bring the country studies similarly up-to-date; although they speak on the basis of data only through 1953, they are not with- out utility in their present state, and we are planning to pub- lish each in multilithed form for limited distribution. The delays entailed in rewriting, recirculating, and again revising the report have been less serious in terms of the per- tinence of the suggestions it advances than one might have ex- pected. We have, of course, been able, on a number of matters, to introduce more recent data than were available to Mr. Fink, but at many points the report speaks as of 1953 and 1954. How- ever, the general pattern presented in the report continues and in some cases the problem is aggravated. Though the generat- ing capacity of the electric power industry has been steadily expanding in the countries studied, it is still not a healthy in- dustry, and the gap between demand and supply still remains wide. The ambitious plans entertained for narrowing that gap are embarrassed, moreover, in most of the countries studied, by sharply increased inflationary pressures. As we believe the report makes plain, these pressures are what have rendered the prevailing regulatory policies and practices-which were car- ried over from the deflationary, sluggish economy of the thir- ties-a dangerously negative influence on electric power indus- try's ability to respond to its post-war opportunity. The suggestions for corrective measures that we have ad- vanced are predicated on two postulates: (1) that the electric power industry in Latin America, whether government- or privately-owned, ought to be self-supporting; and (2) that policies and practices in regulating electric rates should be ef- fective even if and where inflationary pressures will continue. We hope that such pressures will soon subside. If this should happen, the measures we have suggested to counteract the effects of inflation, would not become harmful; they wvould Preface xiii simply not be called into play. Indeed, they would continue as a salutary source of reassurance. The measures we have put forward are designed with a prac- tical end in view: to make the "cost-of-service" principle of rate regulation actually work, even in inflationary economies. We have not attempted in formulating our suggestions to key them to the situations existing in each-or any-one of the countries we have studied. We do not feel equipped to pro- pose specific answers to the particularized problems of indi- vidual countries, and we have not presumed to do so. Yet we have tried to make our suggestions as specific as that limitation would permit. We hope that those persons in government and industry whlo lhave the responsibility of guiding governmental and industry policies may judge our general approach to the problem to be a sound one and find that a number of our specific suggestions point to helpful lines of concrete action. If these hopes can be realized in the countries we have stud- ied, we feel confident that our report and its suggestions will have value in other Latin American republics. We know that the situations prevailing in many of them are not dissimilar to those in the five countries we selected for intensive study. Moreover, though the problems of determining rates for elec- tric power and of financing the industry's expansion in less developed countries outside the Western Hemisphere do not arise within the regulatory framework that is common to the United States and much of Latin America, nevertheless, we believe that our report will have relevance for them as well. That the price of electricity should meet the industry's cost of service and that this requires special provision against infla- tion are ideas that are germane to the economic problems of all parts of the world, however much their pertinence may be concealed by institutional arrangements or their acceptance be deterred by conflict with traditional thinking. In various respects the suggestions advanced in this report go beyond the pattern of regulatory law and practice in the United States which has served as a model for many of the xiv Preface Latin American regulatory systems. Some of the ideas we have put forward are already in use in the United States; others could, we believe, be taken over to some advantage. But, in general, the steadily expanding electric power industry in the United States is already self-supporting to a degree not realized in the countries we have studied, and, it may be added, with rate levels on the average materially higher. Moreover, thoughl provision is not made in many states of the United States against the impact of inflation in the industry, the gradualness of that ailment's progress here and the abundance of capital seeking investment have thus far prevented the industry's fi- nancing troubles, endemic in Latin America, from becoming widespead or serious in the United States. While, therefore, we believe our report may be of interest to readers concerned only with the electric industry of the United States, we urge that its suggestions be considered by them with clear recognition of the very significant differences between the economy of the United States and the economies of the Latin American re- publics. David F. Cavers Harvard Law School Cambridge, Massachusetts Table of Contents FOREWORD V PREFACE Vii CHAPTER I THE PURPOSE AND NATURE OF THE STUDY 1 PART I PROBLEMS OF THE INDUSTRY CHAPTER II THE INDUSTRY AND ITS SETTING 11 1. Historical Development of the Industry 12 2. Present Organization of the Industry: The Shift to Public Ownership 16 3. The Changing Environment of the Industry 19 CHAPTER III FINANCING ELECTRICAL EXPANSION 27 1. The Private Financing of Privately-Owned Electrical Suppliers 29 A. The Brazilian Experience 29 B. The Mexican Experience 32 C. Present Prospects for Private Financing in the Countries Studied 37 2. The Financing of Government-Owned Suppliers of Electricity 39 A. Brazil: Conflicting Objectives and Divided Powers 41 B. Colombia: Municipal Ownership and "Public Service" 45 C. Chile: The Experience of Endesa 47 xv xvi. Table of Contents D. Mexico: The Comision Federal de Electricidad (CFE) 50 E. Costa Rica: The Instituto Costarricense de Electricidad (ICE) 55 3. Special Government Financing Measures 57 A. "Mixed" Corporations 57 B. Development Banks 60 C. Special Taxes 64 4. Loans by International Lending Agencies 66 CHAPTER IV RATES 69 1. The Trend of Electricity Prices in Latin America 71 2. Comparative Electrical Rates in 1954 77 3. Electricity Rates and Costs: Latin America and the United States 83 A. Relative Costs of Capital 83 B. Relative Efficiency in Generation and Technical Progress 85 C. Costs per Kilowatt for New Generating Capacity 87 D. Load Factors 89 E. Customer Costs 90 4. Rate Structures 93 A. Rates for Industrial Power 93 B. Residential Rate Structures 97 C. Electricity Rate Structures and Social Policy 108 CHAPTER V RATE REGULATION IN THE ELECTRIC INDUSTRY 11 0 The Rate-Making Process 112 The Controversy over the Allowable Return: The Rate Base 114 The Rate of Return 117 The Significance of "Leverage" 120 The "Dividend Pay-out Ratio" and Retained Earnings 122 Table of Contents xvii Income Taxes 123 Operating Expenses and "Regulatory Lag" 124 Depreciation 127 Foreign Exchange 129 PART 2 SUGGESTIONS FOR CORRECTIVE ACTION CHAPTER VI GENERAL CONSIDERATIONS 135 The Problem of Capital Expansion 135 The Cost-of-Service Principle and Factors Impeding Its Application 141 The Corrective Measures Proposed and the Problem of Timing 147 CHAPTER VII AUTOMATIC COMPENSATORY ADJUSTMENTS 150 1. Operation and Maintenance Expense 156 2. Taxes 158 3. The Cost of Capital: Adjustments for Changes in Foreign Exchange Rates 159 CHAPTER VIII MEASURES FOR MEETING THE COST OF CAPITAL: THE INVESTMENT BASE 161 1. A Critique of the Traditional Approaches to Determining the Cost of Capital 162 2. A Suggested Alternative to the Traditional Approaches to the Cost of Capital 168 3. How the Suggested System Might Be Instituted and Operated 175 CHAPTER IX MEASURES FOR MEETING THE COST OF CAPITAL: THE RATE OF RETURN 181 1. Determining the Rate of Return on "Equity" for the Privately-Owned Utility 183 2. Determining the Rate of Return on "Equity" for the Government-Owned Utility 189 xviii Table of Contents CHAPTER X HANDLING THE PROBLEMS OF DEPRECIATION 200 Three Functions of Depreciation 201 Depreciation and "Fair Value" 202 Depreciation and Original Cost 202 The Depreciation Reserve and the Rate Base 203 A Suggestion for Separating Depreciation Accounting from the Calculation and Adjustment of the Investment Base 206 Adjusting Depreciation Expense and Reserve for Price Changes 210 CHAPTER XI SUGGESTIONS FOR OBTAINING NEW CAPITAL FOR EXPANSION 213 1. Securities to Protect the Investor Against Inflation 215 A. Cost-of-Living and Cost-of-Electricity Securities 216 B. Securities Payable in Dollars or with Obligations Measured by the External Value of the Domestic Currency 218 C. Convertible Debentures 219 D. Participating Preferred Stock 220 E. Securities Issued by Municipal Enterprises 222 2. Potentialities for Institutional Investment in Electric Utilities 223 3. Greater Participation by Industry in Electric Utility Financing 226 Joint Ownership of Electric Plants 231 Lease-back of Generating Station 234 4. Consumer Contributions to Capital Through Construction Surcharges 235 CHAPTER XII A SUMMARY OF SUGGESTIONS FOR CORRECTIVE ACTION 244 Table of Contents xix APPENDIX POST-WAR USE OF INDEXED BONDS FOR FINANCING PUBLIC UTILITY EXPANSION IN FRANCE 259 1. Index Based on Gold Value of the Franc 261 2. Participating Bonds or Debentures, with Index Based on Gross Revenues 262 3. Bonds Price-Indexed as to Interest and Principal 263 4. Bonds Price-Indexed as to Principal Only 264 INDEX 269 LIST OF TABLES Table 2.1. Latin America and the Rest of the World: Cumula- tive Annual Rate of Increase in Per Capita Consumption of Electricity 20 Table 2.2. Per Capita Consumption of Public Utility Elec- tricity: Brazil, Chile, Colombia, Costa Rica, Mexico 22 Table 2.3. Cumulative Average Annual Rates of Urban and Rural Population Growth, Latest Inter-censal Period, Brazil, Colombia and Mexico 24 Table 2.4. Percentage Increases in Industrial Output, and in Industrial Use of Electricity 25 Table 2.5. Brazil: Rank of Selected Manufacturing Industries, by Rate of Growth, 1939-49, and by Ratio of Value of Elec- tricity Used to Value Added by Manufacture, 1949 26 Table 3.1. Average Rates of Return on Net Worth Brazilian Corporations, 1948-52 30 Table 3.2. Brazilian Capital Issues for New Money, 1948-52 31 Table 3.3. Ratios of Profit to Invested Capital in Mexico, by Industry, 1953 and 1954 33 Table 3.4. Mexican Light & Power Co.-Rates of Return, 1950-56 35 Table 3.5. Chile: Empresa Nacional de Electricidad, S. A., Selected Financial Data, 1953-1956 48 Table 3.6. Chile: Empresa Nacional de Electricidad, S. A., Selected Financial Data Corrected for Price Changes, 1953- 1956 49 xx Table of Contents Table 3.7. Mexican Public Surpluses and Deficits, 1936-51 51 Table 3.8. Mexico: Federal Electricity Commission, Invest- ment and Gross Operating Profits, 1938-1953 53 Table 3.9. Costa Rica: Distribution of Central Government Internal Funded Debt, April 30, 1954 55 Table 4.1. Monthly Residential Electrical Bills in the United States and in Five Latin American Countries, 1929 and 1939 72 Table 4.2. Monthly Electrical Bills for 100 kwh Residential Consumption Selected Latin American Supply Areas, 1939 73 Table 4.3. Indexes of Trend of Real Electrical Rates, 1939= 100 75 Table 4.4. Official and Purchasing Power Parity Exchange Rates, Brazil, Chile, and Colombia, 1954 78 Table 4.5. Monthly Electrical Bills in Brazil, Chile, Colombia and Costa Rica, in U. S. Dollar Equivalents, mid-1954 78 Table 4.6. Monthly Electrical Bills in Mexico, 1954, in Dol- lar Equivalents 80 Table 4.7. Estimates Weighted Average Electricity Bills, 1954, Five Latin American Countries Relative to the United States 81 Table 4.8. Capital Costs per Kilowatt of New Generating Ca- pacity in Selected Latin American Countries 1954 87 Table 4.9 Cost of Electricity in Relation to Value Added by Manufacture and Gross Value of Manufacturing Output 94 Table 4.10. Ratio of Cost of Electricity to Value Added by Manufacture, Brazil and Mexico 95 Table 4.11. Expenditure on Electricity as a Percentage of Total Expenditure by Specified Types of Residential Con- sumers 98 Table 4.12. Residential Rate Increases, Cia. Chilena de Elec- tricidad, 1931-1954 104 Table 4.13. Medellin, Colombia: Residential Consumption and Assessed Value of Residences, October, 1951 107 Table 4.14. Small Residential Consumers, Medellin, Colom- bia, and Cia. Chilena de Electricidad 108 Table A-1. New Issues through the French Capital Market. 1950-1954 260 ELECTRIC POWER REGULATION IN. LATIN AMERICA CHAPTER I The Purpose and Nature of the Study In the rapidly expanding economies of Latin America, Asia, and Africa, one of the impediments to economic develop- ment and social betterment has been the shortage of electric power. Since the end of World War II, there has been an unprecedented expansion of the electric power industry in the less-developed countries of the world. Despite this, existing facilities have continued to lag behind the increasing demand. Yet such expansion of the electric power industry as has been possible has entailed huge capital investments, both by the privately-owned segment of the industry and by the govern- ments concerned. In order to finance this program, the amounts available for financing expansion from earnings and depreciation allow- ances have not been large in relation to the need. A complex of factors has deterred new investments in the industry by private investors. Thus, the low and uncertain return in the electric power industry as compared with returns available in unregulated enterprises has inhibited private local investment. The level of return available, plus the uncertainty of remitting earnings, has deterred private foreign capital from investing in the industry. Under these circumstances, electric power enter- prises have had to rely more and more for new local capital I 2 The Purpose and Nature of the Study on public funds, sometimes the proceeds of excise taxes but more often the result of extension of government credit. The chief sources of foreign funds to finance expansion in the case of both private and public enterprises have been the Interna- tional Bank for Reconstruction and Development ("World Bank") and the United States Export-Import Bank of Wash- ington ("Eximbank"). The existence of these international credit institutions and their continued willingness to devote a part of their resources to finance electrical development mean that the problem of achieving adequate, sound programs for the domestic financ- ing of the electric power industry is a manageable one, not that the problem does not exist or may safely be ignored. These institutions are mainly sources of capital for foreign ex- penditures and not for the locally-incurred costs which can readily amount to half or more of the total cost of an electric installation. Moreover, though their resources are substantial, the calls on them are many. Not only would it be unrealistic to suppose that they could carry the full burden of financing electric expansion in the less-developed nations, but also, in determining the share of their resources that should be devoted to electric expansion, both the banks and the borrowing coun- tries will doubtless wish to consider the many other domestic uses to which these funds might be put, undertakings which, unlike the electric industry, cannot hope to be self-supporting. The need for domestic capital by the electric power indus- try in the less-developed countries will continue to grow with the continued growth of the national economies. Except in the immediate future, therefore, the problem of obtaining funds is not one of emergency financing; long-term policies are called for. If the power-producing enterprises in a country, whether publicly'or privately owned, cannot attract domestic capital or are deterred from doing so by political and economic uncer- tainties, then the government must assume the burden of fur- nishing new capital to the industry, whether or not it con- structs the new power facilities itself. Adoption of that course The Purpose and Nature of the Study 3 places electric power in competition for the government's limited capital resources with schools, roads, harbors, sanitation and other non-income-producing needs of a developing economy. Further, unless governments are to finance the necessary ex- pansion by inflationary methods, they must obtain a larger share of the nation's private savings, either through the sale of government securities or through increased taxation. Electric utility expansion need not, however, be a stimulant to inflation. A government may determine that it is in its na- tional interest for electric power enterprises to be self-support- ing. In that case, the industry, whether publicly- or privately- owned, must be allowed to earn such rates as will (1) permit internal generation of cash through appropriate depreciation accruals and retained earnings to aid sub- stantially in financing expansion; and (2) enable it to attract fresh capital from investors, not once but repeatedly. Whether the electric industry in the less-developed coun- tries is now and can remain self-supporting in this sense de- pends at least in part on the existing regulatory laws governing the industry and on their administration. It may also depend on the adoption of new measures by the governments and the industry which will aid the latter in accumulating and attract- ing the capital it needs. Difficulty in obtaining sufficient capital to finance the nec- essary expansion of the electric power industry has been a post-war problem in practically all countries. Thus, in 1953, the Organization for European Economic Cooperation in de- scribing the problem of OEEC countries stated: Since the war, the financing of capital expenditures in the electricity sector seems to have been affected mainly by the following factors: (a) The higher cost of living and heavier taxes have re- duced private savings and, consequently, the supply of money available from all sources for investment; (b) Owing to the currency devaluations which have oc- curred in various O.E.E.C. countries and, in some 4 The Purpose and Nature of the Study cases, threaten to continue, private investors in the countries concerned are reluctant to finance long- term projects. As electricity supply undertakings must mainly rely on long-term loans, they have been par- ticularly affected by this cautious attitude on the part of the investor; (c) The economic policy adopted by governments in sev- eral countries has hampered the rise of electricity prices and so difficulties have arisen in some cases in raising fresh loans for new plant.' Conditions such as these are revealed in this study of the regulation of the electric power industry in Latin America. Therefore, special emphasis has been given to the industry's problems in accumulating and attracting sufficient capital to meet the pressure upon it for expansion. Although this pres- sure is to be found in most rapidly developing, capital-defi- cient countries, the need to keep the study within manageable dimensions has led not only to its being restricted to Latin America but also to its being focussed on five Latin American countries: Brazil, Chile, Colombia, Costa Rica, and Mexico. These were selected for intensive analysis in the belief that the conclusions which might be drawn from their study would prove to be pertinent to conditions in other countries. The selection of the Latin American countries as the field of study was influenced, in part, by the great importance in Latin America of recent and prospective electric powver devel- opment; the extent to which loans of the World Bank and Eximbank to these countries have been for electric power; the similarity in regulatory practices to those followed, at one time or another, in the United States; and the diversity in the eco- 1 OEEC, The Price of Electricity and Its Relation to Investmien7t in the Elec- trictly Supply Industry (Paris, Oct. 1954) p. 5. A study recently published by OEEC, The Trend of the Selling Price of Electricity and its Relation to the Financing of New Plou71t (Ilaris, Nov. 1958), r'eveals that thie problems reportecd in the above quota1tio1 wCrC still persisting in certain of the OEEC counitries. In its "Conclusions," the stutly states (p. 23): "110. The dificculty of raising capital has been felt auiiily in coutitries where electuicity undertakinigs are not yet au- thorized to sell at a price which covers not only productioni costs proper, but all other charges as well, including a fair rettirn on the capital invested." The Purpose and Nature of the Study 5 nomic-environments which afford a basis for comparison with other areas. However, in the years that have intervened since the selection was made, inflationary pressures appear to have been stronger in four of the five countries selected than in Latin America as a whole. This report, largely on the basis of the individual country studies, directs attention to a succession of problems common, in varying degrees, to virtually all electric enterprises under regulation in the less-developed countries. The report proffers no panacea for these problems. Rather, it undertakes to out- line a number of methods to which governmental agencies and privately-owned electric power companies may turn when confronted by the difficulties perplexing the electric power in- dustry in rapidly expanding economies. In outlining these methods and analyzing the problems to which they are ad- dressed, no distinction is made between government and privately-owned enterprises except to the extent that particular problems relate chiefly to the one or to the other. The ques- tion whether the electric power industry is to be owned pri- vately or by governmental entities is a matter of broad national policy which must take account of national and local condi- tions. This report takes no position on that issue. In the course of the report, references are made, from time to time, to comparable practices or results in the United States. This is due to the fact that the existing regulatory practices in the countries studied are commonly patterned after those ex- isting in the United States at the times when the particular laws were enacted. It must be recognized, however, that any such comparison must be in the light of the significant differ- ences between a relatively mature economy such as that of the United States and the rapidly expanding economies of coun- tries in the earlier stages of economic growth. For example, conditions existing in a country with ample capital available for long-term investment at fixed rates of return, in private as well as governmental securities, afford a dubious guide for establishing standards in a country where the amounts avail- 6 The Purpose and Nature of the Study able for such investment are limited and capital markets are poorly developed. Another basic point of difference is that which exists between a capital-importing and a capital-export- ing country. The most important manifestation of these differences from the standpoint of the expansion of the electric power industry is the rate of return an enterprise must be allowed, if it is to raise new capital. In any country, this rate of return is also a function of the risk involved. In the United States, experience has demonstrated that the electric power industry is one of low risk resulting from high stability of earnings; therefore the rate of return allowed may also be relatively lowV. Unhappily, in the countries studied regulatory policies ill-suited to econo- mies characterized by price instability have rendered electric power an industry of high risk relative to return. In most of these countries, the task of executing an adequate electrical expansion program has been made difficult by steady, rapid urbanization and industrialization. These processes have continuously enlarged the demand for electric power while simultaneously increasing the demand for capital for invest- ment. Yet, in the same countries, the electric industry's access to the investment capital it has needed to keep abreast of the mounting demand has been denied by the dogma that low rates for electrical energy are essential to industrial develop- ment. Not only has the electric industry been unable to earn returns high enough to attract domestic capital, but at the same time it has been prevented by low depreciation rates from accumulating much capital internally. The electric industry's difficulties have been compounded in some of the countries studied by serious and progressive in- flation, resulting in persistent increases in the industry's costs. The existing machinery for electric rate regulation has not been geared to providing rate increases promptly in response to these cost increases. As a consequence, even the low return on investment which the electric rates, as established, had al- lowed has seldom been realized for more than brief periods of The Purpose and Nature of the Study 7 time. Where inflationary pressures have been accompanied by deterioration in a country's exchange position, the constantly rising levels of domestic prices have been matched by con- stantly rising local-currency costs for foreign exchange. These have increased the cost of imported fuel and equipment and the expense of remitting payments to the holders of such of the industry's securities as are payable in foreign currencies. Can an objective system of regulation under law be designed that will permit the electric industry to overcome these handi- caps and to achieve its rightful position as a healthy, self-sus- taining segment of a growing national economy? This will not be easy. It cannot be achieved by simply applying patchwork amendments to old laws and temporary adjustments to emer- gency situations. Since the industry's chronic ailments stem from low rate levels, low depreciation allowances, and slow rate adjustments, the need to raise electric rates in the process of setting the industry on its feet seems plain. But the question may be asked whether the rate increases that are required would impose a serious handicap on Latin American econo- mies in their efforts to achieve a more rounded development and a higher standard of living? Our study discloses four cir- cumstances that should quiet concern on this score. First, the electric rates now charged by most public utility suppliers in the countries studied are startlingly low in com- parison to those prevailing in the United States, and their level is not to be explained in terms of cost advantages. Second, the revenues derived by each system from the very large number of small residential consumers whom it supplies constitute a very small proportion of the system's total reve- nues. Accordingly, if, as a matter of social policy, proportionate rate increases were not imposed on consumers in this category, the resulting increase in the burden on all other consumer categories would be a slight one. Third, shortages in supply and the erratic quality of electric service in many communities have led to the widespread in- stallation of diesel generators for private supply, thereby en- 8 The Purpose and Nature of the Study tailing costs for electricity well above the level which public utilities would have to charge after the needed upward adjust- ment in their rates. Fourth-to state a fact that runs counter to widespread pop- ular belief-the percentage which the cost of electricity bears to the value added by manufacture in all but a very few indus- tries (e.g., metallurgy and paper) is so small that even electric rate increases that seemed large percentage-wise would require no significant increases in the prices of manufactured goods. The last two points may be restated in terms of one of the conclusions reached in a 1958 OEEC study of The Trend of the Selling Price of Electricity and its Relation to the Financ- ing of New Plant. "If a price policy which prevented elec- tricity undertakings from raising funds for development were to lead to a shortage of power, it would place a much heavier burden on economic activity than would the increase in in- come which electricity undertakings could obtain by bringing their prices into line with their costs."2 It seems clear that Latin America's need is for more elec- tricity, not for cheap electricity. 2 Id., p. 24. PART I Problems of the Industry CHAPTER II The Industry and Its Setting The present-day problems of the electric industry in Latin America must be examined both in the perspective of the in- dustry's past and in its social and economic context. This chap- ter will pursue that objective, although it will undertake only to note those major trends affecting the industry that bear most directly on the subject of our study. A brief retrospective look at the industry's development suffices to reveal significant shifts in the sources of its capital. Thus, in its rapid early growth, the industry outran the re- sources of domestic private investment, and foreign private investment furnished the capital for the large-scale expansion that was required. Then came an interlude when capacity was more than adequate to meet demand, when the main problem was to enlarge the market for power. However, when the great upsurge of demand came after World War II, neither domestic nor foreign private capital could be attracted to the industry in the volume needed. Public sources of capital were resorted to, and the capital thus supplied was largely put at the disposal of public authorities. This process, as will appear, has not succeeded in satisfying the requirements of a multiplying population and a growing industry. Indeed, the fast pace of urban growth would entail 11 12 Part 1. Problems of the Industry a large program of electrical expansion simply to keep abreast of rising residential requirements. Latin American industry is also growing. Moreover, it is turning increasingly to products and methods of production that use more electrical energy. Nevertheless, despite these tendencies, in recent years, the growth in per capita use of electricity has been less in Latin America than in any other major area among the non-com- munist economies. 1. Historical Development of the Industry Local private capital was responsible for founding the vast majority of public utility suppliers of electricity in Brazil, Chile, Costa Rica, and Mexico. Colombia is the sole exception to this rule among the five countries studied. In terms of num- bers of electricity suppliers, local private ownership is still dominant in Brazil and important, though decreasingly so, in Chile and Mexico. But the percentage of total output con- tributed by local privately-owned companies is very small in every country studied. The occasion for the first appearance of large-scale foreign investment also varied from area to area. Santiago, Chile, is typical of many large cities whose first need for such invest- ment was to finance unified ownership and operation of urban electric transit facilities which supplied their own power. In the Mexico City area, the requirement of a large investment to tap a low-cost but remote power source was a more impor- tant factor than the transit problem. In the Sao Paulo and Rio de Janeiro areas of Brazil, possibilities of low-cost hydroelec- tricity and a need for large-scale transit electrification com- bined to produce the immense electrical utility, whose corpo- rate title-the Brazilian Traction, Light & Power Company -illustrates the importance of electric traction to the develop- ment of the larger urban systems. The Industry and Its Setting 13 The development of electricity supply systems to serve large areas has been much slower in Latin America than in the United States. Mexico had several area suppliers even before World War II: Mexican Light & Power Company ("Mex- light"), which sold electricity to a number of mines as well as to customers in Mexico City; the Torreon-Chihuahua system of the American & Foreign Power Company, with a scattered mining and irrigation load; the same company's Guanajuato system, emphasizing mining but also serving industry and with some irrigation load; and its Puebla-Veracruz system, origi- nally built up to carry an industrial and traction load. In Brazil, Chile, and Costa Rica, area supply was confined to the countryside around the metropolis. In Colombia, no really large supply areas existed. Mines in Brazil, Chile, and Colom- bia generally produced their own electric power, unlike many of the more important mines in Mexico. Even the larger locally-owned companies in Brazil and Chile generally served, as they still serve, scattered cities and towns instead of compact and integrated geographical regions. Various subsidiaries of the American & Foreign Power Com- pany have already been mentioned. Although none of these subsidiaries is individually as large as Brazilian Traction, taken together they form the most extensive group of electricity- supply enterprises under common private ownership in Latin America. In the countries studied, they include undertakings supplying Recife, Salvador (Bahia), Porto Alegre, Belo Hori- zonte, and other important cities in Brazil, the Santiago-Val- paraiso area in Chile, a long list of important cities in Mexico, Barranquilla and numerous smaller Colombian cities, and the San Jose district in Costa Rica. Other subsidiaries operate in Cuba, Guatemala, Panama, Venezuela, Ecuador, and Argen- tina. American & Foreign Power's Latin American properties were originally acquired by the Electric Bond & Share Com- pany in the late 1920's and early 1930's. Many of the American & Foreign Power's subsidiaries were acquired from other for- 14 Part 1. Problems of the Industry eign investors, usually British. As an illustration: the present Compafiia Chilena, the American & Foreign Power subsidiary in Chile, was acquired from a British holding company which in turn represented a link in a historical chain of ownership which runs back to an original German investment. The largest of the privately-owned Mexican electric systems, Mexlight, is controlled by the Belgian Sofina interests, was incorporated in Canada, and has its head offices in Toronto. Brazilian Traction was from the beginning, and still is, a Canadian corporation independent of any holding company. As of December 31, 1955, 68.89% of its ordinary shares were registered. Almost two-thirds of these shares, or 42.49% of the total, were held in Canada. Just over one-fifth, or 14.91% of the total, were held in the United States, and roughly one- sixth, or 10.78% of the total, were held in the United King- dom.' The International Power Company, another holding company with important Latin American subsidiaries in Mex- ico and in countries not included in this study, is a Canadian corporation with head offices in Montreal. The Argentine in- dustry has not only included several American Sc Foreign Power subsidiaries but also companies which have drawn their equity capital from a number of Western European countries. The importance of foreign private ownership in the electric industry in Latin America may be traced back to the original importance of foreign technical assistance. This appears to have been especially true in the case of Canadian investment; Ca- nadian investors could provide technical assistance in hydro- electric development based on Canada's extensive early ex- perience with water power. Capital shortages were another important factor. Foreign private investors appeared in areas where the supply of domestic capital was insufficient to finance large hydroelectric projects or to meet the high initial costs of transit electrification. During the 1930's, this flow of capital was greatly reduced. The generating capacity which had been provided was ample 1 Brazilian Traction, Light R Power Co., Ltd., Annual Report, 1955, p. 6. The Industry and Its Setting 15 in most areas; the need was for greater utilization of existing facilities. Moreover, public utility holding companies in the United States, many in the throes of reorganization, were in no position to raise new capital for any type of investment, domestic or foreign. Ability to earn profits in Latin America was no longer accompanied by any assurance that these profits could be transferred abroad at anything like the rates of ex- change prevailing when the investment was first made. In ad- dition, all of the countries studied were moving in the direction of stricter regulation of public utility prices and profits during the 1930's. Certain regulatory principles which were gaining ground in the United States in this period appear with con- siderable uniformity in the regulatory systems being established in the Latin American countries studied: in particular, the idea of limiting earnings to the maximum permitted rate of return on a "rate base" expressed in original local-currency cost. By the end of World War II, power shortages were appearing in most of the larger cities in the countries studied. Capital assets had not been readily obtainable during the war; popu- lation and industry had expanded, and were still expanding. Foreign companies could neither satisfy the demand for ex- pansion from earnings nor sell adequate amounts of new se- curity issues. Companies owned by local private capital usu- ally had even poorer prospects for raising substantial sums of money. Moreover, the objective of retaining the hydro re- sources of a nation within the public domain through the creation of publicly-owned hydroelectric generating enter- prises had a strong public appeal, as was the case in certain regions of the United States. Therefore the burden of expan- sion was shifted to public bodies. 16 Part 1. Problems of the Industry 2. Present Organization of the Industry: The Shift to Public Ownership The growth of publicly-owned suppliers has been very marked in all of the countries studied since 1945: a. Brazil had little public ownership, aside from small mu- nicipal plants, prior to 1945. By the end of 1953, only 6% of all new generating plant completed was owned by the major public suppliers. But by that time generating capacity under construction or in an advanced stage of planning was to be over 50% publicly-owned. Once these projects have been completed, over a fourth of all utility generating plant in Brazil will be publicly-owned. Public ownership in Brazil involves both the Federal Gov- ernment and the States. At the end of 1955, the Paulo Afonso project (CHESF) in Northeastern Brazil, largely financed by the Federal Government, had 180,000 kw installed and a further 180,000 kw under construction. Seven other public cor- porations owned by six different States had a further 150,000 kw installed and 489,000 kw under construction. None of this capacity was in existence at the end of World War 11.2 b. Chile had no significant public ownership when in 1943 the Corporaci6n de Fomento de la Producci6n adopted its first national electrification plan. The Empresa Nacional de Elec- tricidad, S.A. (ENDESA) was organized as a subsidiary of the Corporaci6n de Fomento in 1944 and entrusted with the pub- licly-financed share of Chilean electrical development.3 By the end of 1953, Endesa already owned 48.3% of the public utility 2 Economic Commission for Latin America (ECLA), Energy in Latin America (UN, Dept. of Econ. & Soc. Affairs, Geneva 1957) Sales No. 1957. II. G. 2, Table 69, p. 89 (hereafter cited as "ECLA, Energy in Latin America") . For the Spanish language edition of the same work, see CEPAL, La Energia en America Latina (No. de venta: 1957. II. G. 2.). 3 For a report of this development and the present state of the industry as well as plans for its development, see Endesa, Plan de Electrificacion del Pais (2d ed. 1956). The Industry and Its Setting 17 generating capacity of the country. Public ownership had ac- counted for 207,000 kw, or 88%, of the 235,000 kw expansion of generating plant in the eight years 1946-53. In the period of Endesa's second plan, 1953-64, it expects to finish 103,720 kw of generating capacity which was still under construction at the end of 1952, and to provide an additional 453,970 kw. Endesa assumes that privately-owned electricity suppliers will add 181,000 kw over this period.4 The Compafifa Chilena de Electricidad is supposed to provide most of this privately- financed expansion. Its earnings record in recent years makes this forecast of its ability to expand seem optimistic. c. In Colombia, public ownership has always been very im- portant. Since World War II, the municipality of BogotA has bought out private investors in the company supplying that city with electricity, and the municipality of Cali has taken over the distribution of electricity from the Compafiia Colom- biana de Electricidad, an American &c Foreign Power subsid- iary. Three of the four largest cities-BogotA, Medellin, and Cali-now have municipal supply of electricity, while Barran- quilla is served by the Compafiia Colombiana. Colombian electrical development in, the last ten years has emphasized joint financing of generating plant by the central government, departments, and municipalities. Some of these new "Centrales" also distribute electricity. At the end of 1953, total installed capacity of Colombian public service plants was 362,000 kw, owvned by 581 different enterprises. Only 33 of these had ca- pacity of 1,000 kw or more, and the municipal enterprises of Medellin and Bogota alone accounted for 45.2%/ of national capacity. The Compafifa Colombiana, the largest privately- financed supplier in Colombia, provided an additional 13.9%. d. Costa Rica had some municipal ownership before World War II, but by far the largest electrical enterprise was the Compafifa Nacional de Electricidad, a subsidiary of American & Foreign Power. The Instituto Costarricense de Electrici- dad (ICE), financed by the central government, had only 4 Id., Tables 79, facing p. 252, and 80, p. 258. 18 Part 1. Problems of the Industry 10,000 kw in operation by the end of 1955, or 18% of national capacity, but it had a further 42,000 kw under construction.5 e. In Mexico, most publicly-owned facilities are concen- trated in the hands of the Comisi6n Federal de Electricidad (CFE) and the Nueva Compafiia Electrica de Chapala. At the end of 1955, the CFE accounted for 31% of national generat- ing capacity as a result of an increase from 44,000 kw in 1946 to 604,000 kw, while the Chapala company provided a further 5% of the national capacity.6 The 552,000 kw which the CFE had under construction at the end of 1955 almost equalled its total expansion over the previous decade. Although Mexlight was not far short of the CFE in generating capacity, with 490,095 kw at the end of 1955, it was also by far the largest customer for the output of the CFE. Certain features of this trend toward public ownership de- serve special attention: (i) Most of the funds required for expansion of public ownership have been provided by governments. The contribu- tion from gross earnings has been small. The enterprises have raised almost no funds from private investors. Central govern- ments have made all the public contribution in Chile, Costa Rica and Mexico; in Brazil and Colombia, political subdivi- sions have been important contributors. (ii) The emphasis of public expansion has been on hydro- electric power: almost exclusively so in Chile and Brazil, largely so in Colombia. Although the percentage of thermal plant installed has been quite high in Costa Rica and Mexico, some of this has consisted of emergency diesel installations. Only Mexico places much emphasis on thermal plant in its long-run public planning. Hydroelectric plant is generally more expensive in terms of initial capital costs than thermal plant, though it may require no more foreign exchange and, in fuel-deficit countries, will not require foreign-exchange ex- penditures for fuel. 5 ECLA, Energy in Latin America, Table 69, p. 80. 6 Ibid. The Industry and Its Setting 19 (iii) Public ownership has proceeded much further in gen- eration than in distribution. Publicly-generated electricity is normally carried through to the final consumer only in Colom- bia. Elsewhere privately-owned companies still have the re- sponsibility for providing practically all of the capital required for expansion of distribution systems. In addition, the vertical split in ownership raises new and important questions of the relationship between wholesale and retail prices for electricity. (iv) Although a preference for public over private owner- ship goes far to explain the original establishment of many of the publicly-owned suppliers, especially where the exploitation of natural resources is involved, the rapidity of their growth seems primarily attributable to the acuteness of the power shortages which had arisen in the absence of private financing. Objectives other than power, such as valley development, have not been common. Although irrigation occasionally has been coupled with power, multi-purpose projects of the TVA type are few. 3. The Changing Environment of the Industry The five countries studied differ in their rate and type of economic change: with respect to "industrialization," for ex- ample, Chile probably started first, Brazil and Mexico came next, Colombia is following in turn, and Costa Rica is last. In terms of other attributes, such as percentage of literacy or real income per capita, Costa Rica is at or close to the top of the list. This section will discuss types of economic change which most directly affect the supply of, and demand for, electricity. These may be subdivided into "real" and "monetary" factors. The latter involve problems of money supply, government budgets, private savings, and other general magnitudes bearing on the supply of funds for investment in electricity. These 20 Part 1. Problems of the Industry factors will be taken up as part of the financing problem dis- cussed in Chapter III. The present section discusses three as- pects of "real" economic change that are particularly important in affecting thie demand for electricity: energy use in general and industrialization and urbanization in particular. Energy use. Consumption of electricity in Latin America has increased relative to consumption of energy from other sources, and the rate of increase in Latin American total energy con- sumption has been higher in the last ten or twenty years than in any other non-communist area. But, to quote the Economic Commission for Latin America: While total per capita consumption of energy in Latin America during recent years grew at a higher rate than in the rest of the world, the contrary was true of electric energy. The rate of increase in Latin America was not only lower than the world average, but was the lowest of all.7 In view of the rapid economic development of Latin America since 1929, increases in per capita electricity consumption, as shown by comparative statistics, are strikingly low (see Table 2.1). Table 2.1. Latin America and the Rest of the World: Cumu- lative Annual Rate of Increase in Per Capita Con- sumption of Electricity 1929-37 1937-49 1949-54 Latin America ....................... 3.7% 4.8% 5.6% Western Europe ..................... 4.5 2.9 8.7 United States ........................ 4.9 6.1 7.7 Other-except USSR, China, Korea .... 6.8 3.8 8.8 Total world, except USSR, etc .... 4.0% 4.1% 8.7% Source: ECLA, Energy in Latin America, Table 16, p. 32. 7 ECLA, Energy in Latin America, p. 32 (Italics supplied). The Industry and Its Setting 21 Per capita gross product in most of Latin America has in- creased rapidly in the last ten or twenty years, and, as the Economic Commission for Latin America points out, elec- tricity has a high income elasticity of demand.8 Therefore per capita electricity consumption should have increased at least as fast in Latin America as in the rest of the world. Paradoxi- cally, however, it has, instead, increased more slowly, except for 1937-49. The explanation is not hard to find. The demand for elec- tricity in Latin America has been held in check by inadequacy of supply-especially of public utility supply. This lag in supply has in turn been connected with the outstanding eco- nomic attribute of the industry: its high ratio of capital in- vestment to gross or net value of annual production. Electric- ity is a capital-intensive industry. Therefore, and because it has had to expand rapidly, its claim on investment resources has grown far faster than its contribution to current income. For example, in Mexico, in 1950, electricity and gas accounted for 0.5% of net national product yet absorbed 8%/, of all in- vestment; in Chile, in 1950, the comparable proportions were 0.66%, and 7.01 %.9 The difference between electricity's output contribution and capital needs is also very marked within the energy sector it- self. The Economic Commission for Latin America estimates total investment in the development of energy resources at $875 million in 1955, or about 10% of gross national invest- ment for all Latin American countries combined. Electricity which provided 20% of the energy supply alone was expected to account for 60% of this investment. 10 Urbanization has been an outstanding feature of Latin American economic development over the last generation. 8 "Income elasticity of demand" is a measure relating proportionate changes in the consumption of any one commodity to proportionate changes in per capita real income. A high income elasticity of demand for electricity woould indicate a faster rate of growth for electricity consumption than for per capita real income. 9 ECLA, Energy in Latin America, Table 66, p. 86. 10Id., p. 14. 22 Part 1. Problems of the Industry This is particularly important for this study since public elec- tricity supply in Latin America has been and is almost entirely urban supply. A pattern of low total agricultural consumption, very few agricultural users, and high consumption per agri- cultural user is repeated in all the supply areas for which data are available.'" The vast mass of the rural population has no access to commercially-distributed electricity. Usually they could not afford to buy electricity if it were available. They could use it for productive purposes only in connection with a revolution in agricultural organization and techniques.12 This almost complete absence of electricity in rural life has led the Economic Commission for Latin America to classify all consumption of electricity, except industrial consumption in the narrow sense, as "urban use." The ECLA statistics for the countries included in this study are summarized in Table 2.2. Table 2.2. Per Capita Consumption of Public Utility Elec- tricity: Brazil, Chile, Colombia, Costa Rica, Mexico Per Capita Consumption of Rate of Annual Increase in Efectricity (kwh) Per Capita Consumption Indus- Indus- Year trial Urban Period trial Urban Total Brazil ............ 1952 80 84 1939-52 6.8% 7.5% 7.1% Chile ............ 1950 58 155 1939-50 5.9 6.3 6.2 Colombia .1952.. 1952 38 67 1939-52 12.7 7.6 9.6 Costa Rica . 1951.. 1951 18 129 1947-51 6.5 6.3 6.3 Mexico .1952.. 1952 82 102 1939-52 3.1 8.1 5.5 Source: ECLA, Energy in Latin America, Table 39, p. 51. 11 In Mexico, for example, rural electrification is more important than in any other country studied. Agriculture used 6.9% of all the publicly-supplied elec- tricity consumed in the country in 1952. But agricultural users numiibered only 4,900, or 0.39% of total consumers, and the average agriculttural consumer took 49,501 kwh, a level of consumption pointing clearly to a large irrigationi demand. (Data from unpublished CEE-Mex study.) 12 Rural prosperity is at least as great, and as general, in Costa Rica as in any country studied. Yet, in 1950, only 432 of Costa Rica's 41,058 farms used any mechanical power, and 82.4% of the farms reported no animal or mechanical power whatever. Costa Rica, Censo Agropecuario de 1950, Table 54, p. 134. The Industry and Its Setting 23 This table shows that electrification is at least as closely con- nected with the "urbanization" aspects of economic develop- ment as with its "industrialization" aspects. The rate of in- crease in urban use has been, on the average, about as great as the rate of increase in industrial use. From the standpoint of future investment needs, urban use bulks still larger than the percentages would indicate for every country but Brazil, be- cause the urban rate of increase must be applied to a much larger base. Moreover, many urban uses-especially residential and small commercial uses-far exceed industrial uses in their capital investment requirements per kilowatt-hour sold, and in their additional capital investment requirements per addi- tional kilowatt-hour sold. Finally, urbanization in Latin Amer- ica has not been a simple by-product of industrialization: the Federal District (Rio de Janeiro) in Brazil still had a larger population than Sao Paulo in 1950, even though Sao Paulo is much more important as a manufacturing center. And the Colombian Department of Antioquia (Medellin) contained 25.9% of all Colombian manufacturing personnel in 1945, and the Department of Cundinamarca (Bogota) only 20.5%. Yet BogotA's 1951 population was 638,562, and Medellin's only 328,294. Since Table 2.2 expresses growth in electricity consumption in terms of average per capita consumption for the entire coun- try, it does not reflect the very important contribution to de- mand for electricity of urban growth relative to national growth. These disparities in rate of growth between urban and rural populations are shown in Table 2.3. These average growth rates of 4% to 5% per annum for the cities of Brazil, Colombia, and Mexico are almost half as large as the rates of increase in total "urban" use of electricity re- ported by the Economic Commission for Latin America. The electricity supply industry requires heavy capital investment if it is simply to maintain a constant level of electricity sup- plied per urban inhabitant. Industrialization is discussed after urbanization so that it 24 Part 1. Problems of the Industry Table 2.3. Cumulative Average Annual Rates of Urban and Rural Population Growth, Latest Inter-censal Pe- riod, Brazil, Colombia and Mexico Brazil Colombia Mexico Largest city ................... 4.11% (Rio) 4.17% 4.42% 4.86 (S. Paulo) Other cities over 100,000: latest census. 3.93 5.93 5.21 Other urban. 3.97 3.86 4.77 Average urban .4.18% 4.33% 4.77% Rural .1.77% 1.04% 1.50% Source: See official statistical Anuarios for Brazil and Mexico; for Colombia, see Censo de Colombia, 1938, Resumen, and Economria y Estadistica, Jan.- June 1954, p. 23. will be clear that more investment in new electrical facilities in the countries studied has been required by non-industrial than by industrial uses. But a rapid growth of modern indus- try requires a rapid growth in the supply of electricity. If this is not available from public utility sources, industry will install private plant in spite of higher costs per kilowatt hour. A distinctive feature of Latin American industrial power use has been the expansion in electricity consumed per unit of industrial output. Available data are summarized in Table 2.4. In Brazil, more intensive use of electricity accounted for much of the increase in industrial consumption, and in Chile, more intensive use accounted for most of it. Only Mexico seems close to the pattern set in the United States, and even this resemblance has a special explanation: the Mexican data include the relatively important but declining mining con- sumption. This increased intensity of use for electricity in industry has been produced by a number of causes: replacement of handi- crafts by factory production; replacement of non-electrical prime movers by purchased or privately-generated electricity; The Industry and Its Setting 25 greater power use per worker and per unit of output in es- tablished industries. One of the most important factors in this shift to electricity has been a shift in the composition of in- dustrial output. Industries using large amounts of electricity are growing relative to industries using small amounts. Some indication of the character of this shift in Brazil is afforded by Table 2.5. Thus industrial structure is closely related to the electricity supply problem. If heavy power-users are growing relative to light power-users, and in addition the use of electricity is in- creasing in each separate industry, then the power problem is over-simplified unless industrialization is considered in terms of "what kind?" as well as "how much?" The trends shown in Table 2.5 have been accompanied, in Brazil and in other countries studied, by severe industrial power shortages. In this environment, stimulation of heavy power-users can mean dep- rivation of power to industries which could add more to na- tional economic development. Table 2.4. Percentage Increases in Industrial Output, and in Industrial Use of Electricity (Public Utility and Private Plant Supplies) Annual average increase (compounded) Use of electricity Period Industrial per unit of Industrial use Country Covered outpuit industrial oUtpult of electricity Brazil ............. 1939-52 7.30% 5.87% 18.67% Chile .............. 1940-48 4.62 7.92 12.90 Mexico ............ 1943-49 5.25 2.21 7.58 United States ....... 1939-47 7.32 1.65 9.06 Source: See Brazil, Berenhauser and Pereira, Aspectos do Pro blema da Energia, Aguase Energia Eldtrica (Rio de Janeiro, July, 1953) p. 9. Chile, Pan- orama Econdmico, Year VII, pp. 329, 331 (June 19, 1953), citing Direc- ci6n General de Estadistica. Mexico, International Bank for Recon- struction and Development, Combined Mexican Working Party, The Economic Development of Mexico (1953), Table 77, p. 275, and Table 78, p. 276. United States, U. S. Census of Manufactures, 1947, Fuels and Electric Energy Consumed, p. 3. 26 Part 1. Problems of the Industry Table 2.5. Brazil: Rank of Selected Manufacturing Industries, by Rate of Growth, 1939-49, and by Ratio of Value of Electricity Used to Value Added by Manufac- ture, 1949 Rank in ratio of cost of elec- Rank in rate tricity used to value added of growth (ratios in parentheses) Rubber products ......... ......... 1 7 (1.27) Metallurgy ....................... 2 2 (1.80) Transport equipment ....... ...... 3 5 Non-metallic minerals ....... ...... 4 4 (1.46) Paper and products ........ ....... 5 1 (3.32) Beverages ........................ 6 10 (0.74) Textiles .......................... 7 3 (1.49) Printing and publishing ...... ..... 8 9 (0.89) Food processing .......... ......... 9 6 (1.31) Leather .......................... 10 8 (1.11) Source: Rate of growth statistics from Joint Brazil-U. S. Commission, The De- velopment of Brazil (Washington, D. C., 1955) p. 293. Value added ratios computed from 1949 Census of Production data. CHAPTER III Financing Electrical Expansion Electricity suppliers usually depend on sources external to the industry for at least part of the funds they require for expan- sion. They must grow rapidly to keep up with the demands of their customers, and they are not allowed to respond to the growth of demand by raising prices freely, and thus financing new investment to meet the demand out of earnings. More- over, the capital needs of the electricity supply industry do not fit readily into the pattern of commercial bank lending. A hydro project may be under construction for a decade and have an economic life of a century. Short-term lenders cannot be expected to meet such long-term requirements. Normally much of the expansion of the electric power in- dustry is financed by the issuance of long-term securities: bonds and debentures and, in the case of privately-owned suppliers, preferred and common stock. But in Latin America the ability of these securities to attract domestic, not to mention foreign, savings for investment in the electric industry has long since been seriously impaired, if not totally destroyed, by the low level of electric rates and earnings during a period marked by increasing prices and profits for unregulated industries. Na- turally, it has been to the unregulated industries that investors have turned. 27 28 Part 1. Problems of the Industry In this situation government financial assistance to the elec- tric industry has seemed to provide the obvious way out. This assistance has taken two forms: aid from the general budgets of central or local governments, and aid from the proceeds of special taxes or from development banks. But no government in the countries studied has been successful in tapping the flow of national savings, through either taxation or borrowing, on the scale that is needed to finance the required electrical investment. Therefore government aid for electrical expansion has been furnished, in whole or part, through the creation of new purchasing power by the central bank. This has created a paradox: the electricity supply industry has contributed to in- flationary pressures because it has been one of the chief suffer- ers from inflation. This chapter will illustrate these general comments within the framework of each national economy. The major topics to be considered are: (1) Private financing of privately-owned electrical suppliers; (2) The financing of government-owned electrical suppliers; (3) Special government financing meas- ures-"mixed" corporations, development banks, electrifica- tion taxes. No attempt will be made to review the experience of every country studied with respect to each of these topics. Instead the discussion will concentrate on those countries whose experience most clearly demonstrates the character of the electrical financing problem. At the close of the chapter, a fourth type of financing will be noted: loans by the World Bank and Eximbank. The availability of loans from these sources, however helpful, provides no escape from the prob- lems considered in the rest of this chapter. Financing Electrical Expansion 29 1. The Private Financing of Privately-Owned Electrical Suppliers Every country studied has suffered from acute electric power shortages at some time or other since World War II. In sev- eral important supply areas, shortage has been the rule and abundance the exception. Even when generating capacity has been adequate, service to ultimate consumers has often dete- riorated because of bottlenecks in distribution capacity. The rapid rate of installation of private plant in some of the most important industrial areas of Brazil and Colombia points to an almost continuous struggle to obtain enough public utility electricity, and frequent statements of government officials and industry leaders testify to the shortage of capacity. Why, in view of the continuing pressure of unsatisfied de- mand for electricity, has private financing of expansion been so inadequate? The answer to this question varies, in detail, from country to country. The experience of Brazil is repre- sentative of the problem in an environment of severe inflation. Mexico provides an example of the difficulties confronting private financing even when general inflationary pressures are considerably less. A. THE BRAZILIAN EXPERIENCE Brazilian data for the period before the onset of extreme inflation are reasonably complete. Table 3.1 shows average rates of return on net worth for an important group of elec- tricity demanders (corporations engaged in manufacturing), for electricity suppliers, and for Brazilian corporations gener- ally during this period. By 1952, the discrepancy between rates of return in elec- tricity supply and in manufacturing was greater than these 30 Part 1. Problems of the Industry Table 3.1. Average Rates of Return on Net Worth Brazilian Corporations, 1948-52 1948 1949 1950 1951 1952 Manufacturing .......... 13.2% 12.9% 17.3% 18.8% 12.8%7o Electricity supply ....... 7.5 9.0 9.8 9.6 9.2 General average of all corporations ....... 13.6% 13.5% 14.4% 17.9% 13.4% Source: Conjuntura Econ6mica, Year III, No. 10 (Oct. 1949) p. 24; Year IV, No. 12 (Dec. 1950) pp. 20-21; Year VI, No. 12 (Dec. 1952) pp. 27-28; (In- ternat. Ed.) Year II, No. 2 (Feb. 1955) p. 83. statistics indicate. Law 1474 of November, 1951, permitted up- ward revaluation of fixed assets on the basis of coefficients ranging from 8.0 for 1925-29 to 1.5 for 1945-46. Many manu- facturing concerns revalued their assets accordingly, but the electricity supply industry did not.' Moreover, as might be anticipated in an economy with rapidly-expanding output and rising prices, manufacturing corporations distributed a relatively small percentage of their profits: dividend payments were only 36.6% of manufacturing earnings after taxes in 1949, 32.3% in 1950, 43.2% in 1951, and 38.8% in 1952.2 Over this span of years, retained earnings in manufacturing were at least as high as a percentage of net worth as total earnings in electricity supply. Therefore the rate of manufacturing expansion from reinvestment could ex- ceed the rate of electricity supply expansion even if electricity suppliers paid no dividends at all. And manufacturing expan- sion, in a country like Brazil, is only one source of increased industrial demand for electricity. This initial lead of manufacturing was greatly increased by the operation of Brazilian capital markets during this period. Table 3.2 shows the relative values of security issues for new 1 Conjuntura Economica (Internat. Ed.) Year I, No. 2 (Feb. 1954) p. 87. 2 Id., Year IV, No. 12 (Dec. 1950); Year VII, No. 2 (Feb. 1953); id. (Internat. Ed.) Year I, No. 2 (Feb. 1954) p. 87. Financing Electrical Expansion 31 money by manufacturing corporations and by electricity sup- ply corporations between 1948 and 1952. The ratio for the United States has been added for purposes of comparison. Table 3.2. Brazilian Capital Issues for New Money, 1948-52 (Federal District and Sdo Paulo only) 1948 1949 1950 1951 1952 1. Manufacturing (Cr. mn.) ...... 1538 988 1294 3404 3762 2. Electridty supply (Cr. mn.) ..... 470 36 121 349 82 3. % Line 2 to line I ....... ..... 3.0% 3.6% 9.4% 10.3%o 2.2% 4. United States new issues: % of electridty supply to mfg . ....... 70.5 187.4 673.6 61.1 53.5 * Excludes Cr. 400 million issue of Companhia Hidro-Eletrica do SSo Francisco. Although this company is legally a "mixed" corporation, with securities available to both public and private buyers, practically all of its Cr. 400 million issue in 1948 was in fact purchased by Brazilian government units or agencies. Source: Lines I and 2, Conjuntura Econ6mica, Year III, No. 2 (Feb. 1949) p. 32; Year IV, No. 2 (Feb. 1950); Year V, No. I Jan. 1951) p. 46; Year VI, No. I (Jan. 1952) p. 44; Year VII, No. I (Jan. 1953) p. 56; Line 4 from Loughlin F. McHugh, "Financing Corporate Business," Survey of Cur- rent Business (April 1954) p. 16, and Moody's Public Utility Manual, 1953, p. a20. Electricity supply enterprises in Brazil raised only one-tenth to one-fortieth as much money from private investors as manu- facturing corporations in this period. Suppliers of electricity did not even receive a share of new money proportionate to their existing assets: at the end of 1952, fixed assets of electric- ity supply corporations had a value equal to 12.3% of the fixed assets owned by manufacturing corporations.3 This record contrasts sharply with that of the United States. Rates of return allowed to electrical utilities in the United States are deliberately kept below the level which would be 3 Conjuntura Econ6mica (Internat. Ed.), Year II, No. 2 (Feb. 1955) pp. 81, 83. 32 Part 1. Problems of the Industry required to permit the degree of self-financing achieved by manufacturing concerns during periods of rapid expansion. But, as the figures indicate, these rates of return are nonethe- less high enough to permit electrical utilities to compete effectively for new capital. In Brazil a functioning capital market exists, but the Brazil- ian experience makes it plain that the mere existence of such a market does not provide an answer to the problem of utility financing. Instead, by furnishing a ready outlet for private sav- ings in the securities of profitable industrial and commercial companies, the market has made it even harder for the electric and other regulated industries to tap these savings for their capital needs. As industrial and commercial uses of electric power have been expanding, the Brazilian and other Latin American economies have been faced by a choice between per- petual industrial power shortages and continued dependence on government funds. B. THE MEXICAN EXPERIENCE Mexico has suffered less from inflation and from unbalanced economic development than Brazil since World War II. But its experience in the private financing of electrical expansion has been similar to that of Brazil in a number of ways. Table 3.3 presents an industry-by-industry comparison of profit rates in Mexico. These percentages show an even greater disadvantage for the electrical industry than those reported for Brazil. Even if electricity suppliers had reinvested all their earnings, it seems unlikely that the resulting expansion in electric plant would have kept abreast of the expansion in de- mand for power resulting simply from the growth in manu- facturing investment, neglecting the increases in demand arising from qualitative changes in manufacturing and the growth in urban population and in non-industrial uses of elec- tricity. Financing Electrical Expansion 33 Table 3.3. Ratios of Profit to Invested Capital in Mexico, by Industry, 1953 and 1954 Industry Number of companies 1953 1954 1. Electric energy ............ ........... 11 4.7% 3.0% 2. Mining .............................. 4 10.1 13.6 3. Steel ................................ 7 13.4 16.5 4. Construction materials ........ ........ 12 19.2 16.3 5. Manufacturing ............ ........... 51 15.5 20.3 6. Finance .............................. 71 23.1 26.6 General average (or total) ........ 160 16.3% 19.1% Source: Nacional Financiera, S. A. El Mercado de Valores, Year XV, p. 445 (Sept. 19, 1955). Mexico is of further interest because it provides a particu- larly good example of the problem of establishing a domestic capital market for the securities of the electrical industry. The Mexican Light & Power Company is by far the largest privately- owned utility system in the country. The company and the governmental investment bank (Nacional Financiera) have been interested in promoting the growth of ownership of the company's securities within Mexico. For example: On March 29, 1957, Nacional Financiera agreed to lend the Company one hundred million pesos to finance the Company's local construction requirements during 1957 and 1958. The loan bears interest at 8%7 per annum and matures in semi- annual installments beginning in 1960 and ending in 1981. As both Nacional Financiera and the Company desire to encour- age investment by the Mexican public in the Company's securities, the loan agreement provides that this loan may be funded at the request of Nacional Financiera any time up to March 29, 1961, by delivery of 8%, convertible debentures. These debentures will be convertible, during the three years immediately following their delivery to Nacional Financiera, into common shares of the Company at the rate of four com- 34 Part 1. Problems of the Industry mon shares for each 1,000 pesos principal amount of deben- tures, which represents a price of $20 U. S. per share at the present rate of exchange. This conversion right will only apply when the debentures are held by private investors.4 For a time at least, the Mexican regulatory authorities ap- peared more aware of the dangers of undue time lags in rate adjustments than were the regulatory authorities of some of the other countries studied. After the peso devaluation of April, 1954, the Company received a 21.2% general rate in- crease effective in October, plus "certain tax advantages for the years 1954 and 1955 in order to reduce the amount of the rate increase."5 It received permission for a further rate in- crease averaging 12½2% effective January 1, 1957.6 These in- creases may be compared with advances of Mexican wholesale and retail prices averaging about 40% between 1953 and 1957. However, a 15% wage increase plus other economic benefits was granted in March, 1958, and it was not until November that a compensating rate increase was allowed effective Feb- ruary 1, 1959, with the partial consolation of provision for "recuperation over a period of twelve months of the additional wage costs incurred since" March.7 Even before the 1958 wage increase, the problem of selling the company's securities in Mexico had not been solved, as is shown by Table 3.4. 4 Mexican Light & Power Co., Ltd., Annual Report, 1956, p. 3. 5 Id., 1954, p. 3. 6 Id., 1956, p. 3. 7 See Mexican Light & Power Co., Letter to shareholders, Dec. 18, 1958. The provision for recuperation in the rate increase underlines the need for automa- ticity in the making of adjustments. Here the need for higher rates caused by the wvage increase is ultimately recognized, but for a year in which the surcharge is in effect, the burden of the higher rates on the consumers is doubled (and not always on the same consumers who enjoyed the unduly low rates before the in- crease). In the period prior to the rate order, the utility's earning power and credit standing have been jeopardized. Moreover, in any country where the pur- chasing power of money was steadily falling, the real value of the "iecuperative" rate increase would be less than the cost of the wage increase to the company be- fore the surcharge became effective. Financing Electrical Expansion 35 Table 3.4. Mexican Light & Power Co.-Rates of Return, 1950-56 1950 1951 1952 1953 1954 1955 1956 Ratio of net operating revenue to rate base allowed for tariff purposes, assuming use of: 1. Company depreciation rates .... ............... 3.0% 3.5% 4.4% 5.7% 4.5% 7.1% 6.8% 2. Regulatory depreciation al- lowance .................. 4.1 5.0 6.2 7.8 7.0 10.1 10.0 Ratio of net operating revenue to company net capital assets: 3. Including intangibles* .... 3.0 3.1 3.5 4.7 3.5 5.5 '5.2 4. Excluding intangibles .... 3.9 3.8 4.2 5.6 4.1 6.4 6.1 Ratio of net profit after pfd. dividends to common stock and surplus: 5. Equity reported by Com- pany .. .......... 0.6 1.6 3.2 5.5 2.6 5.6 5.2 6. Line 5 less intangibles .... 1.1 2.3 5.3 8.7 4.0 8.4 7.5 *"Rights, contracts, franchises, and goodwill." Source: Mexican Light & Power Co., Ltd., Annual Reports. In recent years, the Tariff Commission has established the objective of permitting the Mexican Light 8c Power Company to earn 14.7% before income taxes, or approximately 11.5% after taxes. Actual earnings, on the basis of the Tariff Com- mission's own depreciation allowance, reached a peak for the period summarized of 10.1% in 1955. But the Commission's "retirement allowance" was 10 million pesos in 1950 and still 10 million pesos in 1956, after the restoration in 1954 of re- ductions in the allowance made in 1952 and 1953. The Com- pany's rate base, as computed by the Tariff Commission, had doubled between 1950 and 1956. So the.Commission's "retire- ment allowance," which was equal to 1.4% of the rate base in 1950, had dropped to only 0.7% in 1956. Meanwhile the Com- 36 Part 1. Problems of the Industry pany's own depreciation allowance for the years 1950 through 1956 varied over a range of 1.8% to 2.3% of gross tangible fixed assets as reported in its annual balance sheet.8 By 1956, the discrepancy between the Company's depreciation charge and the Tariff Commission's allowance for retirements and replacements was approximately Canadian $3,400,000, or close to half of the Company's Canadian $7,300,000 gross operating revenue. Moreover, the 1954 devaluation of the Mexican peso had reversed the relationship between the Company's rate base, as calculated by the Tariff Commission, and the Company's net tangible fixed assets as recorded on its own books. At the end of 1953, the Commission's rate base was the higher of the two, doubtless because of the very low annual charge-off resulting from the Commission's minimal retirement allowance. The devaluation changed neither the Commission's rate base, ex- pressed in pesos, nor the Company's net fixed asset valuation, expressed in Canadian dollars. Consequently, the rate base be- came lower than the Company's own asset valuation at the new lower exchange rate of the Mexican peso. Yet the Mexlight example also illustrates the fact that, in terms of gross revenue, the margin between an adequate and an inadequate return may not be wide. The Company's net operating revenue in 1956 was $7.3 million. An 11.5% return on its net tangible fixed assets, supposedly the goal of Mexican regulatory policy, would have required additional operating revenue of $6.35 million, plus a further addition to revenue to cover income taxes on these extra earnings. This $6.35 million would have increased the Company's net operating revenues by 84%/,, less applicable taxes, yet to produce this amount would have required an increase in the Company's gross op- erating revenue of only 18%. 8 Even the highest rate of 2.3% is lower than the average rate for electrical utilities in the United States. Financing Electrical Expansion 37 C. PRESENT PROSPECTS FOR PRIVATE FINANCING IN THE COUNTRIES STUDIED Electric utility regulation in Latin America began after the most important supply enterprises had passed to companies financed by foreign capital and owned by foreign investors. During the period of regulatory growth, in the 1930's, new foreign funds would probably not have been available for Latin American utilities almost regardless of their rates of return. The habit of purchasing securities in national enterprises wvas just developing, and therefore domestic funds probably could not have made a significant contribution. Brazilian Trac- tion, which was an outstanding case of expansion during this period, financed its growth almost entirely from depreciation accruals and retained earnings. Therefore, from the beginning, the regulatory pattern understandably developed without ref- erence to any possible connection between allowable earnings and the attraction of capital. In Brazil, Chile, and Colombia, the pace of inflation is now much faster than it was twenty years ago, and thus the problem of private financing of a regu- lated industry has been rendered inherently more difficult. Yet, though Brazil has recently permitted one revaluation (subject to tax) of assets in the rate base, it has not abandoned original cruzeiro cost of assets as an appropriate denominator for computing rates of return, and in Chile, regulatory action has lagged so far behind inflation that a theoretical regulatory allowance for price increases has become almost meaningless in practice. Since World War II, the private foreign-owned utilities have shown much more interest in the possibility of selling securi- ties to local investors. The Mexican Light & Power Company interest in attracting Mexican investors has already been dis- cussed. The American & Foreign Power Company managed to sell fairly large blocks of subsidiary common stock in Brazil before the onset of extreme inflation, as well as a small amount 38 Part 1. Problems of the Industry of subsidiary bonds in Costa Rica. The Company has also sold substantial volumes of bonds of its Cuban subsidiary in that country. Brazilian Traction, with a strong tradition of self- financing through reinvestment of an unusually high percentage of earnings, obtained both Canadian and Brazilian legislation by late 1956 enabling it to shift the head offices of two sub- sidiaries to Brazil so that it would be "possible for the first time for the Brazilian public to be offered participation by purchase of shares and securities in the undertakings of these subsidiary companies...." But this changed attitude on the part of the companies has been confronted by regulatory policies which, in effect, dis- criminate against domestic investors. Companies have been able to obtain rate adjustments in one way or another, which make some allowance for increased costs of foreign transactions whenever local currency units are devalued. Payments of in- terest on foreign debt are often handled in this way as well as by the provision of specially favorable exchange rates for suclh transfers. But regulatory policy with respect to domestic in- vestors has been discouraging in its effect. Since there are domestic savings in all of the countries studied which could be mobilized for electrical investment, this attitude is indeed paradoxical. Part 2 of this study will therefore explore pos- sible changes in regulatory measures that would open the way to domestic financing of the industry, public as well as private, from private sources. Before turning to the next topic, however, a brief report should be made of recent developments which suggest that, where economic prospects are encouraging and regulatory poli- cies are compatible with financial soundness, Latin American electric companies will be able to turn again to United States private investment sources for at least a part of the funds they need. In its Annual Report for 1957, the American & Foreign Power Co. announced (at p. 9) its re-entry "into the private capital markets of the United States for the first time in over 9 Brazilian Traction, Light & Power Co., Ltd., Annual Report, 1956, p. 6. Financing Electrical Expansion 39 25 years." The Metropolitan Life Insurance Co. and two other investors agreed to join with the Eximbank in purchasing $52 million of an authorized issue of $75 million of 57/8% Sinking Fund Debentures. Proceeds were to be invested in Cuba, Colombia, Costa Rica, Ecuador, Guatemala, Panama, and Venezuela where, as President H. B. Sargent explained in his letter to shareholders, "currencies have been relatively stable and where earnings of our companies as a group have been generally satisfactory." In its Financial Report for June 30, 1958, the Company reported that its subsidiary, Compafiia Cubana de Electricidad, had successfully placed 350,000 shares of common stock "with a limited number of private investors in the United States," of which the Company was one-for 53,000 shares. Hopeful as these signs are, the amounts are small in relation to the total need and the conditions limiting the use of the proceeds make it plain that this help can be obtained only in relatively stable economies. 2. The Financing of Government-Owned Suppliers of Electricity The shift in government finance of the electricity supply industry in Latin American countries studied has been accom- panied by the development of a number of different channels for government aid: direct government financing of publicly- owned supply enterprises; government financing via inter- mediaries; government contributions to "mixed" corporations, which combine both private and public ownership; and loans by the government or by a government investment bank (Na- cional Financiera in Mexico; the Development Bank in Brazil). But the main channel for government funds in all the coun- tries studied has been from the regular budget via govern- ment-owned corporations. This general case will provide the theme for this section. Special government financial arrange- ments will be discussed in the succeeding section. 40 Part 1. Problems of the Industry Once the financing of the electricity supply industry be- comes a charge on the government budget, the problems in- volved in the industry's financing, pricing policies, and rates of return on investment become different from those experi- enced when the industry is privately financed: First, privately-owned suppliers of electricity have no con- trol over the banking system of the countries in which they operate, and so have no control over the creation of money. A government-owned electricity supplier could, of course, be required to finance expansion independently of both the government budget and the banking system. The banking system might even be in a position to offset any inflation which might otherwise result from electrical investment programs; it could reduce its loans in other areas of the economy hav- ing an inflationary potential. However, in the type of direct government financing of electrical expansion which, as we shall show, is habitual in most of the countries studied, this institutional control is lacking. Governments can, and do, require the banking system to create new money as an aid to government finance. In general, governments do not require banks to offset increased loans for electricity supply or other purposes. Under these conditions, increases in purchasing power created by the banking system are often not matched by increases in the supply of goods or services available for purchase. Unless offset by other factors, this contributes to a continuing inflationary chase of money after goods. Second, once the financing of electricity supply is under- taken by governments which do not attempt to meet the com- petitive tests imposed by an independent money or capital market, the way is open for indefinite subsidy. This subsidy may be "inadvertent," in the sense that costs are not defined properly or inclusively when government suppliers ascertain their revenue requirements and so the resulting deficit has to be covered, somehow, by the government. The subsidy may, on the other hand, be the result of a deliberate act of public policy. Or these two reasons for subsidization may be confused Financing Electrical Expansion 41 and combined as when government financing is required be- cause the industry being assisted has not been permitted to cover all its costs, including the cost of capital, and therefore cannot raise funds for expansion from any non-governmental source. This inadvertent subsidy is then used as proof that the industry should be subsidized. A. BRAZIL: CONFLICTING OBJECTIVES AND DIVIDED POWERS This tendency to leap from the demonstrated fact that the industry is not paying its own way to the conclusion that it should not, in the national interest, be expected to pay its own way has been particularly evident in Brazil. The explanatory report accompanying the 1954 proposal for a national elec- trical program begins its statement of the case by observing: ". . . the rate of return permitted by Brazilian law for capital invested in electricity causes serious difficulties for the expan- sion of the industry...." The report then reviews previous government efforts to assist privately-owned electrical enter- prises, and concludes: Without abandoning the basic criterion of profitability . . .the Plan cannot turn back at the prospect of undertak- ings temporarily with little or no profit, or only indirectly profitable . . . the promotional effects of electric energy . ought to be weighed as a second form of profit. . ..10 The step that is lacking in this analysis is, of course, an in- quiry whether the existing legal limitation on the electric in- dustry's rate of return must be allowed to keep on causing "serious difficulties for the expansion of the industry." Cer- tainly there seems no economic necessity. Why should an industry supplying a needed, valuable service to other indus- tries which are operating profitably be forced to confront the prospect of "little or no profit"? There is no showing of any 10 Presidencia da Repbiblica, Plano Nacional de Eletrificagrzo e Centrais Eldtri- cas Brasileiras, S. A.M... Memoria Justificativa" (1954) pp. 48, 60. (Translation supplied.) 42 Part 1. Problems of the Industry incompatibility between a profit to the electric industry suf- ficient to permit its expansion and, at the same time, prosperity and expansion in the industries it supplies. The report points out that ... An abundant supply of electricity could be the point of departure for the creation of industries producing import substitutes, . .. particularly electro-metallurgical and electro- chemical, . . . electrification of railways and urban trans- port."1 But to note the possible uses for an abundant supply of elec- tricity is not to demonstrate that this supply must be subsi- dized. And, while the problem of defining objectives is peculi- arly for the country which is to pursue them, we cannot refrain from observing that the industries specifically mentioned as meriting promotion are the most gluttonous of all users of electric power. In a scheme of priorities in a country suffering from shortages of both capital and power, one might suppose that industries which required less lavish outlays of both capi- tal and power would be the first to be encouraged. Moreover, in spite of Brazilian efforts to coordinate elec- trical development through such instruments as the National Electrification Fund and the Development Bank, the split be- tween Federal and State responsibility in Brazil has tended to blur the meaning of electrical costs and to enhance inflationary pressures. No single publicly-owned supplier in Brazil initiates and carries out a national electrification policy. One obvious re- sult appears in the 1952 annual report of the Comissao Estadual de Energia Eletrica (CEEE) of the State of Rio Grande do Sul: In general, the contribution of the Federal Government represents one-fourth of the investment of the State Govern- ment and, as the value of the facilities constructed with Federal funds will not be computed in the price per kilo- watt-hour, the importance of this in the establishment of tariffs is obvious.12 11 Id., p. 56. 12 CEEE, Relatdrio, 1952, p. 16. (Translation supplied.) Financing Electrical Expansion 43 This in effect treats electricity as a free good if it comes from plant built with Federal funds. The situation of a number of Brazilian utilities in fields other than the electric industry has also been a deterrent to continued investment. Some exam- ples follow. A 1948 discussion of Brazilian railways noted that: . . . Most of the lines have an operating deficit, without con- sideration for depreciation or return on the invested capital. Only one company, the Paulista, is fully paying its way with- out financial aid from state or federal government. Another privately-owned line, the Mogiana, has been paying divi- dends for several years, but the management is so concerned about the outlook that it is ready to turn its burden over to the state.13 Total operating deficits of the Federally-owned railroads amounted to Cr. 5 billion, or 8% of the Federal budget, in 1954. The reason: "the present tariff scale bears no relation- ship with present operating costs."'4 Also in 1954, the Sio Paulo Gas Company decided to cancel its contract in spite of authorization for a rate increase. The service was therefore to revert to the City of Sao Paulo.15 Street car service in Saio Paulo had already followed the same course. In 1954, again: . .. Water shortages in the principal large cities continued ... principally because of unrealistic tariff schedules. In two cases, where water supply has been operated by private capi- tal, the companies concerned cancelled their contracts and turned the service over to the local municipal authorities.16 These instances suggest that Brazilian public-utility control has developed a spiral: (1) Control of utility prices results in a lag behind the general price inflation. This makes outside 13 George Wythe, Royce A. Wight and Harold M. Midkiff, Brazil: An Expand- ing Economy (New York, 1949) p. 192. 14 Conjuntura Economica (Internat. Ed.) Year II, No. 2 (Feb. 1955) p. 89. 15 Ibid. 16 Ibid. 44 Part 1. Problems of the Industry financing both more necessary and more difficult; as the lag in- creases, it finally produces operating deficits. (2) These indus- tries are then transferred to the public sector of the economy if the operating deficits are hopeless; at least the public sector must provide expansion funds if the industry cannot. (3) The spread between prices charged by these industries and other prices continues or even widens, thereby reducing the internal flow of expansion funds or increasing deficits. (4) Funds to meet deficits and for expansion are obtained in part or en- tirely by the creation of new money and new bank reserves. (5) This further stimulates inflation and pushes still other public utility enterprises over the brink. The end result for Brazilian public finance is expounded by an authoritative Brazilian source: It is generally recognized that the main reason for chronic deficits is to be found in investment expenditures . . . Sup- plementary resources have usually been obtained through bank loans, generally . . . from the Bank of Brazil (i.e., generally by the creation of funds which soon add to the reserves of other Brazilian banks and enable them to expand their operations by a multiple of the original loan).17 Once this stage is reached, the original question of how to finance public utility expansion in an inflationary environ- ment becomes merged with the larger question of how to con- trol the inflation in view of the methods used to finance public utility expansion. In principle, State control of electrical utility expansion should have less inflationary consequences than Federal con- trol, since the States supposedly cannot create money directly or indirectly. This is doubtless true of the smaller and poorer states. However, it seems not to be true, in effect, for a large and relatively wealthy State such as Sio Paulo. In spite of its wealth, this one State accounted for Cr. 16 billion of the Cr. 21.3 billion combined budget deficits of all Brazilian governmental units-including the Federal government-between 1950 and 17 Id., p. 63. Financing Electrical Expansion 45 1954.18 The State financed its deficits for a number of years prior to 1953 by the issuance of rotating bonds, acceptable in the future in payment of taxes. By the beginning of 1953, Cr. 10.7 billion of these bonds were outstanding. They were selling at a discount of 40% in the market even though they could be used at par for tax payments. The State was falling behind in payment of salaries and ordinary commercial bills-and so the Federal Government was forced to come to the rescue, with a credit for Cr. 5 billion from the Bank of Brazil which was usable as a substitute for half of the rotating bonds in circula- tion.19 At the very time of this crisis, the State was pToceeding with plans for construction of State-owned electricity generating plants, with an ultimate proposed capacity of 1,100,000 horse- power, over a period of four years.20 This planned capacity was about equal to the total electric generating plant then in existence in the State, all of it privately-owned. The State's plans, if carried out in the scheduled four years, would have added a minimum of Cr. 2 billion a year to State spending. The State had deficits of Cr. 4.7 billion in 1953, Cr. 3.8 billion in spite of a 10% tax surcharge in 1954, and budgeted for a Cr. 2.5 billion deficit in 1955.21 As far as Brazil is concerned, then, the basic problem has been to try to finance public investment by running deficits, there being no voluntary purchasers for governmental securi- ties. This has been greatly complicated by the failure of State and Federal development programs to be harmonized and coordinated. B. COLOMBIA: MUNICIPAL OWNERSHIP AND "PUBLIC SERVICE" This lack of precision with respect to the character of elec- tricity costs, and to the rationale behind government financing 18 Id., p. 72. 19 Id., Year VIII, No. 3 (March 1954) pp. 4-5. 20 d., p. 19. 21 Id., Year II, No. 2 (Feb. 1955) p. 72. 46 Part 1. Problems of the Industry of the electricity supply industry, has also been noted in Colom- bia. Successive directors of the Colombian central government agency for the finance of electrical expansion have complained bitterly about the attitude of many of the municipal suppliers who are dominant in the Colombian industry. For example, in 1954: . . . in our country . .. there exists an equivocal idea of what is the "public service." Many people believe that since elec- tric energy is a public service and the Government has met the cost of the plants, energy ought to be used like a park or a street. The majority of the enterprises supplying electric energy have too-low tariffs, which neither permit them to recover the capital invested nor to obtain any profit. Consequently, the enterprises ruin themselves, there is no improvement of service due to lack of maintenance or of plant enlargements, and service becomes worse day by day until it finally disap- pears. Then the undertakers, who are almost always ex- tremely poor municipalities, turn to the Central Govern- ment for so-called "aid" for a new electric plant.22 Deliberate policy may have had some influence on this situa- tion, but, according to a foreign expert who investigated the Colombian industry, lack of knowledge with respect to the nature of the costs of providing electricity has been an im- portant contributing factor. Returning to one aspect of this ignorance this commentator observed: . . . the prescribed rates [of depreciation] have been used only in few cases. Certain publicly-owned concerns made wholly insufficient reserves: others made no provisions at all. The officers of some concerns appeared to have no knowl- edge whatever of the purposes of a depreciation reserve.23 22 Instituto Nacional de Aprovechamiento de Aguas y Fomento Electrico, In- forme del Gerente, June 1953 to June 1954 (Bogota, 1954) p. 65. (Tianslation supplied.) Other reports of the Instituto indicate that this has been a major prob- lem for every Director. 23 Cecil A. Ellis, Public Utilities in Colombia (UN, Tech. Assist. Admn., New York, 1953) p. 24. Financing Electrical Expansion 47 The danger is, then, that government subsidies will con- tinue, and even grow, through simple failure to define what is meant by subsidy. C. CHILE: THE EXPERIENCE OF ENDESA Endesa (Empresa Nacional de Electricidad, S. A.) is a "mixed" corporation almost wholly owned by the Corporaci6n de Fomento, an autonomous agency of the Chilean govern- ment. Endesa, the chosen instrument for publicly-financed electrical development in Chile, offers a case of a government- owned electrical utility which has added to the government deficit at critical periods in spite of its own strenuous efforts to achieve a more nearly self-supporting basis. Endesa's first problem is the rate of expansion expected of it. Other suppliers of electricity in Chile have had very limited funds for financing expansion and practically no prospects for successful security sales. Thus Endesa has been forced not only to provide by far the largest proportion of new public utility generating plant; it has also had to take over distribution net- works in various Chilean towns and cities, after the bankruptcy of their previous private owners, and to rehabilitate them. Therefore, for any given rate of increase in national electrical capacity, Endesa must budget for a considerably higher rate of increase in its own capacity. Endesa's second problem has been that of obtaining per- mission for rapid rate increases in a period of extreme infla- tion. As a consequence of its efforts, Endesa's return to average equity rose substantially during the years 1954-56, rising from 5.0% in 1954 to 8.0% in 1955 and 11.9% in 1956. These per- centage rates of earnings are even more striking in view of successive upward revaluations of assets over these years, to balance the increased peso liabilities resulting from successive revaluations of Endesa's foreign debt from the original ex- change rate of Ps. 31.10 to the U. S. dollar to a rate of Ps. 500 48 Part 1. Problems of the Industry to the dollar at the end of 1956. This revaluation increased the peso total of a U. S. $10.6 million debt from Ps. 333 million at the end of 1953 to Ps. 5.3 billion at the end of 1956-an increase due to revaluation (matched on the asset side) of almost Ps. 5 billion, or a substantial share of the recorded value of total fixed assets-Ps. 17.7 billion-at the end of 1956. Not only has Endesa thoroughly been alive to the impact of inflation on rate of return and making considerable headway against it, but, unlike many government-owned suppliers in other countries, Endesa has been keenly aware of the im- portance of depreciation as a source of funds. Beginning in 1954, it has even added an extraordinary depreciation charge to its normal depreciation to allow for cost increases since the original installation of its assets. Over the three year period 1954-56, the sum of normal and extraordinary depreciation reached the unusually high levels of 5.03% of average depreci- able assets in 1954, 5.93% in 1955, and 6.48% in 1956. Endesa's third problem has been that of obtaining adequate funds from the Corporaci6n de Fomento, and eventually from the Chilean government, to finance its expansion plans. Since, despite the policies it has followed, Endesa has not been able to generate more than a fraction of the needed funds from its own resources and has had no securities to offer which might attract private Chilean investors, it has perforce become only another net claimant for funds from a government budget al- ready in deficit. Table 3.5. Chile: Empresa Nacional de Electricidad, S. A., Se- lected Financial Data, 1953-1956 1953 1954 1955 1956 1. Profit and depredation (mn. pesos) ... 249 430 953 2,024 2. New investment (mn. pesos) .......... 1,232 2,664 3,436 6,461 3. % Line I to Line 2 .................. 20.2% 16.1% 27.2% 31.3% Source: Empresa Nacional de Electricidad, S. A., Memoria y Balance General for 1953, 1954, 1955 and 1956. Financing Electrical Expansion 49 Tables 3.5 and 3.6 express the Endesa position in two ways: as reported, in Chilean pesos, and in "real" terms, reduced to a 1953 base by correction for subsequent increases in the San- tiago cost of living. The statistics appear to show a winning battle against in- flation. Absolutely and relatively, Endesa was apparently financing more of a greatly increased new investment from combined profit and depreciation in three years out of the four.24 Thanks to this determined policy, Endesa protected it- self from a crushing load of new financing. As Table 3.6 shows, however, there remained a very substantial continuing need for new capital from sources outside Endesa. Table 3.6. Chile: Empresa Nacional de Electricidad, S. A., Se- lected Financial Data Corrected for Price Changes, 1953-1956 1953 1954 1955 1956 1. Profit and depreciation (mn. pesos) ...... 249 250 310 430 2. New investment (mn. pesos) ....... ...... 1,232 1,549 1,138 1,372 3. New investment requiring external financ- ing (Line 2 minus Line 1) (mn. pesos) .. 983 1,299 828 942 Source: Same as Table 3.5 plus International Monetary Fund (IMF) Interna- tional Financial Statistics, for correction of reported data by propor- tionate increases in Santiago cost of living after 1953 (100): '54 = 172; '55 = 302; '56 = 471. On a "real" basis, the sum of Endesa's profit and deprecia- tion increased by almost 75% between 1953 and 1956 and the volume of new investment showed no consistent gain over this period. The clue to Endesa's continuing need for external financing may lie in the fact that profit and depreciation com- bined were only 20.2% of new investment in 1953. Therefore 24Since practically all of Endesa's share capital is owned by the Corporaci6n de Fomento, and since the Corporaci6n ploughs back its dividends each year, divi- dend payments need not be considered as an offset. 50 Part 1. Problems of the Industry their substantial relative gain over the next three years still left Endesa seriously dependent on outside financing and so posed a continuing problem for the Chilean economy to solve. D. MEXICO: THE COMISION FEDERAL DE ELECTRICIDAD (CFE) Mexico's experience with governmental electricity develop- ment has been almost the opposite of Chile's. In Chile, since 1953 the government electricity enterprise has struggled might- ily to place itself on a self-supporting basis but has been handi- capped by a low point of beginning, rapid inflation, and the absence of a market for new electrical securities of the familiar types. The rate of inflation in Mexico has been much less than Chile's in recent years, and government budgets have been un- der increasingly strong control. Yet in contrast with Endesa, CFE, the government electrical enterprise, has made a remark- ably small contribution to the financing of its own expansion, especially in the period of rapid inflation extending for some years after the end of World War II. The outstanding feature of Mexican electrical development in terms of its claims on the Federal budget is that this devel- opment is only one phase of a heavy government investment program. The Mexican government is responsible, through Petr6leos Mexicanos (Pemex), for practically all of the pro- duction, refining, and distribution of petroleum carried on in the country; it has carried out a large-scale irrigation program; it has invested heavy sums in railroad modernization and in highways. Owing to the character of Mexico's natural resources as well as to the Mexican government's conception of its role in the economy, this heavy investment load will continue or even increase. But, unlike the central governments of Brazil and Chile, the Mexican Federal government does have a market for its obligations. The character of this market must be con- sidered in appraising the government's role in financing elec- trical development. Financing Electrical Expansion 51 The most systematic study of the Mexican economy ever undertaken, that of the Combined Mexican Working Party, established under the auspices of the World Bank and the Mexican Government, revealed the shift in the nature of the Mexican Government's fiscal problem over the critical period 1939-51. This shift is evident in Table 3.7. Table 3.7. Mexican Public Surpluses and Deficits, 1939-51 (millions of pesos; deficit = -; surplus = +) 1939-44 1945-48 1949-51 Federal government .............. -464 -617 + 34 Local governments .............. - 23 - 46 - 39 Autonomous government agencies and enterprises + 55 -548 -945 Net total ..... -432 -1211 -950 Source: Tnternational Bank for Reconstruction and Development, Combined Mexican Working Party, The Economic Development of Mexico (1953), (hereafter cited as "IBRD, The Economic Development of Mexico") p. 109. The category "Autonomous government agencies and enter- prises" includes the large areas of government investment in commercial or quasi-commercial enterprises which were mainly taken over as government responsibilities during the latter part of the 1930's. The table shows a progressive shift in the contribution of the newer investment enterprises to govern- ment deficits: at first, almost no contributions; then half of the total; then practically all of the total. These deficits were accompanied by a massive increase in the Bank of Mexico's holding of government debt: the Bank's claims on the government rose from Ps. 123 million at the end of 1937 to Ps. 1,657 million at the end of 1951.25 This increase alone is equal to over half of the government deficits shown in Table 3.7. After rising to a peak of Ps. 2,381 million at the end 25 IMF, International Financial Statistics (Feb. 1958) p. 164. 52 Part 1. Problems of the Industry of 1954, the Bank of Mexico's claims on the government sub- sequently declined as the claims of other banks rose to leave total bank claims on the government at about the 1954 level. Mexico, then, provides a case of recourse to increases in money supply for the finance of government investment which must have made an important contribution to Mexican inflation. But non-bank lenders, domestic and foreign, also showed some absorptive power for government debt, and recently the gov- ernment has been able to manage its investment program with- out reliance on large deficits. The period of large government deficits, and of heavy bor- rowing from the Bank of Mexico, was by no coincidence a period of rapid price increases in Mexico. Between 1937 and 1953, the cost of living index increased at a cumulative average rate of almost 11% a year, as compared with an average of less than 4% over the same period in the United States. Therefore the finance of the electricity supply industry over this period is of special interest. Government investment in CFE, substantial though it was, represented a relatively small fraction of the total government investment in Mexico in the 1946-1951 period. However, CFE loans constituted over one-half the government's borrowing abroad, and the government's net contribution to CFE ac- counted for as much as three-quarters of the government's net internal financing in the 1949-51 period. Indeed, the gov- ernment's subsidy to CFE in that period must assume at least some of the responsibility for the inflationary increase in money supply during that period.26 26 CFE's investment, as a percentage of all government investment, was 3.7% in 1946-48 and 8.04% in 1949-51. CFE's borrowing abroad as a percentage of the government's total foreign borrowing in the first period was 58.1%, in the latter period, 54.8%. The government subsidy to CFE was Ps. 348 million in the latter period, or 75.2% of the total internal financing of the public sector, Ps. 462 mil- lion. In the same period, the increase in all claims against the government except the Bank of Mexico's was Ps. 170 million, while the Bank's claims increased by Ps. 292 million. See IBRD, The Economic Development of Mexico, Tables 19, 20, 21, 22, 72, 114, 120, 121. Financing Electrical Expansion 53 In more recent years, the public sector of the Mexican econ- omy taken as a unit (i.e. including all levels of government, plus autonomous government enterprises) has avoided any significant net borrowing from the Bank of Mexico. Therefore, more recent data for the CFE cannot be related directly to the problem of inflation. But the CFE still provides an example of a policy of government electrification which is the opposite of the energetic struggle against inflation which has characterized the policy of Endesa in Chile. Table 3.8 compares the CFE in- vestment to its gross operating profits (operating revenues less operating expenses) before deduction of depreciation and in- terest. Table 3.8. Mexico: Federal Electricity Commission, Invest- ment and Gross Operating Profits, 1938-1953 (mil- lions of pesos) Annual Cumulative Annual Cumulative Ycar investment investment gross profits gross profits 1938-44 ........ - 45.0 - - 1945 .......... 14.0 59.0 4.1 4.1 1946 .......... 26.9 85.9 3.3 7.4 1947 .......... 52.9 138.8 1.4 8.8 1948 .......... 73.8 212.6 0.9 9.7 1949 .......... 142.9 355.5 2.6 12.3 1950 .......... 212.0 567.5 9.5 21.8 1951 .......... 262.9 830.4 3.6 25.4 1952 .......... 129.0 959.4 14.4 39.8 1953 .......... 236.9 1,196.3 8.1 47.9 Sources: For 1939-1950, IBRD, The Economic Development of Mexico, Table 72; for 1938 and 1951-53, unpublished CEE-Mex data, except "Annual Investment" in 1952 and 1953, Nacional Financiera, Informe Anual, 20th Gen. Ass., 1954. CFE's reported profits have been considerably higher in some years, but these have included as profits the interest al- lowed on the cost of plant in the course of construction. It is, to be sure, a valid accounting procedure to show the full cost 54 Part 1. Problems of the Industry of construction by imputing to it interest that would be in- curred on the capital tied up in construction before the plant can produce revenues. However, this cost certainly is not a source of expansion funds. Indeed, such interest has nothing to do with the operating budget. As shown in Table 3.8 above, the results of CFE's low rate policy, which has relieved its cus- tomers of having to share in the cost of its expansion, have in- cluded the following: 1. CFE's gross profit exceeded 2% on its cumulative invest- ment at the previous year-end only in 1945, 1946, and 1950. 2. CFE's cumulative gross operating profits at the end of 1953 were Ps. 47.9 million, less than the figure of Ps. 64.8 million obtained by assuming 2% depreciation each year on cumulative investment at the start of the year. Therefore, over a critical period of its development, the CFE was not even re- ceiving a net cash flow from operations equal to the require- ments of a relatively low depreciation rate. CFE's own de- preciation calculations have involved a rate of much less than 2%. 3. The CFE gross operating profit over the years 1950-53 never came close to covering interest charges on CFE debt. Therefore, in this rapid period of expansion, CFE was a source of net cash deficit to the Mexican government before any funds were available to finance new investment. The record of CFE during these years 1938-53 is tantamount to treating the capital required for electrification as virtually a free good. CFE has done very little with respect to rural electrification, although it has taken over or started distribu- tion undertakings in a number of smaller cities and towns. Most of its investment has gone into generating plants to feed the major distribution systems. Therefore the government sub- sidy must, in the main, have been for the benefit of consumers in certain larger communities. The purchasing power of these communities, which include most of the important industrial areas of Mexico, would have been adequate to provide much higher rates of return to CFE than those it actually earned. Financing Electrical Expansion 55 E. COSTA RICA: THE INSTITUTO COSTARRICENSE DE ELECTRICIDAD (ICE) The Costa Rican price level has been more nearly stable than that of the United States since 1948. By any method of measurement, its general record of government budgetary con- trol has been better in the last decade than that of any of the other countries we have studied. The Costa Rican cost of liv- ing index in 1957 was only 6% higher than in 1951. Central bank claims on the government were lower at the end of 1957 than at the end of 1948, and substantial commercial bank holdings of public obligations were completely liquidated over this period. In 1954, the internal funded redeemable debt of the central government was distributed as indicated in Table 3.9. Table 3.9. Costa Rica: Distribution of Central Government Internal Funded Debt, April 30, 1954 (colones 000) Holder Amount Per cent Government ....................................... 11,455.5 8.42 Central bank (in trust) ........5....................25,117.5 18.46 National bank (commerdal department) ...... ...... 20.0 0.01 National Insurance Institute ......... ............... 4,632.6 3.40 Social Security Fund .............. ................. 6,556.3 4.82 Other public and semi-public ......... .............. 2,657.3 1.95 Private holders (including pension funds) ...... ..... 85,632.8 62.94 Total .......................................... 136,073.0 100.00% Source: IBRD data. This distribution of public debt is markedly different from that in any other of the countries studied. It appears to reflect a different approach to the problems of government budgeting 56 Part 1. Problems of the Industry and debt management. This inference is supported by the view of a distinguished Mexican economist that Costa Rica has re- cently been the exception to the general Central and South American rules that government securities are sold to the cen- tral bank. In Costa Rica, the central bank has neither sup- ported the market nor guaranteed redemption.27 This policy was combined with a balanced budget in 1949, and a series of budget surpluses in the next four years, in place of the very heavy deficits of the period 1940-48.28 The absence of govern- ment intervention was accompanied by a rise in price of Costa Rican 6% government obligations from 70-75%, of nominal value at the end of 1950 to 96-98% of nominal value at the end of 1954.29 This combination of a stable price index, budget surpluses, reduction in bank holdings of government debt, and a consid- erable rise in open-market prices for government bonds, would seem to provide the perfect background for the sale to private investors of securities in the nationalized sector of the Costa Rica electric power industry. Yet the legislation providing funds for construction of the La Garita hydro-electric plant of the Instituto Costarricense de Electricidad (ICE) first stipu- lated that local-currency costs of the project, estimated at 30 million colones, were to be met in four equal contributions from the government's annual budgets, and then provided that the colon counterpart of foreign exchange required for ex- ternal costs was to be obtained from issuance of 4% bonds. As might be expected in a 6-6½/2% environment, this 4% issue was to be taken by the banking system. Thus Costa Rica, in its first large-scale venture into publicly- 27 Victor L. Urquidi, "Financing Investment in Mexico and Central America," in A. Curtis Wilgus, editor, The Caribbean: Its Economy (Gainesville, Fla., 1954), p. 101. 28 Stacy May, Just Faaland, Albert R. Koch, Howard L. Parsons, and Clarence Senior. Costa Rica: A Study in Economic Development (New York, 1952) p. 282. Also App. A. 29 Price data provided by Ministerio de Economia y Hacienda, Direcci6n Ge- neral de Economia. Financing Electrical Expansion 57 financed electricity supply and in spite of unusually favorable prospects for mobilizing national savings for electricity ex- pansion, resorted to two subsidies, one overt, one concealed. Half the cost of the project was to be met directly from the government budget, and the nominally self-supporting half of the project was to be financed at a subnormal rate of interest. The introduction of this concealed financing subsidy in addi- tion to the overt construction subsidy destroyed an opportunity to relate at least a part of the nation's electrical expansion to the flow of voluntary national savings. 3. Special Government Financing Measures The preceding section was concerned with the financing of electrical expansion in the general context of governmental budgetary positions and markets for government securities. In addition, various special financing devices have been tried, of which three are of particular importance: "mixed" corpora- tions, with securities offered for both private and public sub- scription, in Brazil, Chile and Colombia; development banks with special sources of funds, in Brazil and Mexico; special taxes, earmarked for use in electrical development, also in Brazil and Mexico. These arrangements were introduced in an effort to separate the financing of electrical expansion from the general problems of the governmental budget. Therefore, they deserve special consideration. A. "MIXED" CORPORATIONS The "mixed" corporation, in Latin America as in other countries, has been designed to combine a separate appeal for the savings of private investors with access to government 58 Part 1. Problems of the Industry funds. The method has also provided the opportunity for government control without complete government ownership. For the realization of the advantages in financing that the structure of the "mixed" corporation offers, the "mixed" corporation must be allowed to earn a rate of return sufficient to attract private investors. This, of course, depends in turn upon the rates the corporation is allowed to charge for elec- tricity. However, the reluctance which the government-owned corporations have shown to seek needed rate increases seems also to have characterized the behavior of the "mixed" cor- porations. Endesa in Chile appears unique in waging an ag- gressive battle for protection against general inflation, but it did so only after it had become completely dependent on the government for its financing. More favorable treatment for private than for government investment may have been the original intention which led to the establishment of some of the "mixed" corporations. The simplest way to provide such treatment is through the issuance of bonds or preferred stock for general subscription. The gov- ernment's investment in common stock then provides the first cushion against inadequate earnings. But such bond and pre- ferred stock issues have attracted very little private investment, either because the rates of return offered have been too low or because they have been fixed, in money terms, in an environ- ment of general inflation. Two Brazilian examples illustrate the rise and fall of the "mixed" corporation. The Companhia Hidro-Eletrica do Sao Francisco was the first government-sponsored project in Brazil to appeal for non-governmental funds. Of the 200,000 prefer- ence shares originally offered, governments or government agencies took all but the 18,465 shares subscribed by the gen- eral public and the 2,350 taken by non-government banks.30 At Cr. 100 per preference share, the total non-government in- vestment amounted to only Cr. 2,815,000 out of the initial 30 Companhia Hidro-El6trica do Sao Francisco, Power Market Survey, March 1949. Financing Electrical Expansion 59 1948 financing of the company, in Brazil, of Cr. 400 million. The company has not obtained any of the additional funds it has needed from private Brazilian investors. CEMIG, another "mixed" corporation controlled by the State of Minas Gerais, has received practically all of its capital from the State government. But its subsidiaries had raised Cr. 97.85 million from private investors by October 31, 1953, in addition to Cr. 439.15 million contributed by CEMIG.31 A State Law of December 14, 1951 guaranteed a minimum divi- dend of 6% to all private stockholders in CEMIG and its sub- sidiaries. But this was certainly too little: in 1953, State of Minas Gerais 7% Economic Recovery bonds were selling to yield 14% or more.32 Chilean experience has been similar. Early in 1954, the gov- ernment investment in Endesa, via the Corporaci6n de Fo- mento, was twenty times that of the Chilean public. The reason is easily explained: between 1949 and 1953, Endesa's rate of profit on paid-in capital varied between 2.64% and 5.37%, its dividend rate on common stock rose from 2.26% to 3.50%, and the dividend rate on its preferred stock was 7%.33 These rates were all far below yields on either Chilean gov- ernment bonds or short-term bank loans during this period, and the preferred stock offered no protection against inflation. Colombian experience, by mid-1954, was like that of Brazil and Chile. "Mixed" corporations controlled 16 complete elec- tricity supply projects, representing a total investment of Ps. 85,375,000, of which all but the equivalent of Ps. 14,654,000 in loans from the International Bank for Reconstruction and Development had been provided from Colombian sources. Yet 31 Gov. Juscelino Kubitscbek de Oliveira of the State of Minas Gerais (now President of the Republic), "Politica de eletrificaqao," Aguas e Energia Eldtrica, Year V, p. 16 Jan. 1954). 32 For information as to the 6% dividend guarantees, see Joint Brazil-U. S. Commission, Project No. 11, Itutinga . . .; for bond yields, see Anuario Esta- tistico do Brasil, 1954, p. 227. 33 Panorama Econ6mica, Year V, p. 161; Year VI, p. 289; Year VII, p. 349; Year VIII, p. 230. 60 Part 1. Problems of the Industry only four enterprises had sold stock to non-governmental buy- ers, for a total private investment of Ps. 631,000-1% of the total value of stock sold by the "mixed" enterprises. In addi- tion two enterprises had obtained loans from Colombian non- governmental sources for a total of Ps. 4,335,000. The experiences of Brazil, Chile, and Colombia confirm the generalization that "mixed" corporations must either earn a rate of return which attracts private investors, or at least offer attractive securities to these investors. A rare example of a government enterprise acting on this second principle is not legally a "mixed" corporation and does not supply electricity. On May 20, 1958, it was announced that Petroleos Mexicanos would offer Ps. 200 million in bonds whose values would be adjusted in accordance with movements in the Bank of Mex- ico's index number of wholesale prices.34 Unless some such approach is employed in the electricity supply industry, the legally "mixed" corporation will remain tied to governmental budgets. B. DEVELOPMENT BANKS For present purposes, the significant aspect of development banks is the source, and not the use, of funds. Development banks can add a new dimension to electrical financing only to the extent that they can tap new sources of capital. The two most interesting instances of the use of develop- ment banks to support electrical development have been in Brazil and in Mexico. The Brazilian National Development Bank (Banco Nacional de Desenvolvimento Econ6mico) was established in 1952 and began large-scale lending operations in 1953. In addition to miscellaneous sources of funds, its financ- ing was to be obtained from (1) the proceeds of additional taxes, especially an income tax surcharge, and (2) compulsory 34 Nadonal Financiera, El Mercado de Valores, Year XVIII, pp. 244, 251 (May 26, 1958); New York Times, June 6, 1958, p. 31. Financing Electrical Expansion 61 contributions of up to 4% of the total deposits of the Federal Savings Bank, up to 25% of the annual increase in the tech- nical reserves of insurance and capitalization companies, and up to 3% of the annual receipts of the Social Security Insti- tutes, calculated before any Federal government payments to the Institutes. The income tax surcharge was expected to provide more funds than all other sources combined.35 The demand for the Bank's services may be indicated by that fact that loans sought by all types of borrowers amounted to Cr. 19,858 million by the end of 1953, or more than the Bank's anticipated receipts of loanable funds for the five years through 1957. The Cr. 2,132 million lent by the Bank in 1953 fell short of the CT. 3,442 million requested for electrical de- velopment alone.36 Thus the Bank could not be expected to carry anything like the whole burden of financing electrical ex- pansion in Brazil, nor is Brazilian planning based on any such expectation. But the Bank has already become a significant ele- ment in electrical finance. Therefore its sources of funds de- serve attention. First, the device of supporting the Bank through earmarking special tax sources amounts, in effect, to the use of taxes for electrical development. This device is discussed in the next section. Second, a forced contribution from insurance and capitaliza- tion companies and social security institutes seems much less desirable as a long-run measure than the creation of electrical securities which would voluntarily attract capital from these sources. This problem of attracting voluntary support will be considered in Part II. Nacional Financiera, in Mexico, ". . . is a complex institu- tion which goes far beyond the ordinary functions of a develop- ment bank."87 Even in its financing of economic development, 35 Material in this paragraph is from Banco Nacional do Desenvolvimento Economico, Exposiido sobre o Programa de Reaparelhamento Econ6mico, 1953 (Rio de Janeiro, 1954). 36 ld., pp. 66-71. 37 William Diamond, Development Banks (Baltimore, 1957), p. 118. 62 Part 1. Problems of the Industry it uses several devices not employed in Brazil or in any of the other Latin American countries studied. Although Nacional Financiera was first organized in 1934, it did not begin large-scale operations until 1941. At that time, the intention was to concentrate on security underwriting. The stumbling block was a familiar one: . . . In Mexico, the [stock] exchange played a useful role, in enabling the National Industrial Financing Corporation to sell shares in a third of its enterprises to the public. The corporation was eager to sell to private investors, whether local or foreign, but the public was not much interested in public utility stocks.38 The next approach was to deposit Nacional Financiera holdings of bonds and stocks in a "common fund," against which "certificates of participation" were issued for sale to out- siders. This investment trust function still did not eliminate the basic problem of earnings on securities held in the com- mon fund, however, and all certificates of participation now outstanding bear interest at fixed rates. They differ from the usual bond only in that every issue since March 1946 has had no fixed maturity. Instead, Nacional Financiera has stood ready to repurchase certificates of participation at sight and at par. The certificate of participation has become a mainstay of the Mexican capital market, and an outlet for substantial non-bank saving as well as for banking funds. Interest rates have been low by Mexican standards: 6% on outstanding certificates issued before 1951, and 5% since, as compared with the cus- tomary 8%. The price of this acceptance at a low interest rate has been the creation of demand liabilities or "additional liquidity," as the Mexicans put it, which may create troubles for the Mexican economy, just when financial crises already exist. In 1954, a year in which the external value of the peso was substantially reduced, certificates of participation outstand- 38 UN, Dept. of Econ. Affairs, Domestic Financing of Economic Developmeni (New York, 1950) p. 72. Financing Electrical Expansion 63 ing dropped by 28%, those in the hands of non-financial hold- ers declining by 350%. This difficulty has been faced in two different ways: by the issuance of "titulos financieros"-in effect, bonds with fixed interest and fixed maturity dates-and, beginning May, 1956, by the issuance of "Certificados de Copropiedad Industrial." The "titulos financieros" now outstanding are almost entirely denominated in U. S. dollars, and their use has been mainly re- served for periods when certificates of participation could not be sold. During 1954, the peso value of "titulos" increased by 120%., due in small part to revaluation in pesos of dollar ob- ligations and in large part to a shift of the Mexican public to dollar securities. The "certificates of industrial co-proprietor- ship," on the other hand, are payable as to principal and in- terest in Mexican pesos. Nacional Financiera does not guaran- tee to redeem them at any set price before maturity, and, since one-half the underlying investment is in industrial equities and one-half in securities with fixed yields, the certificates pro- vide a variable return. They are callable five years after issue in cash or in the underlying securities. This new shift of Nacional Financiera, from guaranteed re- purchase to an "open market" type of security, has been made to avoid the potential increase in money supply associated with the debt instruments previously issued. According to Nacional Financiera's own explanation: With the aim of modifying the structure of the securities market, which has been characterized by a high degree of liquidity, Nacional Financiera began in 1956 a new policy of security issuance, to introduce new types of securities ... which would be appropriate to the needs of longer term finance for economic development.39 As a matter of financial mechanics, then, Nacional Financiera is rearranging its offerings in an effort to attract long-term voluntary savings as an offset to the long-term investment it 39 Nacional Financiera, S.A., Informe Anual, 1957 (Mexico, D.F.) p. 58. (Trans- lation supplied.) 64 Part 1. Problems of the Industry finances. Since Nacional Financiera has been a major source of finance for the electricity supply in Mexico, the basic question is clearly raised: can electricity earnings be attractive enough to enable Nacional Financiera to sell its own securities based, in part at least, on the earnings of electricity suppliers? That this is economically feasible in all the countries we have con- sidered is the theme of Part II of this study. And if this policy were adopted, the indexed security issue by Petr6leos Mexi- canos mentioned above would provide a precedent to be con- sulted in determining the nature of the issue. C. SPECIAL TAXES The relationship of taxation to electrical development is many-sided. In Chile, a tax on the consumption of electricity is not used for electrical development. In Mexico, the tax is both on and for electricity. It provides a major source of reve- nue for the investment program of the Federal Electricity Commission. In Brazil, the relationship between taxation and electricity is extremely complex. The Development Bank relies on non-electrical tax sources for some of its funds, and a por- tion of its receipts, in turn, are invested in the electrical in- dustry; various States have special tax measures to assist elec- trification; finally, the National Electrification Fund involves further special tax and expenditure arrangements. In Mexico, the Federal Electricity Commission (CFE) has been granted the proceeds of two different electricity taxes: a 10% surcharge on electrical bills, and the so-called "entero anual" of 2% on the capital value of assets used in the indus- try. The 100% tax has been of considerable, though declining, importance in the financing of Mexican electrical expansion. In the period 1939-50 the yield of this tax accounted for 12.25% of total financing of the Mexican electrical industry, for 17.6% of the domestic financing, and for 23.8% of the Financing Electrical Expansion 65 domestic financing of the public portion of the industry.40 The "entero anual" has been much less important as a source of funds. Suppliers may be authorized not to pay the entire sum due under this tax, or any of it, and payment has in fact been scaled down or forgiven in many cases in an effort to keep rates of return to suppliers up while final prices to consumers are kept down. The yield of the "entero anual" was a negligible 0.02% of total industry financing between 1939 and 1950.41 In Brazil, several States have levied special taxes on other commodities for electricity. For example, the State of Minas Gerais has granted a 4/14 (28.6%) share in the "State Re- habilitation Tax" to the State-owned power supplier and its subsidiaries. The rehabilitation tax was originally at the rate of 0.6% on sales and other transactions, and was raised in 1951 to 1.4%. CEMIG, the State-controlled "mixed corporation" which is the beneficiary of this tax, has estimated that the tax would contribute Cr. 6,057 million toward the Cr. 9,398 mil- lion total cost of its investment program between 1955 and 1964. CEMIG's depreciation accruals, earnings, and security sales were to provide only Cr. 592 million.42 The State of Rio Grande do Sul introduced a similar electrification tax, at the rate of 10% on all other taxes except export taxes, late in 1950.43 The Brazilian National Electrification Plan of 1953 involves a tax on electricity of the Mexican type to raise funds for elec- trical development. The Mexican tax on electricity is preferable to the Minas Gerais and Rio Grande do Sul taxes for electricity to the ex- tent that the former diminishes the artificial stimulus to in- creases in the consumption of a scarce commodity which re- sults from below cost rates at the same time as it provides 40 Percentages derived from Crist6bal Lara Beaiitell, La 7ndustria de Energia Electrica (Mexico, Fondo de Cultura Econ6mica, 1953), Table 40, p. 107. 41 Ibid. 42 Centrais Eletricas Minas Gerais (CEMIG) , Expansdo do Programa de Eletri- ficacdo do Estado de Minas Gerais (Nov. 1954), p. 34. 43 Comissao Estadual de Energia E1ltrica (CEEE), Relat6rio, 1952, pp. 8, 4, 21. 66 Part 1. Problems of the Industry funds for expansion of supply. Unfortunately, the CFE has never developed significant earning power of its own, and the evidence of Mexican electrical development suggests that CFE has used its receipts from the electrification tax, as well as from the general budget, as a means of rate subsidy rather than as a source of funds for expansion. The tax financing in the Brazilian States of Minas Gerais and Rio Grande do Sul is preferable to the absence of any financial plan in the original power program of the State of Sao Paulo. But the methods used in both States are inferior to the Mexican system because they do nothing to diminish the artificial stimulus, noted above, to growth in demand for elec- tricity. In addition, they place a burden on all transactions, including those involving country-dwellers who have no elec- tricity and will have none, for the benefit of certain industries and certain towns. The taxes may be defended as a means of providing capital for the state enterprises during a transitional period while they are developing independent earning power. As permanent arrangements, which they are apparently in- tended to be, they would subsidize heavy users of electricity at the expense of the rest of the population. Moreover, if such taxes are an important element in the budgets of public sup- pliers, they tend to gear the rate of electrical development to the flow of tax receipts and not to the rate of expansion in demand for the service. 4. Loans by International Lending Agencies Both the International Bank for Reconstruction and Devel- opment and the Export-Import Bank of Washington have added a very substantial flow of "official" international capital to the private international capital which began entering Latin Amer- ica even before World War I. By mid-1958, $395 million of the $798 million in loans made by the World Bank to Western Financing Electrical Expansion 67 Hemisphere countries-or practically 50%-were for electric power.44 Although the relative share of electric power loans was less for the Export-Import Bank, the absolute totals were nevertheless impressive: on June 30, 1958 at least $334 million in loans had been authorized for Latin American bor- rowers, $238 million had been disbursed and $197 million were outstanding.45 In addition, export credits have been available from both public and private sources in various countries to provide shorter-term financing for purchases of electrical equip- ment. These loans, a major part of which have gone to the five countries which we have studied, have been valuable in filling the vacuum left by the near-collapse of private international purchase of new obligations issued by Latin American enter- prises. But their great size relative to the total loans of the organizations providing them should not be allowed to obscure the basic comparison with the electric power industry's annual need for investment funds. This comparison will be based on three estimates of the Economic Commission for Latin America: (1) that electric power investment in all Latin America for the year 1955 amounted to the equivalent of $525 million; (2) that it will total about $5,400 million, at 1954 prices, over the period 1955- 1965; and (3) that the foreign exchange component of this ten-year outlay will be the equivalent of about $2,800 mil- lion.46 The sum of net World Bank loans and of Export- Import Bank Loan disbursements for Latin American power purposes was about $633 million in mid-1958. This sum, built up over a decade or more is not greatly in excess of actual and projected annual outlays. This ten-year total is only a little 44 See International Bank for Reconstruction and Development, Thirteenth Annual Report, 1957-1958 (Washington, D. C.), p. 16. 45 Compiled from Export-Import Bank of Washington, Report to the Congress for the Twelve Months Ending June 30,1958, Pt. 2, pp. 2-81. Loans have been in- cluded only if they are specifically listed as available to public-utility suppliers of electricity. 46 ECLA, Energy in Latin America, p. 14. 68 Part 1. Problems of the Industry more than twice the anticipated annual outlays in foreign ex- change that would be required for the projected expansion of Latin American electric facilities. Unless these lending agencies should expand their loans to Latin America for electric power purposes vastly beyond all previous experience, the bulk of total outlays must come from other sources. Even the rate of foreign-currency expenditures for electric power has been far above the net foreign-currency proceeds of these loans. But even if all the foreign exchange costs were covered, half the financial problem would still have to be solved. The ratio of foreign exchange to total costs in Latin American power financing is roughly 50%. Moreover, such a sudden access to foreign capital would leave the electric industry in many Latin American countries exposed to the problem of meeting a rising cost of capital on a current basis. These international loans have been, and would presumably continue to be, repayable in United States dollars or other currencies foreign to Latin America. The depreciation of a national currency in terms of the currency of payment would automatically raise the na- tional currency requirements of foreign debt service. In sum, the availability of international credit on a large scale still leaves a serious quantitative problem of public util- ity financing and does not eliminate the special difficulties that spring from inflationary pressures on exchange and price levels. Public international lenders can make important contributions to the financing of electric power in Latin America by supple- menting domestic capital, but they can neither replace it nor permit the industry to escape the choice between operation on a self-sustaining basis and dependence on government sub- sidization. And it must be remembered that the real bene- ficiaries of government subsidies are not the electric enter- prises which receive them but their customers, residential and business, to whom the subsidies are passed on in the form of low rates. How low these rates actually are is developed in the chapter which follows. CHAPTER IV Rates The Economic Commission for Latin America has emphasized the significance of electrical rates for the distribution of na- tional energy consumption, and for economic development generally. As for the influence of electrical rates on demand for electricity: When, from a technical point of view, there is a high elasticity of substitution among different forns of energy, the price is the decisive factor determining use. Govern- ments, through regulations, price control and rationing, play a key role in this sector, even when they do not intervene much in the production and marketing process. . . . The establishment and maintenance of a structure of relative prices-as between energy and other goods and services of a country, and, particularly, between different forms of en- ergy within the energy sector-determines . . . the volume and proportion of consumption of each form. If the price structure is inadequate, does not reflect faithfully the inter- relationship of technical and economic factors, it may pro- duce disturbances in the whole system of energy consump- tion.1 1 ECLA, Energy in Latin America, p. 39. In the recent study by a group of European experts, OEEC, The Trend of the Selling Price of Electricity and its Relation to Financing of New Plant (Paris, Nov. 1958), the same warning is given: "89. An economic policy purely designed to keep down the prices of energy arti- 69 70 Part 1. Problems of the Industry The Commission recognizes that in Latin America, the problem of increasing the supply of electric energy is basically a problem of providing a supply of capital for the industry. ECLA remarks, on the relationship between prices of electric- ity and the growth of supply: Electricity makes a comparatively small contribution to income, while requiring a large capital outlay, largely be- cause of the policy under which the rates are kept below the general price level, which results in a relatively low income ....2 . . . in the [Latin American] countries examined, the rela- tive contribution of energy services to national income tends to decrease, while at the same time they need increasing proportions of total invested capital.3 Thus electrical rates provide the focal point on which all the problems of the industry-regulatory policy, demand for the service, financing-converge. They are the final result of regu- latory policy, and the cause of ability or inability to finance expansion from within or to attract voluntary savings. In ad- dition, rates have an effect on levels of electrical consumption. The first two sections of this chapter will discuss the trend of electrical rates in Latin America, relative to the trend of other prices, and the level of electrical rates in Latin America relative to electrical rates in the United States. Of course, the ficially might distort competition between the various forms of energy and so lead to their being used for purposes other than those that serve the best inter- ests of the community." (p. 21.) On the constructive side, the report notes: "2. In- stead of seeking to stimulate demand, the policy of the electricity supply industry in most member countries is directed toward another goal, namely, by its pricing policy, to encourage both old and new customers to use electricity more efficiently and to better purpose rather than to create fresh requirements." (p. 9.) 2 ECLA, Energy in Latin America, p. 86. 3 Ibid. Cf. OEEC, op. cit. supra note 1 (p. 23): "109. Some difficulty has been experienced in recent years in bringing power supply and demand into balance, because of difficulties in financing investment." "110. The difficulty of raising capital has been felt mainly in countries where electricity undertakings are not yet authorized to sell at a price which covers not only production costs proper, but all other charges as well, including a fair return on the capital invested." Rates 71 latter comparison is subject to the uncertainties involved in the conversion of exchange rates. The two sections cannot, however, be complete in them- selves, for two reasons. First, as the Economic Commission for Latin America points out, a rational policy for energy con- sumption must relate prices to costs; therefore some attention must be paid to the relative costs of electricity supply in Latin America. This will be done in the third section of the chapter. Second, electricity, unlike many other commodities, is not sold at a flat price of so much per unit; indeed, the term 'flat rate' is usually employed to refer to a fixed charge per month to the customer without reference to the number of kilowatt-hours of electricity he uses. Therefore the general idea of the rate level for electricity must be further refined by some consider- ation of rate structure: relative charges for different classes of consumption, and for different amounts of electricity used by consumers of the same class. The idea of rate structure is implicit in the comparisons of typical monthly bills for various consumer categories to be found in Sections 1 and 2 of this chapter. It is examined explicitly, and in more detail, in Sec- tion 4. 1. The Trend of Electricity Prices in Latin America Rate data are available since 1929 for nine important elec- tricity supply areas in the Latin American countries studied. To facilitate comparison, these data have been used to con- struct typical monthly bills for specified quantities of monthly residential consumption of electricity.4 4 Monthly bills are used instead of average prices paid per kilowatt hour for two main reasons: the composition of consumption (industrial, commercial, resi- dential, etc.) may vary considerably from time to time or from place to place; and growth in consumption per customer will reduce average prices whenever the rate structure provides any of the usual concessions for increased quantities purchased. 72 Part 1. Problems of the Industry These monthly bills are contrasted with the United States weighted average bills for cities of over 50,000 population in Table 4.1. The data in this table do not include statistics for the most important cities in Brazil, Colombia, and Mexico. These are available for 1939, however, and are shown in Table 4.2. Table 4.1. Monthly Residential Electrical Bills in the United States and in Five Latin American Countries, 1929 and 1939 (in U. S. dollars) Nine Latin American Suppliersb United Statesa (weighted average) Median Low High 1929: 25 kwh .......... $ 1.73 $ 3.04 $ 1.25 $ 4.85 100 kwh .......... 5.13 7.51 4.53 19.42 250 kwh .......... 10.69 15.56 10.79 31.78 1939: 25 kwh .......... 1.40 1.36 0.68 1.96 100 kwh .......... 3.96 4.03 1.60 4.51 250 kwh .......... 7.21 7.75 3.21 10.33 a United States statistics are weighted average for selected cities over 50,000 population. b Latin American statistics are for the following supply areas: Brazil: Com- panhia Paulista (Baurd), Porto Alegre, Recife, Salvador (Bahia); Chile: Com- paffia Chilena (Santiago and Valparaiso); Colombia: Barranquilla; Costa Rica: San Jose; Mexico: Chihuahua, Torreon. Exchange rates used for conversion to U. S. dollars (local currency units per dollar): 1929 1939 Brazil ....... 8.48 16.50 Chile ....... 8.26 31.10 Colombia ....... 1.03 1.75 Costa Rica ..... 4.00 5.61 Mexico ....... 2.15 5.18 Sources: For electrical rates: Ebasco International rate files for Latin America; Moody's Public Utility Manual, 1954, for the United States. For ex- change rates: U. N. Statistical Yearbook. Rates 73 Table 4.2. Monthly Electrical Bills for 100 kwh Residential Consumption Selected Latin American Supply Areas, 1939 United States-weighted average ......... ........... $3.96 Median, 5 Latin American countries, Table 4.1 ....... 4.03 Brazil: Rio de Janeiro .................................. 3.82 Sao Paulo .................................... 3.20 Colombia: Bogota ..................... ............... 3.93 M edellin ........................................ 1.02 (1942) Mexico: Mexico City .................................... 1.79 Puebla and Veracruz .............................. 2.08 "For residence with Ps. 6,000 assessed valuation. Sources: Latin American rates: Company data. U. S. weighted average: Moody's Public Utility Manual, 1954. All of the residential rates in Table 4.2 fall below the median of Latin American undertakings in Table 4.1; four of the six are considerably below. The 100 kwh bill is, in gen- eral, fully representative of relationships at other levels of consumption. Moreover, the two preceding tables include supply areas containing the overwhelming proportion of the residential consumers and consumption in their respective countries in 1939. Therefore these two tables permit the follow- ing conclusions: (1) Latin American residential electrical rates were probably, on the average, well above the United States level in 1929. For the nine supply areas from which rate data are available for that year, median rates were about 50%0 higher. (2) By 1939, the median rates for these nine undertakings were very close to the United States weighted average. Data for 1939 for other major supply areas would reduce the Latin American weighted average to a figure below the nine-supplier median. Comparison of relative trends in Latin American and United States electrical rates and costs 74 Part 1. Problems of the Industry of living for the period 1929-54 can therefore be simplified by using 1939 as the base year. Cost-of-living indexes are available for Brazil, Chile and Mexico since 1929, and for Colombia and Costa Rica since 1939. Table 4.3 contains six supply areas whose "real" elec- trical rates (monthly bills for successive dates divided by con- temporaneous cost-of-living indexes) can be traced back to that year. "Real" residential rates for both 25 kwh and 100 kwh monthly consumption dropped in five of the six Latin American cases by more than the rather slight decline that took place in the United States. Salvador, Brazil, provides the only exception. "Real" commercial rates, which are not avail- able for the United States, also dropped in all the Latin Amer- ican supply areas except that of the Compafiia Chilena. "Real" industrial rates rose by 2% in the United States between 1929 and 1939, as compared with declines of over 50% in Recife, Brazil, and Chihuahua, Mexico, of a third in Rio de Janeiro, of almost a quarter in Santiago and Valparaiso, Chile, and of between five and ten per cent in Bauru and Salvador, Brazil. Thus the decline in residential rates in Latin America between 1929 and 1939 was no isolated phenomenon; except in the Compafiia Chilena supply area in Chile, commercial and in- dustrial rates dropped as much or more. Over the fifteen-year period 1939-54, every Latin American country studied, except Mexico, showed relative declines in electrical rates greater than those experienced in the United States, and Brazil, Chile, and Colombia showed much greater declines. In the decade 1939-49, "real" residential rates declined by almost 50% in the United States, but even this decline was ex- ceeded by a wide margin in all the Brazilian supply areas listed in Table 4.3, in the Compafiia Chilena area in Chile, in Bogoti, Colombia, and in Chihuahua and Puebla-Veracruz, Mexico. The drop in San Jose, Costa Rica was 57%. Only in Mexico City, and for 100 kwh residential consumers in the Compafiia Colombiana area in Colombia, was the decline in Rates 75 "real" residential rates between 1939 and 1949 less than in the United States. The Mexico City bill for 100 kwh residential consumption in 1939 was the equivalent of only $1.79, whiclh compares with the 1939 United States weighted average of $3.96. "Real" commercial rates in Latin America fell by al- most exactly as much as residential rates with two exceptions- the drop was less in Mexico City, and greater in San Jose, Costa Rica. Industrial rates fell in similar proportion. The "real" industrial rate decline between 1939 and 1949 was greater than that for the United States for every Latin Amer- ican supply area shown in Table 4.3 except Puebla-Veracruz, Table 4.3. Indexes of Trend of Real Electrical Rates, 1939= 100 ("Real Rates" obtained by dividing indexes of changes in monthly electrical bills by cost-of-living indexes) Industrial Residential Commercial 20 kw per 25 kwh 100 kwh 500 kwh 4000 kwh Brazil* Rio de Janeiro: 1929 . 162 162 162 152 1949 . 28 31 31 28 1954 . 18 20 20 18 Cia. Paulista, Bauru: 1929 . 109 108 108 107 1949 . 25 25 25 17 1954 . 15 15 18 13 Recife: 1929 . 132 132 135 268 1949 . 33 47 34 43 1954 . 21 29 21 30 Salvador (Bahia): 1929 . 86 75 106 106 1949 . 26 26 30 29 1954 . 13 19 21 17 Chile Cia Chilena: 1929 . 147 145 97 131 (Santiago, Val- 1949 . 29 29 28 34 paraiso, etc.) 1954 . 12 21 22 32 (Table 4.3 continued on Page 76.) 76 Part 1. Problems of the Industry Table 4.3 (Cont'd) Industrial Residential Commercial 20 kw per 25 kwh 100 kwh 500 kwh 4000 kwh Costa Rica Cia Nacional: 1949 ......43 43 26 28 (San Jose) 1954 ......43 42 25 28 Colombia Cia Colombiana: 1949 ......50 67 59 50 1954 . 36 50 48 42 BogotA: 1949 . 38 38 38 38 1954 . 26 26 26 26 Mexico Mexico City: 1949 . 59 60 80 47 1954 . 61 65 73 82 Chihuahua: 1929 . 234 270 322 232 1949 . 29 29 29 29 1954 . 27 26 33 38 Puebla-Veracruz: 1949 . 32 36 38 64 1954 . 40 43 68 69 United States Weighted average: 1929 . 103 107 - 98 (cities over 1949 . 52 54 - 51 50,000) 1954 . 48 48 - 51 Brazil: Cost-of-living index used as divisor: Rio de Janeiro, 1929-39; Sao Paulo, 1939-54. Rio de Janeiro power rates are based on average prices per kilo- watt-hour for industrial power. Mexico: Divisor is wholesale price index, 1929-39; retail price index, 1939-54. Sources: Ebasco International (American & Foreign Power Co.) rate files, except for Brazilian Traction rate material for Rio de Janeiro; Mexican Light rate books for Mexico City; Moody's Public Utility Manual, 1956, pp. a-13 and a-14, for the United States. Mexico, and much greater in such capital cities and important industrial centers as Rio de Janeiro, Brazil, Santiago, Chile, and Bogota, Colombia. Between 1949 and 1954, the "real" indexes had practically stabilized in the United States, with a drop of less than 10% in residential rates and complete relative stability for indus- trial users. But a further precipitous decline occurred in Brazil, Chile, and Colombia. Prices for electricity remained stable Rates 77 relative to other prices only in Costa Rica, where prices in general were quite stable, and in Mexico. 2. Comparative Electrical Rates in 1954 Latin American electrical rates, which were relatively high in 1929, had become remarkably low by 1954. This becomes evident when Latin American rates are compared with rates charged in the United States, and is still more obvious if the comparison is made for the general trend of Latin American prices. Fortunately, Latin American rate material for 1954 is much more abundant than the data available for earlier years. Therefore Table 4.5 presents monthly bills for all cities of over 50,000 population in Colombia and Mexico, and for a number of smaller towns in each of these countries, as well as for the most important supply areas in Brazil, Chile, and Costa Rica. By 1954, official exchange rates in Colombia, and especially in Brazil and Chile, had become thoroughly unrealistic. They must be corrected to allow for the extent to which internal inflation, relative to that in other countries, had outstripped external devaluation. The "purchasing power parity" exchange rates used in this study are those employed by the Economic Commission for Latin America in its energy study. Table 4.4 compares them with contemporaneous official exchange rates. Costa Rican electrical bills have been converted by using the official 1954 exchange rate of 5.67 colones per dollar and the free rate of 6.65. Rate comparisons with the United States for Brazil, Chile, Colombia and Costa Rica, using the foregoing bases of con- version, are presented in Table 4.5. A separate table, Table 4.6, compares Mexican rate levels in greater detail. Mexico had, and has, only one peso-dollar exchange rate and does not employ exchange controls; therefore, in Table 4.6, Mexican electrical bills have been converted into U. S. dollars at the official exchange rate only. Since the supply areas listed in Tables 4.5 and 4.6 account 78 Part 1. Problems of the Industry Table 4.4. Official and Purchasing Power Parity Exchange Rates, Brazil, Chile, and Colombia, 1954 (Local currency units per U. S. dollar) Exchange rate Brazil Chile Colombia Official ................... 25.82 110.00 2.51 Purchasing power parity ............ 38.3 245.05 3.13 Lowest available rate of Cr. 18.82 plus minimum agio of Cr. 7.00. Source: ECLA, Energy in Latin America, p. 264. for almost all the electricity consumed in Colombia and Mex- ico, and for over half the consumption of Brazil, Chile, and Costa Rica, these tables provide the basis for an estimate of the relationship between weighted average electricity bills for given levels of consumption in the five Latin American coun- tries, and in the United States, in 1954. This estimate is given in Table 4.7. Table 4.5. Monthly Electrical Bills in Brazil, Chile, Colombia and Costa Rica, in U. S. Dollar Equivalents, mid- 1954 Purchasing power parity exchange rate Residential Commercial Industrial 25 kwh 100 kwh 500 kwhi 400 k%vh/20 kw Brazil Rio de Janeiro ........ $0.60 $2.62 $13.04 $52.33 Sao Paulo ............. 0.45 2.02 10.10 49.63 Bauru (Cia. Paulista) .. 0.81 2.27 15.44 51.64 Recife ................. 1.12 4.38 21.77 97.43 Salvador .............. 0.74 2.92 15.14 65.27 Chile Santiago, etc . .......... 0.29 1.97 13.42 52.04 Colombia Barranquilla .......... 1.35 4.18 11.74 89.40 Bogota ................ 0.72 2.09 15.98 35.14 Medellin .............. 0.84 1.21 12.00 55.12 9 other cities over 5,000: Median ............. 0.80 2.96 10.89 57.51 Rates 79 Range: Low .............. 0.32 1.35 5.60 26.85 High .............. 1.24 4.00 24.28 84.33 21 towns under 50,000:a Median ............. 0.90 2.99 11.18 55.92 Range: Low .............. 0.32 1.28 5.60 19.98 High ............. 1.75 6.38 18.54 130.52 Costa Rica San Jose .............. 0.69b 1.52 7.29 49.43 United States (U. S. Dollars) Cities over 50,000c ..... 1.30 3.73 18.75d 105.770 Tacoma, Wash.t ....... 1.00b 1.70 7.03d 42.400 TVA Supply Areaf ...... 0.75b 2.50 11.75d 52.00e *"Purchasing power parity" exchange rate employed for Costa Rica-free ex- change rate. a All towns with metered power for which rate data are available. Excludes towns served by Cia. Colombiana, whose rates are already included with cities. b Minimum charge. c Average of weighted averages for January 1, 1954 and January 1, 1955. d Estimate based on interpolation. Average price per kwh for consumptions of 3 kw/375 kwh and 6 kw/750 kwh multiplied by 500 kwh. This slightly understates U. S. rates. e Minimum estimate. Represents two-thirds of average of 30 kw/6000 kwh bills for January 1, 1954 and January 1, 1955 for U. S. cities. f January 1, 1955. Also see notes (d) and (e) for adjustments. For sources, see Table 4.6. For residential consumers: where fixed or minimum charges exist, they were converted into kwh charges on the following bases: (1) 25 kwh monthly consumption was assumed to be associated with three rooms, 500 watts maximum demand, five electrical outlets, one wire (Mexico), or 4,500 pesos assessed value prior to 1954 and 9,000 pesos in 1954 (Medellin, Colombia). (2) 100 kwh: five rooms; one kw maximum demand; seven electrical outlets; two wires, 5,000 and 10,000 pesos assessed value. For commercial consumers: five kw maximum demand; 20 toutlets; three wires (Mexico). No night or special use assumed at low rates. For industrial consumers: 1.33 horsepower equals one kw corrected connected load equals registered maximum demand. Low-tension use. No night or special use. No special residential heating, cooking, or refrigeration rates were used in the calculations. Rates not available to new consumers were also excluded. No com- mercial or industrial rates for special uses or users have been included. 80 Part 1. Problems of the Industry Table 4.6. Monthly Electrical Bills in Mexico, 1954, in Dollar, Equivalents (official exchange rate of pesos 12.49/ U. S. $1.00. No free rate) Residential Commercial Industrial Mexico 25kwh lOOkwh 5OOkwh 20kw/4000kwh A. Cities over 50,000 (Number served by each system in parentheses): 1. Mexican Light & Power System (4) .. 0.73 2.29 12.24 83.30 2. American & Foreign Power Systems: a. Puebla-Veracruz (4) ...... ....... 0.63 1.74 8.76 38.36 b. Torreon (1) ........ ............ 0.56 1.61 7.60 44.23 Chihuahua (1) ....... ........... 0.40 1.14 5.45 34.20 c. Guanajuato (2) ....... .......... 0.83 2.26 10.30 34.69 3. Chapala Co. System (1) (govt. owned) 0.62 2.03 10.68 40.64 4. Monterrey Co. (1) ....... .......... 0.44 1.26 5.24 37.60 5. Companies in U. S. border cities (4): Unweighted average ....... ......... 0.59 2.04 11.33 49.12 Range: Low .......... ............. 0.49 1.62 10.76 40.48 High .......... ............. 0.68 2.63 11.86 60.80 6. American & Foreign Power Co.-iso- lated subsidiaries (5): Median ............. ............. 0.88 2.12 10.43 54.36 Range: Low ......... ........... 0.67 1.76 9.22 30.38 High ......... ............ 1.02 2.62 10.82 66.99 7. CFE: Jalapa (1) ............. .... 0.28 0.89 - 26.26 S. Morelia Co. (1) (govt. owned) 0.54 1.16 5.92 36.96 B. Towns under 50,000 1. CFE (33;31 with power rates) Median ...........................76 2.31 11.77 62.32 Range: Low0 .....................50 1.30 4.21 31.40 High .......................52 4.00 15.68 99.20 2. Other (29 with power rates, of which 5 govt. owned): Median ....................0.72 2.09 8.09 47.00 Range: Low ............. ..... 0.42 1.08 3.00 25.60 High ......... ............ 1.28a 3.37 15.22 90.27 United States (cities over 50,000) ...... 1.30 3.73 18.75b 105.77b a Minimum charge. b See preceding table, notes (c) and (d) . For sources, see p. 81. Rates 81 Table 4.7. Estimates Weighted Average Electricity Bills, 1954, Five Latin American Countries Relative to the United States (United States = 100) Exchange rate Purchasing power Official parity TBrazil ......... 80 to 110 55 to 75 Chile .110 to 150 50 to 70 Colombia ......... 60 to 110 50 to 85 Costa Rica .50 to 60 40 to 50 Mexico. 45 to 60 45 to 60 United States .1 0 100 100 As these estimates indicate, the comparison is rather irregu- lar for different types of customer and for different quantities of consumption. But two general conclusions emerge: in some categories, weighted average rates in the five Latin American countries were only half, or less, of the United States average, while in almost every case the Latin American rates were less than three-quarters of the U. S. weighted average. A fair over- all summary figure would be 60%, or less, for Latin American rates relative to the United States. Moreover, although the range of Latin American bills is extremely great, very few are above the United States weighted average for any type of consumer or level of consumption. In the smallest residential category, Barranquilla, Colombia, is the only city of over 50,000 population whose rates are avail- Sources for Tables 4.5 and 4.6: FOT Brazil, rate files of Ebasco International (American & Foreign Power Co.) and Brazilian Traction; for Chile and Costa Rica, rate files of Ebasco International; for Colombia, material supplied by Instituto de Aprovechamiento de Aguas y Fomento Electrico; for Mexico, rate files of the Comisi6n Federal de Electricidad. For the United States, Federal Power Commission, various rate publica- tions. 82 Part 1. Problems of the Industry able with a bill above the United States level at "purchasing power parity" exchange rates, and even in this case the differ- ence is minor. Part of the explanation for the consistent ap- pearance of lower Latin American bills is the absence, or low level, of meter rents, fixed charges, and minimum bills in most Latin American supply areas. But even for the 100 kwh resi- dential consumption group, only Barranquilla, one other Co- lombian city, and Recife, Brazil, are above the United States weighted average. For 500 kwh commercial consumption, Recife and one Colombian city are the sole representatives of bills higher than the United States weighted average. Finally, for power, every single Latin American bill in a city of over 50,000 is below the United States weighted average, and the same is true for the smaller towns of Colombia and Mexico with one exception in the former country. Table 4.5 also presents monthly bills for Tacoma, Washing- ton, and the low "standard" retail rates in the Tennessee Valley Authority supply area as examples of the lowest rates charged in the United States. Tacoma had the lowest United States rates in almost every category in 1954 and 1955. Yet every Latin American country listed in Table 4.5 provides one or more bills below the Tacoma level for at least one category of service. All of the three countries-Brazil, Colombia, and Mex- ico-from which more than one set of rates were obtained show at least one bill below the Tacoma level for each category of service listed. As for bills in the TVA area, these would be close to both the medians and the weighted averages for most Latin American countries. The very lowest rates in Latin America crop up without relation to geography: they are scattered almost at random over the countries studied, and range from Santiago, Chile, to northern Mexico. The very lowest rates also appear in some of the smaller towns of Colombia and Mexico-the only two countries for which a wide selection of rates is available-as well as in some of the largest cities. Rates 83 3. Electricity Rates and Costs: Latin America and the United States Are the differences in electric rate levels between the United States and the Latin American countries studied to be ex- plained in terms of higher costs in the United States? An analy- sis of relative costs may be based on cost trends through time, on relative cost trends in different areas, or on relative cost levels in different areas at any one time. All of these cost meas- ures are relevant to the price discussion in the preceding sec- tions, and all will be used to some extent. Although the cost of supplying electricity is affected by many factors, only a few are important. Among the most significant, for purposes of a comparison of Latin America and the United States, are the following: (1) relative costs of capital; (2) rela- tive efficiency in generation and technical progress; (3) costs per kilowatt for new generating capacity (reflecting differences in the inherent qualities of different hydro-electric sites and transportation costs and wage rates as well as technological differences) ; (4) relative ability to reduce costs per kilowatt- hour by spreading capacity costs over more output via higher load factors; (5) relative ability to reduce costs per kilowatt- hour by spreading customer costs over more output via higher use per customer. A. RELATIVE COSTS OF CAPITAL The cost of capital is a major factor in the electric industry because of its high ratio of investment in plant to revenues- roughly four times that of the average manufacturing indus- try.5 Moreover, capital costs are especially heavy in hydroelec- 5 In 1954, the ratio of privately owned electric utility plant per dollar of clectric revenue in the United States was 4.29. Federal Power Comm'n, Statistics of Electric Utilities in the United States, 1954. Privately Owned Companies (Washington, D. C.), p. x. 84 Part 1. Problems of the Industry tric generation, the mode of generation on which the Latin American countries studied chiefly rely.6 Since Latin American economies have been developing rap- idly, the demand for capital has been high, and it has been in scarce supply. Naturally, rates of interest and returns expected on equity capital have been much higher than in the United States. Lack of capital markets in the countries studied in which potential investment uses can compete freely for com- mand over the flow of national savings makes statistical com- parison difficult, but such evidence as is available leaves no doubt that capital costs in the United States are much lower. The debt securities of electric utilities in the United States were yielding only a little over 3% in 1954, the year as of which the rate comparisons in Section 2 of this clhapter were made.7 In Mexico, where inflation has not dried up the market for long-term fixed-interest obligations, as it has in Brazil and Chile, the interest rate for the obligations of mortgage banks and other credit-worthy debtors has been 8%. In Costa Rica, during the period of price stability and government surpluses after 1948, prices of government bonds advanced in a rela- tively free market until yields were reduced to 6% to 6½2%, but this was at least twice as high as the yields on United States Government bonds in the same period. In the absence of distorting factors the cost of equity capital, because of the greater risk involved, should be distinctly higher than the cost of senior debt securities. In the United States, earnings available for common stocks in electric utilities aver- aged 10.5% after taxes in 1954.8 Obviously, the cost of capital 6 The ratios of thermal to total electric generation in terms of kilowatt-hours consumed, were the following in 1954: Brazil, 19.7%; Chile, 43%; Colombia, 32.6%; Costa Rica, 16%; Mexico, 56.1%. Computed from data in ECLA, Energy in Latin America, Table 15, p. 31. In the United States in recent years, hydro has represenited about 21% of all electricity generated. 7 Federal Power Commission, op. cit. supra note 5, Table 5, states the average interest rate on long-term debt in 1954 to have been 3.2%. 8 Ibid. Market prices of utilities stocks had risen well above the book valtie of the shares so that the return to the investor, if computed on market value, was considerably below 10.5%. Rates 85 would have to start well above this level to attract fresh equity investment in Latin America if a market for such capital could be re-established. If the returns actually earned on equity cap- ital in electric utilities in Latin America have tended to be lower than those earned in the United States, that does not mean that the cost of capital is lower in Latin America. Rather, it means that the cost of capital there is not being fully met. Is the higher cost of capital in Latin America being offset by greater efficiency of generating plants, by lower costs per kilowatt for new generating capacity, or by lower customer costs? These possibilities will be examined briefly in succeed- ing sections. B. RELATIVE EFFICIENCY IN GENERATION AND TECHNICAL PROGRESS The cost of generating electricity varies widely from installa- tion to installation. An important factor in the case of hydro- electric plants is the extent to which effective generating capa- city is below installed capacity. On this point, the Economic Commission for Latin America reports:9 "With regard to Latin America, the information available is very inadequate, although the utilization factor is known to be low. In other words, owing to seasonal variations in river flows and the lack of reservoirs, constant power amounts to only a fraction of installed power. In most cases the factor fluctuates around 30 to 40 per cent." In the United States, the widespread availability of low-cost fuel has made it possible to rely on hydroelectric installations chiefly where conditions for generation and use are relatively favorable or where the costs of an installation can be shared by one or more additional purposes, e.g., flood control and navi- gation. 9 ECLA, Energy in Latin America, p. 67. 86 Part 1. Problems of the Industry The efficiency of thermal plants in the use of fuel has been steadily increasing, but, as ECLA points out,'0 "Latin America is often tardy in introducing technical advances, mainly for financial reasons." "An examination of the situation in various countries," the Commission continues, "shows that efficiency in thermo- electric generation is, with few exceptions, very low in Latin America. Compared to average efficiencies of 21% to 25% in Western Europe and 30O% in the United States-and even higher in the more efficient power stations-efficiency in Latin America is 15%, to 18% and even less. Furthermore, this deficiency is not confined to the less industrially devel- oped Latin American countries but is also seen in the most advanced." The handicaps indicated by the foregoing cannot be over- come in Latin America by virtue of lower labor costs since, in the case of hydroelectricity, labor costs are minimal" and, in the case of thermal electricity, the higher labor costs per employee in the United States can be spread over a much larger output.12 But do the Latin American utilities have any advantage in equipment costs? Here, again, the evidence indicates higher Latin American costs. Most capital equipment for electrical expansion has al- ways been imported into Latin America, and still is. Local manufacture of electrical equipment is beginning to replace certain types of imports in Brazil and Mexico, but neither country has yet become a pioneer in any phase of electrical manufacture; on the contrary, the domestic industries are pro- 10 Id., p. 76. 11 In the production costs of United States hydro installations, labor costs typically range from 0.20 mills to o.3o mills per kwh. Ordinarily maintenance cost (in which labor may be an important factor) is lower still. See Federal Power Comm'n, Hydroelectric Plant Construction Cost ancl Annual Production Expenses, First Annual Supplement, 1957. 12 Sixteen predominantly thermal electric systems in the United States (with total electric sales of more than 63.000,000 mkwh) showed a cost for salaries, wages and benefits in production expense of only 0.70 mills per kwh. Based on data compiled by the American & Foreign Power Co., Inc., in cooperation with the 16 systems. Rates 87 tected by exchange controls (in Brazil) and by high tariffs (in Mexico) . The rate of technological progress in the Latin American countries studied is reduced by their greater reliance on hydroelectric generation, which reached high technical efficiencies long ago and cannot match the continuing improve- ments in efficiency of thermal plants. Moreover, the pressure of power demand in Latin America has compelled obsolete plant to be retained in service. C. COSTS PER KILOWATT FOR NEW GENERATING CAPACITY These costs are partly determined by relative technical progress, but they are also affected by relative wage rates, transportation, the characteristics of different hydroelectric sites, etc. Table 4.8 summarizes cost evidence compiled by the United Nations Economic Commission for Latin America. Table 4.8. Capital Costs per Kilowatt of New Generating Ca- pacity in Selected Latin American Countries, 1954 (U. S. dollars) Hydro Steam No. of Cost range per No. of Cost range per Projects kilowatt projects kilowatt Brazil ............... 8 $2524508 1 $194 Colombia .......... 3 288- 850 - - Costa Rica ......... 1 305 - - Chile ............... 6 162- 336 2 284- 333 Mexico ............ 7 146- 685 7 137- 486 Source: ECLA, Energy in Latin America, App. XIV. Tables 4, 5, pp. 262-267. Purchasing power parity exchange rates were used in conversion, as in Table 4.4, supra. Weighted average costs for the projects summarized in Table 4.8 are: Brazil, hydro, $346; Chile, hydro, $241, and steam, $290; Colombia, hydro, $349; Mexico, hydro, $209, and steam, 88 Part 1. Problems of the Industry $179. The Mexican hydro average is influenced by an exten- sion of an existing plant, and dominated by run-of-river installations; in terms of costs of firm power, the Mexican averages would be considerably higher. These weighted averages may be compared with the 25 most thermally-efficient steam plants in the United States for which capital costs per kilowatt are available, as of the end of 1954. All were completed during the 1950's. The range of capital costs for these plants is from $121 to $227 per kilowatt, with both a median and a weighted average of $160 per kilowatt.13 Even assuming an identical rate of return in the United States and in Latin America, these United States plants could produce more cheaply at normal load factors than those of any Latin American country studied, with the possible exception of hydro plants in Chile. Moreover, the economies of large generating plants are still far from realization in Latin America outside of a few metro- politan areas. The average size of U. S. public utility generating plants is at least twenty times as large as the average for Brazil, Chile, Colombia, or Costa Rica, and at least ten times as large as the average for Mexico. Although the Latin American averages are pulled down by a multitude of very small plants which do not add much, as a group, to output, some of these small plants belong to enterprises which charge very low rates. For example, the lowest power bills reported in Table 4.6 for Mexico were charged by independent suppliers, with small plants, in towns of under 50,000 population. Thus, in terms either of a comparison with costs of new gen- erating plants in the United States, or of general reasoning from the economies of plant size, most areas of Latin America are at a disadvantage. This disadvantage has certainly not been reduced enough over the last ten or twenty years to explain the decline of Latin American electricity bills relative to other Latin American prices or to electricity bills in the United 13 Federal Power commission, 35th Annual Report, 1955, Table 1, p. 154. For data showing the capital costs of thermal and hydro generating plants and of total systems in Europe, see OEEC, op. cit. supra note 1, Tables 5, 6. Rates 89 States. The Latin American emphasis on hydroelectric devel- opment may even work against future cost reduction: many of the low cost hydro sources have already been developed, so new sites may involve either higher costs per kilowatt of gen- erating capacity or additional expense for long transmission lines to load centers. D. LOAD FACTORS Load factor (the ratio of average actual use to maximum use) has special importance for electrical costs per kilowatt-hour both because of the high percentage of capital to total costs in the supply of electricity and because of the technical necessity to produce electricity at the same time as it is used. An in- crease in load factor, by increasing the ratio of average actual use to maximum use, correspondingly reduces the cost per kilowatt-hour of plant which must in any case be available to supply maximum demand. Load factors in the United States and in Latin America must be interpreted differently. The bulk of the electricity supplied in the Latin American countries studied is hydro-electricity. Much of this is derived from run-of-river hydro plants, or from plants with storage reservoirs but without substantial inter- connected thermal capacity. Under these conditions, improve- ments in load factor may be limited by variability in supply as well as fluctuation in demand. Moreover, load factors in the United States have increased over the last twenty years due to more widespread interconnection of generating plants and to increased load diversity. Improvements in many Latin Amer- ican supply areas have been achieved because of the unfortu- nate necessity to chop off demand peaks due to inadequate capacity. Load factors are not conspicuously above the United States average anywhere in Latin America.14 14The ECLA study indicates that, though the load factor is "constantly ris- ing" in Latin America, it is still below the United States. ECLA, Energy in Latin America, p. 78. 90 Part 1. Problems of the Industry E. CUSTOMER COSTS Increased use per customer spreads certain costs, from meter- reading to general distribution outlays, over more kilowatt- hours. If additional sales yield additional profits, then this de- crease in customer costs per kilowatt-hour may permit rate reductions. Here, again, a comparison of Latin America and the United States is unfavorable to Latin America, for two reasons: 1. In Latin America, the growth of urban population and the extension of electrical service to urban dwellers who pre- viously received no supply account for a relatively large per- centage of total growth in consumption, whereas in the United States the growth in number of electricity consumers is a very small element in total growth. 2. Latin American consumptions per consumer are generally far below the average for the United States; even though Latin American electrical rates are much lower, Latin American in- comes are lower still. In Medellin, Colombia, and San Jos6, Costa Rica, average residential consumptions are 50% above the weighted average for the United States. But in the "nor- mal" metropolitan area in the countries studied, residential consumption per user is only half as great as in the United States. This ratio holds for the Brazilian Traction supply area, for Bogota, Barranquilla, and Cali, Colombia, and for Santiago and Valparaiso, Chile. Average consumption in smaller towns is considerably lower, and average consumption per residence in Mexico City is only about 30% as high as in the United States. Average consumptions per commercial and industrial user are also considerably lower in every Latin American supply area reporting data. From the standpoint of customer costs, Latin American sup- ply areas have only two advantages over the United States: serv- ice tends to be confined to thickly-settled urban areas, which may yield a high density of electrical use per unit of distribu- Rates 91 tion facilities even if use per customer is low; and wage rates are much lower. But both of these advantages should be par- ticularly important in reducing the costs of low-voltage distri- bution, the type of service which involves the largest outlays of labor as an operating expense and on distribution capital per kilowatt-hour sold. Yet the lowest relative Latin American monthly bills discussed in Section 2 of this chapter were for industrial rather than for residential service-i.e., for precisely that group of users who are least affected by distribution costs. Labor is a more important factor in customer costs than in production costs. However, greater efficiency in the use of labor and especially the ability to reduce labor costs per kilowatt- hour by spreading them over a much larger consumption of electricity per customer has enabled electric utilities in the United States to hold labor costs down, despite rising salary and wage scales. Although it is difficult to ascertain and com- pare labor costs with exactitude, such data as have been ob- tained indicate that salaries and wages (including benefits), though lower per employee in Latin America, constitute a higher cost per kilowatt-hour in the total of operating costs exclusive of production and transmission costs and taxes. These costs (which comprise administrative and general expense, dis- tribution expense, customer accounting and collection expense and sales promotion expense) also appear to be higher when not merely the labor cost element but the totals of these ex- penses are compared.'5 15 Labor costs (including benefits) as a part of all costs other than produc- tion, transmission, and taxes, were as follows for American & Foreign Power Co. subsidiaries in 1957: Brazil, 2.20 mills per kwh; Chile, 3.56 mills per kwh; Colombia, 1.83 mills per kwh; Costa Rica, 2.69 mills per kwh; and Mexico, 2.17 mills per kwh. In the United States, 17 exclusively electric systems had an aver- age labor cost of 1.93 per kwh for the same categories of expense in 1958 (when salary and wage scales had risen somewhat above 1957). Total operating costs in the same categories averaged 3.29 mills for the same United States systems in 1958. In 1957, the corresponding figures for the same Latin American subsidi- aries were as follows: Brazil, 4.00 mills per kwh; Chile, 5.31 mills per kwh; Colombia, 3.29 mills per kwh (equal to the 1958 United States level) ; Costa 92 Part 1. Problems of the Industry One final aspect of customer costs should be noted. A fa- miliar rule-of-thumb, in the United States, is the expenditure of a dollar on distribution plant for every dollar spent on thermal generation. This ratio would not hold in Latin Amer- ica, owving to the importance of hydro, but this fact alone cannot explain the extremely lop-sided ratios-5:1 and even higher-of generation to distribution investment in many Latin American supply areas. Just as load factors on generating plant have been improved by lopping off the peak of the load, so distribution costs have been held down by failures to expand capacity with growth in demand. This is compatible witlh cheap service, but it is not compatible with high-quality service. # * * This consideration of cost factors has unearthed no outstand- ing improvement in Latin American electrical supply whichi would account for the precipitous downward trend of "real" electrical rates, nor has it exposed any obvious reasons why electricity should be any cheaper in Latin America than in the United States. On the contrary, it has shown that the United States enjoys a net cost advantage. Despite various expedients employed to mitigate the effect of low rates on particular suppliers, the result has been similar in every country. "Cheap" electricity has produced low earn- ings and poor service. The effective cost to the economy has often been very high in terms of inconvenience, delay, and damage, plus the extra cost of private plant. Rica, 4.11 mills per kwh; and Mexico, 4.29 mills per kwh. These costs were based on data supplied by the American & Foreign Power Co. from its own sources and 1958 data compiled by it from 17 cooperating electric systems in the United States (with total electric sales of nearly 65,000,000 mkwh). Conversion of Latin American costs into dollars was at official exchange rates prevailing on December 31, 1957. These rates were more representative of internal costs than the 1954 rates which had compelled us to resort to the pur- chasing power parity basis for conversion. Of course, if the exchange rates used overstated the internal purchasing power of the national currencies, they would, to that extent, overstate the Latin American operating costs noted above. Rates 93 4. Rate Structures Since the benefits of low or subsidized electric rates are not equally distributed among electric consumers, it is important to examine rate structures. This discussion has two aspects: the relative distribution of the costs of electricity supply among different consumer classifications (e.g., industrial, commercial, residential, etc.), and the presence, or absence, of "promo- tional" rate structures available to any one consumer group. Industrial and commercial rates in Latin America vary in detail, but they do not involve any novel principles. The "Hop- kinson" tariff, which combines a maximum demand charge with an energy charge, is the standard for power charges in Latin America as it is elsewhere. Residential rates, on the other hand, deserve special attention. They sometimes take unusual forms in Latin America, they affect by far the largest group of consumers, and their structures can sometimes have a pronounced effect on the total consumption of electricity. Therefore, this section will briefly discuss certain general as- pects of industrial rates and then concentrate on the special problems of residential rate structure. A. RATES FOR INDUSTRIAL POWER Industrial power rates in Latin America raise two separate issues: the general issue of adequacy as a source of funds to finance public utility expansion of facilities as an alternative to private plants; and the special issue of relative prices to be charged industries using large amounts of electricity. The general issue of adequate funds is, of course, the theme of this entire discussion. It will be examined here only to the extent that it relates to the rate structure of public utility sup- pliers of electricity, on the one hand, and to the cost structure of power-users, on the other. 94 Part 1. Problems of the Industry The first point to be emphasized is the low average ratio of electricity costs to gross value of industrial output or value added by manufacture. This is evident in Table 4.9, in which, it is to be noted, the ratio of electricity cost to gross value plainly shows how small an impact on the price of commodities would result from even a sharp increase in industrial rates for elec- tricity. And it should be observed that these data do not re- flect the fall in industrial rates relative to prices which Table 4.3 shows took place in several of the countries studied after 1949. Table 4.9. Cost of Electricity in Relation to Value Added by Manufacture and Gross Value of Manufacturing Output Percent electricity cost to Value Gross Country Year added value Brazil: 1939 .................... 2.78% 1.23% 1949 .................... 1.37 0.60 Chile: 1946 .................... - 1.1* Colombia: 1945 .................... 2.43 0.55 Costa Rica: Oct. 1950-Sept. 1951 ...... 0.92P 0.30v Mexico: 1945 .................... 2.11 0.63 United States: 1939 .................... 1.89 1947 .................... 1.28P - OEEC nations: 1953 .................... 1.8** 1.1 *Percentage of total production costs. "Percentage of gross national product. PPurchased electricity only. Sources: Brazil: Anuario Estatistico do Brasil, 1949, p. 165, and 1953, pp. 124-126. Chile: ECLA, Economic Survey of Latin Arnerica, Annex i, Industrial Development in Chile (mimeographed), App. 5. Colombia: Direcci6n General del Censo, Primer Censo Industrial de Colombia-1945. Resumen General, pp. 1o5, 135. Costa Rica: Ministerio de Economia y Hacienda, Censo de Comercio e Industrias de 1952, passim. Mexico: Anuario Esta- distico 1951-52, pp. 651 0. U. S.: Dept. of Commerce, Bureau of Census, Census of Manufacturers, 1947, Indexes of Production, App. B, p. 66 and Fuels and Electric Energy Consumed, p. 3; OEEC: The Trend of the Selling Price of Electricity (Paris, Nov. 1958), p. 17 and Table 34. Rates 95 Table 4.10. Ratio of Cost of Electricity to Value Added by Manufacture, Brazil and Mexico Brazil, 1949 Mexico, 1945 Food ..................... 1.31% Beverages ......... ........... 0.74 Tobacco ..... ............. 0.33 0.27 Textiles ................. 1.49 1.33 Clothing ................. 0.56 0.74 Leather ................. 1.11 1.20 Wood and furniture ............. ..... 1.28 1.22 Paper and products ........... ....... 3.32 5.84 Printing and publishing .................. 0.89 0.95 Rubber products ............... ............ 1.27 4.01 Chemicals and drugs ........... ............. 1.21 1.25 Non-metallic minerals .......... ............. 1.46 2.43 Metallurgy, metal-working ........ ........... 1.80 1.97 Electrical ................................... 1.55 2.00 Miscellaneous ..................... .......... 1.50 0.61 Sources: Census of Production, Mexico, 1945, and Brazil, 1949-50. Aside from the high percentage for electricity in Mexican food-processing, probably caused by the widespread use of elec- tricity for corn-grinding mills, both the Brazilian and the Mexican figures show much higher percentages for electricity in the "new" industries than in the "old." But even the highest figure in the table would be much too low for aluminum re- duction, or for specialized electro-chemical and electro-metal- lurgical operations. There is danger that publicly-financed suppliers will be under extra pressure to grant particularly low rates for par- ticularly large-scale power users even when the average level of rates is inadequate to cover costs, and power shortages are holding back industrial development which could add much more to the total value of manufacturing output. Unusual stress on the wvonders to be performed by electricity, especially by means of its use in a few industries which are very greedy 96 Part 1. Problems of the Industry users, neglects the fact that the cost of hydroelectricity is mainly the cost of capital. If capital scarcity is one of the main factors holding back the rate of economic development, lavishing capital on a few industries is an almost certain method of slowing this rate. Finally, low rates for public utility supplies of electricity may be accompanied by great cost discrepancies among differ- ent manufacturing concerns even if they are apparently subject to the same electricity tariffs. Even if all users have the same load factors, use electricity at the same times, etc., cost differ- ences can arise if the public utility supply of electricity is limited and of low quality. The Brazilian Traction supply area supplies examples of both the qualitative and quantitative problems. Brazilian Trac- tion, before World War II, had always provided the most abundant supply of power obtainable in Brazil, and was prob- ably an important factor in the. unusually rapid industrial de- velopment of its supply area. By the early 1950's, its capacity for raising funds was swamped by the growth in demand for electricity. In 1952 and 1953, circuits had to be disconnected without warning, and both voltage and frequency declined during period of overload. The results ranged from burned- out motors to the experience of a maker of polystyrene during power cut-offs: ... At such a time the material flowing through the produc- tion equipment solidifies, causing a shutdown of no less than ten days and up to three weeks, while polystyrene is manu- ally chipped out of the vessels.'6 The discrepancy between Brazilian Traction power rates and private generation was enough to alter the whole cost structure of many firms: . . . A steel foundry reports that while it plans to install a diesel motor to provide compressed air, it cannot afford die- 16Joint Brazil-U. S. Economic Development Comm'n, The Development of Brazil (Washington, D. C., 1954), p. 158. Rates 97 sel generated electric power for its electric furnaces. Light company power costs this foundry about Cr. $0.14 per kwh, while it calculates that the total cost of diesel power, includ- ing amortization of equipment, would come to about Cr. $1.30 per kwh, or nearly 10 times as much.17 In spite of cost differences of this order, Brazilian industry had installed an estimated 200,000 kw of small thermal plant in the brief period between the end of World War II and 1953, or over 10% of all electrical generating capacity in Brazil in the latter year.18 Thus cost discrimination from the standpoint of individual manufacturers may become extreme even if public utility sup- pliers of electricity do not engage in rate discrimination. From the standpoint of power users and of the whole economy, this cost discrimination is an important problem-especially if it operates to the disadvantage of the most rapidly-developing and most progressive industrial sectors. B. RESIDENTIAL RATE STRUCTURES The main factor determining the expenditure on electricity of household consumers is probably the income of each house- hold. This holds true for Latin America as it holds, in general, for other countries. This expenditure may be considerably in- creased or diminished by the relative prices of electricity and other energy sources, but for electricity, unlike most of its competitors, this is a matter of rate structure as well as rate level. Two different rate structures may produce the same monthly bill for some particular amount of electricity con- sumed per month, yet one structure may be "promotional" and provide very low rates for additional consumption, with highl average consumption resulting, while the other may be "non- promotional" or even "anti-promotional" and be associated 17 Ibid. 18 Id., pp. 163, 164. 98 Part 1. Problems of the Industry Table 4.11. Expenditure on Electricity as a Percentage of Total Expenditure by Specified Types of Resi- dential Consumers Consuming group % spent for Country and city investigated Date electricity 1. Brazil: S5o Paulo .... Wage-earners June 1934 1.3% 2. Rio de Janeiro a 1946 1.0 3. Chile: Santiago ..... Employees March 1928 1.67 4. Wage-earners Sept.-Oct. 1946 0.30 5. Colombia: Barranquilla Wage-earners Oct. 1946 0.87 6. Medellin ..... Wage-earners June 1938 0.87 7. Bogota ..... Middle class July 1940 1.44 8. 1946 0.94 9. Mexico: Mexico City .. - 1950 1.0 10. Costa Rica: San Jose ...... Wage and salary earners 1949 1.45 11. U.S.: Major cities .. Moderateincome 1934-1936 1.6 12. Aug. 1948 0.9 13. Total personal consumption expenditure 1935 1.24 14. 1948 0.88 15. 1953 1.17 16. OEEC nations: Total value of consumer goods 1953 1.1 *Data are for a family of four in Rio de Janeiro in 1946, with a monthly income of Cr. 1500. The average monthly wage in Rio de Janeiro in 1946 was Cr. 800. In- ternational Labor Organization, Yearbook of Labor Statistics, 1953, p. 147. **In computing this item, the consuming group was given the same weight as that given in the preceding line. Soutrce: Line 1: U. S. Bureau of Labor Statistics, 68 Monthly Labor Review (Jan. 1947) p. 18. Line 2: Conjuntura Econ6mica, Year I1I, No. 3 (March 1949) p. 7. Lines 3, 4: Panorama Ec6nomico, Year IV, No. 29 (Sept.-Oct. 1950) p. 213. Lines 5-8: Colombia, Direcci6n Nacional de Estadistica, Anales de Economia y Estadistica, Supp. to Vol. III, No. 6, pp. 31-34; to Nos. 19 and 20, pp. 33, 63; to Nos. 40-42, pp. 45, 75. Line 9: Banco de Mexico data. Line 10: U. S. Buieau of Labor Statistics, 71 Monthly Labor Review (Oct. 1950) pp. 443-444, (203 wage-earning, and 55 salary earn- ing, families) . Lines 11, 12: U. S. Department of Labor. Lines 13-15: U. S. Department of Commerce, National Income, 1954; Line 16: OEEC, The Trend of the Selling Price of Electricity (Paris, Nov. 1958), p. 18. with low average consumption. Thus, even if income were the sole determinant of expenditure on electricity (which, of Rates 99 course, it is not), the average quantity of electricity consumed per customer would depend on both the level and the struc- ture of prices charged for this consumption. Family budget studies in Latin American cities show that the share of total family expenditures devoted to electricity, even in working-class households, is not far from the percent- age reported in studies for the United States. In both cases, the proportion of electricity bills to total outlays is in a range from a little under to not much over 1%. The available evi- dence is given in Table 4.11. Although this table shows a low average ratio of working- class expenditure on electricity to working class income, this cannot be explained on the ground that most working-class households do not use electricity. Even in 1938, 72.25% of the households in the Medellin sample used electricity, and the percentage for Barranquilla in 1946 was 83.01%.19 Correction of the Medellin and Barranquilla figures to eliminate all families without electricity would raise the electrical share of total expenditures only to 1.2% and 1.05%, respectively. The widespread urban use of electricity is, in fact, one of the outstanding features of the Latin American industry, and possibly the most important single feature from the standpoint of influences at work on residential rate policy. The relatively widespread consumption in Latin American towns and cities emerges on any basis of comparison: of the ratio of households using electricity to total households; of the ratio of households using electricity to households with access to such services as public water supply and sanitary sewers; even of the ratio of towns in which electricity is available to the total number of towns, or to the number of towns with other common urban services. Brazilian data for December 31, 1950 illustrate the very low level of income required to make electricity an unattainable luxury. In the relatively prosperous southern states, 98.2% of 19 Colombia, Anales de Economia y Estadistica, Supp. to Nos. 40-42, p. 75, and Supp. to Vol. III, No. 6, p. 62. 100 Part 1. Problems of the Industry all residences in state capitals were supplied with electricity. In the less prosperous north, the ratio was still 49.2%. Even in non-capital cities, 89.2% in the south, 27.9% in the north, and an average of 60.7% for all of Brazil used electricity.20 The Brazilian record is confirmed by the high ratio of elec- tricity consumers to total households consistently reported by cost-of-living studies in the other countries studied. The general availability of electricity in urban communities is in contrast to the much less widespread distribution of other basic services. Brazil again provides information on the dis- tribution of electricity, public water supply, and sanitary sewer services among working-class households sampled in 1952. In four out of 26 north Brazilian cities, no households investi- gated used electricity, but piped water was entirely absent in nine cities and sanitary sewers in 18. Of 40 cities of south Brazil, none reported an entire absence of electricity, although one reported no piped water to any working-class household surveyed and eight reported no sanitary sewers. Electricity was used by 70% or more of the households in 30 of these 40 cities, a ratio achieved in only eleven for piped water and in only six for sanitary sewers. Only six of the 66 cities in the survey reported any working-class use of piped gas, and in each of these six the proportion of users was under 10%. Users of electricity made up 67% of the sample in Recife, and over 85% in the other five cities.21 The same relationships appear in studies of availability of electricity to at least some consumers. Even in the relatively impoverished north of Brazil, all state capitals and 585 of the 693 smaller communities covered by the 1950 survey had elec- tricity service of a sort, while in south Brazil not only all the state capitals but, in addition, 1,072 of 1,187 other communi- ties had electricity.22 In Colombia, in 1949-50, every munici- pality of the 56 with populations of 10,000 or more had elec- 20 Conijtntura Economica, Year VIII, No. 4 (April 1954) p. 39. 21 Anuario Estatistico do Brasil, 1954, pp. 338-339. 22 Conijunzttra Economnica, Year VIII, No. 4, p. 39. Rates 101 tricity supply, while only 46 had public water systems and 37 had public sewer systems. Of 279 municipalities with popula- tions between 2,000 and 10,000, 189 had electricity, 134 had water systems, and 119 had sewer systems.23 These data pose two questions which bear on the problem of residential rate structure for electricity: (1) Given the cheapness of electricity, which must be a factor in its very widespread use, can any rate structure be devised which yields adequate revenues to finance expansion and is also politically feasible? (2) Given the extreme poverty of most Latin Ameri- can municipal governments, as displayed by their inability to provide water supplies and other public improvements, does the use of public funds for further electrical development yield higher social benefits than could be obtained from alter- native expenditures, allowing the burden of expanding elec- tric supply to fall on electricity consumers, residential and in- dustrial? This second problem has been disguised when pri- vately-owned companies provide electricity. With rates kept too low to permit expansion by resort to private investment, the private company has typically turned for aid, not to the local, but to the national government or its agencies. But the problem remains in its essentials, because the governmental funds provided in response to these appeals could otherwise have been used for different municipal purposes. Three particularly interesting cases-of Mexico, of the Chil- ean regulatory authorities with reference to the Compafila Chilena, and of the municipal electrical enterprise in Medellin, Colombia-will be examined in some detail to show the rela- tionship of residential rate structures to the level of residential consumption and to the goals of social policy which have be- come intertwined with rate regulation. The broad question of "social rates" for residential users will be briefly considered in a concluding section. 23 U. S. Foreign Service Despatch No. 47, Bogota, Colombia, July 18, 1951, quoting report of Committee of Economic Development on municipal develop- ment and housing. 102 Part 1. Problems of the Industry Mexico Mexico has approached the problem of residential electricity through two different channels: indirectly, through the pricing policy of Petr6leos Mexicanos for petroleum products which can be used for light and heat, and directly, through the "Mexican cycle" residential electrical tariff devised by the Tariff Commission. Prices of petroleum products sold in Mexico were controlled by the government even before the 1938 expropriation of foreign oil interests, from which Pemex emerged. Since then, Pemex has followed a deliberate policy of encouraging the use of kerosene and liquid gas in place of wood and charcoal-and in place of electricity. Between 1939 and 1946, the workers' cost of living index in Mexico City increased by 166.7%.24 Yet the average price of kerosene in Mexico was 13% below its 1938 level in 1946, and the average price of liquid gas was 5% lower. In addition, sales of liquid gas were expanded from the Mexico City area to most major cities, and consumption was encouraged by the sale of stoves at cost by Pemex. Be- tween 1938 and 1946, domestic consumption of kerosene in- creased ninefold, and of liquid gas twelvefold.25 It is no coin- cidence that sales of electricity per residential customer are lower in Mexico City than in any other capital city in the countries studied, and that average consumptions in Mexican cities are also lower than any others reported except in north- eastern Brazil and a few other low income areas. Meanwhile, the Tariff Commission has been steadily re- shaping the structure of Mexican electricity tariffs, especially for residential users. A small fixed charge per lead-in wire, which gives some recognition to probable size of appliances used, is followed by "cycles" of alternating higher and lower 24 Anuario Estadistico de los Estados Unidos Mexicanos, 1951-1952 (Mexico, D.F., 1954) pp. 938-939. 25J. R. Powell, The Mexican Petroleum Industry, 1938-1950 (Berkeley and Los Angeles, 1956), pp. 107, 108, 189. Rates 103 rates. For example, in the Puebla-Veracruz area, in 1954, the first ten kwh of residential consumption per month cost 21 centavos each, the next ten cost 17 centavos, the next ten again cost 21 centavos, and so on for a total of 30 kwh at the higher charge and another 30 kwh at the lower charge. All additional energy used cost 11 centavos per kwh. The width of each block varies from undertaking to undertaking, with twenty a more common figure than the ten used in this example. So does the level of rates in each block, the spread between high and low rate levels, and the level of the rate of rates charged for "fol- low-on" consumption. In all cases, the combination of fixed charge and cycle rate produces a declining average charge per kwh over the entire range of the cycle. The significance of the system's contribution to rate-making is diminished by the fact that the difference between high and low cycle rates is generally only one-fifth or one-sixth, and the residential cycle stops at fairly low consumption levels. But the system has certain real merits. The fixed charge is higher for consumers with large appliances than for other consumers. Even very small consumers can obtain the financial, and psycho- logical, advantage of lower prices for some of their electricity. If the cycle were extended to higher levels of consumption, the system could be used to provide low average rates for small users without also providing still lower average rates, and possibly still greater total subsidies, to much larger and usu- ally much wealthier residential consumers.26 Chile The residential tariff structure of the Compafila Chilena, serving Santiago and Valparaiso, has been the gradual out- growth of regulatory desires to hold down rates to small con- sumers in the face of severe inflation. Table 4.12 summarizes the total change in Chilean residential rates from the time rate control was introduced, in 1931, to the tariff change au- 26 For a more extensive description, see G. C. Delvaille, "The Mexican Cycle," 61 Public Utilities Fortnightly (1958) p. 811. 104 Part 1. Problems of the Industry thorized by the regulatory authorities effective July 17, 1954. This change was effected by a series of individual steps. Table 4.12. Residential Rate Increases, Cia. Chilena de Elec- tricidad, 1931-1954 Residential use July 1, 1931 July 17, 1954 per month Pesos/kwh Pesos/kwh 1931 = 100 First 12 kwh ..................... 0.95 1.68 177 First 30 kwh, if over 12 kwh ....... 0.95 2.8875 304 Next 20 kwh ..................... 0.95 4.555 480 All over kwh ..................... 0.95 6.74 709 Heating rate ..................... 0.225 - - Lowest residential energy charge ... 0.16 3.5775 2236 Cost of living, Santiago 1929 = 100 - - 3150 Sources: Electrical rates: Cia. Chilena rate files; Santiago cost of living: U. N. Statistical Yearbook. This table reveals a shift from a straight-line meter lighting rate in 1931, buttressed by promotional heating and general residential rates, to a very anti-promotional rate structure by mid-1954. This evidence is corroborated by the history of the company's optional residential tariffs (i.e. tariffs not required by, nor established by, the regulatory authorities): 14 of these were available in 1938; six in 1946; two in 1954. If the rate level for residential electricity were remunera- tive, this rate structure would be very objectionable. Large users of residential electricity on the lighting tariff were re- quired to pay over four times as much per kilowatt-hour for consumptions of over 50 kwh per month as the rate charged to the very smallest users. Moreover, the thirteenth kilowatt- hour per month involved an extra charge on all consumption, for a total incremental cost of 17.3775 pesos for this one unit. Even the lowest rate available to any of the larger residential consumers was over twice as high as the base rate for the very smallest. But even before this highly anti-promotional rate structure is criticized, it must be noted that all Chilean rates are low. A combination of low rate levels and a promotional Rates 105 rate structure would simply have pushed supply-and-demand relationships from bad to worse. Before rates for larger users were increased relative to other rates, the growth of the winter space heating load had forced the introduction of rationing and in addition had pushed the Compafia Chilena into operating losses during its months of heaviest sales. The company's rate structure in 1954 was no model, by any standard, but it at least faced the problem of how to temper the inflationary wind to the small consumer without driving the company into insolv- ency by giving all consumers the same treatment. MedellIn, Colombia Medellin provides an outstanding example of an electric enterprise whose rate level provided adequate earnings at a given level of consumption relative to a plant valued in pre- war pesos, but whose rate structure was totally inadequate to cope with the problem of financing expansion. Moreover, Medellin illustrates the tremendous effect of really low incre- mental rates on residential demand for electricity in any area without cheap alternative sources of heat and light. Finally, the Medellin residential rate structure includes a fixed charge based on the assessed value of the consumer's residence, and thereby provides the best available proof of the extent to which "promotional" rates in a Latin American city are in fact of greatest benefit to the wealthiest users. Before Medellin came face to face with the problem of fi- nancing a 100%, expansion in its generating plant, it was a paradise for residential consumers of electricity. Average re- ceipts per kilowatt-hour sold to residential users were lower than for any other category of electric sales. The residential rate structure consisted of a fixed charge of 30 centavos per month per 1,000 pesos assessed value of property plus 0.7 cen- tavo (equivalent.to less than 4 mills, U. S., at the official ex- change rate) per kilowatt-hour consumed. Although Medellin is the industrial center of Colombia, residential consumers used over half the electricity sold by the municipal undertak- 106 Part 1. Problems of the Industry ing. Their use per customer was some 50% above the United States average. Once Medellin had to face the costs of its new Riogrande hydro plant, the arithmetic of these rates produced a financial impossibility. The old generating plant had a gross value of Ps. 250 per kilowatt; the new plant cost Ps. 800 per kilowatt when finally completed in 1950. Assuming power losses equal to one-sixth of residential sales and a 60% load factor (the average for the entire undertaking) for residential uses. Mede- ilin's pre-1949 rates would have yielded gross revenues, before any operating expenses or depreciation, of only 3.83% on the new investment in generating capacity. Given the high rate of growth in Medellin residential de- mand, the municipal electrical enterprise could eventually have swallowed up all of the city's taxing power on the basis of this rate structure. Therefore, in 1949, Medellin tried its first experiment with anti-promotional rates: the fixed charge and the energy charge for the first 200 kwh per month stayed the same, but energy charges for incremental consumption rose to a peak of 6 centavos per kwh for consumptions of over 1500 kwh per month. A further upward revision of energy rates in 1953 changed the system to a flat one-peso charge for the first 50 kwh per month, with upward steps leading from 2 centavos per kwh to a 12-centavo plateau at over 1200 kwh per month. Here, in general, is a repetition of the Chilean experience. But the differences are important. For many consumers, the zero incremental charge for any of the first 50 kilowatt-hours would hardly operate as a check to consumption. Moreover, reverse blocking (higher incremental rates) began only at the 501 kilowatt-hour level. In May, 1953, 84.3% of all Medellin residential consumers purchased less than this amount.27 Fi- nally, the highest residential rate involved an abrupt swing from the promotional to the punitive. 27 Olarte, Ospina, Arias & Payan, Empresa de Energia Electrica, Medellin, Analisis de las Tarifas (Bogoti, 1953) p. 37. Rates 107 The distribution of Medellin residential consumption groups, by average assessed value of residence for each level of consumption, is available for October 1951. It is fair to assume that these statistics reflect earlier promotional rates with rela- tively little influence from the anti-promotional fillip added in 1951. They are presented in Table 4.13. Table 4.13. Medellin, Colombia: Residential Consumption and Assessed Value of Residences, October 1951 Average assessed value per residence Number of (in thousands Monthly use (kwh) installations of pesos) 0-19 ..................................... 5,163 5 20-49 ........................... I ......... 7,290 4 50-99 ..................................... 7,031 5 100-149 .................................... 5,134 5.75 150-199 .................................... 4,430 6 200-399 .................................... 11,945 8 400-599 .................................... 4,611 14.5 600-799 .................................... 1,844 23.75 800-999 .................................... 949 33.25 1000-1499 ................................... 825 48 1500-1999 ................................... 148 71 2000-4999 ................................... 60 75 5000 and up .....................1........... - 49,431 total 8.85 average Source: Empresa de Energia E1ectrica, Medellfn, Informe, 1951, p. 43. The "assessed value" column in this table clearly breaks at two points: consumptions of 400 kwh per month, and con- sumptions of 1000 kwh per month. The lower income con- sumers below 400 kwh per month provided 82.9% of the total installations, 12.5% of assessed value, and 12.3%o of total use. The correlation between use and assessed value in the middle and higher consumption ranges is obvious. The effect of the original Medellin price scale, then, would have been to concentrate the electrical subsidy on the middle and upper income groups if it had been continued. This is, 108 Part 1. Problems of the Industry in fact, the probable effect of promotional rates in most other parts of Latin America today. C. ELECTRICITY RATE STRUCTURES AND SOCIAL POLICY In spite of the small percentage contribution of electricity to the total cost of living in the countries studied, regulatory desire to hold down this particular cost has probably exerted a strong influence on rates charged for electricity sold to all classes of buyers. This result would be inescapable if electric- ity were sold at the same price to everyone and in all quanti- ties, but the very existence of rate structures makes a different policy economically feasible. Two extreme cases of "anti-promotional" and "promotional" rates have just been reviewed, with the Compafiia Chilena at th-e former end of the spectrum and Medellin, Colombia, before its rate revisions, at the latter end. The relationship of small consumers to total residential consumers, consumption, and revenue in each of these supply areas is summarized below: Table 4.14. Small Residential Consumers, Medellin, Colom- bia, and Cia. Chilena de Electricidad l'ercentages of total residential Consumers Consumption Revenue McdellIfn, Colombia, October 1951 0-19 kwh ..................... 10.4% 0.03% 3.0% (0.2%) 20-49 kwh ..................... 14.8 2.2 3.9 (1.2 ) e Total, 0-49 kwh ............. 25.2% 2.2% 6.9% (1.4%) * Cfa. Chilena, 1953 0-12 kwh ..................... 5.5% 0.4% 0.4% 13-20 kwh ..................... 8.4 1.2 1.7 21-50 kwh ..................... 29.3 10.4 15.6 Total, 0-50 kwh ............. 43.2% 12.0% 17.7% *Pcrcentage of total revenue from energy charges. Source: Company data. Rates 109 The proportion of really small consumers (20 kwh per month and less) to all residential users is surprisingly similar in these two areas in spite of their very different rate structures: 13.9% for the Compafiia Chilena, 10.4% for Medellin. Their con- tribution to residential consumption was negligible in both: 0.03% in Medellin, 1.6% in Compafifa Chilena. Even the 0-50 kwh group in the Compafina Chilena area accounted for only 12% of total residential sales (12% x 33% residential to total) . In both cases, really small consumers could be shielded from inflation to any desired degree without significantly affecting the revenue yield from a rate increase. This is not to say that a policy of stable rates for very small users is economically justifiable. The expense of serving them is unusually high, per kwh, owing to the fixed costs of meters and meter reading, service facilities, etc. They are certainly served by the Compafifa Chilena at a considerable loss, and they are probably not profitable in any supply area. With urbanization and the extension of supply areas, new small con- sumers may continue to appear for some time to take the place of others who graduate to larger consumptions. But at least these consumers place almost no demands on present genera- ting and distribution capacity and are even less significant for the problem of expansion. With the continued rise in real in- comes in Latin America and the continued spread in the use of electricity, they will probably decline in relative numbers and certainly decline in relative consumption. There appears little excuse, economic or social, for non-remunerative rates to all residential consumers in an attempt to aid this sub- category, especially since these very small consumers can be aided directly to any desired degree by appropriate adjustment of the rate structure. CHAPTER V Rate Regulation in the Electric Industry The principle that prices charged by a monopolistic enterprise which performs an essential public service ought not to be left to the discretion of its management has been recognized in the five Latin American countries studied. In all of them the electric industry has come to be subjected to regulation by the central government. The regulatory agencies have been given jurisdiction over the rates charged not only by investor-owned companies but by autonomous government authorities for the sale of electric power for consumption or resale. (In most of the United States, government-owned electric utilities are not subject to rate regulation.') As has been indicated in Chapter II, the five Latin American countries studied have all adopted the general pattern of rate- 1 In the United States, state regulatory commissions have jurisdiction over rates charged by government-owned electric utilities in only 14 states, and, in four of these, rate regulation is confined to sales of electricity outside the municipalities owning the utilities. Many of the states in which large govern- ment-owned electric utilities are located do not impose rate controls. Among the regulating states, however, are New York and Wisconsin, the latter having been a leader in the development of utility regulation. State commissions regulate accounting and depreciation practices of government-owned utilities in a number of states in which no commission control over rates is authorized. Federal Power Commission, State Commission Jurisdiction and Regulation of Electric and Gas Utilities, 1954 (Washington, D. C.), Table 1, p. 18. 110 Rate Regulation in the Electric Industry 111 regulation which had been evolved in the United States, mainly in the first quarter of this century. The chief characteristic of that regulatory system has been reliance on commission- administered formulas for determining whether an increase or decrease in the general level of rates charged by a utility was warranted and, if so, the extent of the change to be approved or required. The standards used in these formulas have been of such generality as to allow a considerable degree of discre- tion to the regulatory commissions. However, in the United States, as a means of protecting the commissions in the exercise of this discretion from industrial and political pressures, the commissions have generally been made independent of direct control by the executive branch of the government. To pro- tect the industry from arbitrary action by the commissions their orders have been subjected to extensive review by the courts. In comparison with the United States, greater reliance has been placed in Latin America on specific statutory standards than on the discretion of the regulatory bodies. Moreover, the latter are ordinarily placed within one of the ministries of the government, and their orders are therefore subject to the ap- proval of a high political official. Judicial review of commis- sion action is relatively limited and infrequent. Yet, despite these institutional differences, rate regulation in the two con- tinents is sufficiently similar to repay comparison. Since, more- over, many of the suggestions advanced in Part II of this report look to possible improvements in the methods of determining the level of electric rates, some attention here to the th-eory and processes of rate regulation seems called for. Because of its prototype role in the development of Latin American regu- lation, our presentation will give special emphasis to practice in the United States.2 2 For a presentation of the economic theory, legal doctrine, and administrative practice of public utility rate regulation, as it has developed in the United States, the following four works by economists may be consulted: I. R. Barnes, The Eco- nomics of Public Utility Regulation (New York, 1942); E. W. Clemens, Economics and Public Utilities (New York, 1950) ; M. G. Glaeser, Public Utilities in Amer- 112 Part 1. Problems of the Industry The Rate-Making Process Whatever their variations, the rate-making formulas used in the five Latin American countries as well as in the United States can best be understood in terms of the process which must be followed in employing them. Typically a rate proceed- ing to pass upon proposed rates requires that the petitioning electric enterprise project its operating costs for the forthcom- ing year. For this purpose, it is likely to build upon its experi- ence in a previous "test period," adjusted to reflect intervening changes in cost and operations. The enterprise then proposes rates at a level high enough to produce total revenues that wvill not only cover these adjusted operating costs and de- preciation expense but will also provide the enterprise with earnings-ordinarily after income taxes-at a rate of return approved by the regulatory body. This total, the amount that must be produced by the rates if the enterprise is to be self- sustaining, we shall call the "revenue requirement." The generation, transmission, and distribution of electric- ican Capitalism (New York, 1957); E. Troxel, Economics of Public Utilities (New York, 1947). The procedure of the United States utility commissions in handling applications for rate increases is described in two works for the lawyer by Francis X. Welch, editor of "Public Utilities Fortnightly," viz. Preparing for the Utility Rate Case (Washington, D. C., 1954), and Conduct of the Utility Rate Case (Washington, D. C., 1955). The prindples governing judicial review of commis- sion action are presented in Davis, Administrative Law (St. Paul, Minn., 2d ed., 1958). For current developments in the regulation of the electric industry from the industry viewpoint, see Public Utilities Fortnightly (Washington, D. C.) . For economic analyses of regulatory problems, one of the best sources is Land Econom- ics (Madison, Wis.) , a quarterly formerly entitled Journal of Land and Public Utility Economics. For the viewpoint of the regulatory officials, in addition to their decisions which are collected and published currently in Public Utilities Reports (Washington, D. C.), see National Association of Railroad and Utilities Commissioners, Proceedings (published annually). Rate Regulation in the Electric Industry 113 ity require a high ratio of capital investment to annual reve- nues-as noted above, it may be 4:1, whereas a 1:1 ratio is common among other industries. Hence, the problem of the rate of return to be allowed on that investment looms large in fixing electric rates. For the same reason, depreciation expense is important. To be sure, depreciation rates for long- lived hydroelectric installations can be lower than for thermal generating plants. However, a relatively higher investment of capital in depreciable property is involved in constructing hydro facilities. Naturally, too, operating costs vary with the type of installation; thermal generators consume fuel and thermal plants need larger staffs than do hydro, yet, since thermal plants are usually located closer to their consumers, they entail fewer losses in transmission. Distribution, of course, involves relatively higher operating costs than generation. The formulas advanced to determine the proper level of earnings to be yielded by proposed electric rates-the cost of capital-have usually followed a two-step procedure. The first step is to ascertain the "present value" of the property "used and useful" in providing the utility's service, or, instead, the amount of money prudently invested in the utility's plant. The total, however ascertained, is known as the "rate base." The second step is to determine a "fair rate of return" to be earned on this rate base. The rate base is then multiplied by the rate of return, a percentage figure. This produces a money figure which is included in the revenue requirement. (It should be noted, incidentally, that the return thus allowed is before, not after, interest on the utility's long-term debt. The signifi- cance of this fact will be discussed below.) The management will ordinarily fix the particular rates which in the aggregate should yield the revenue requirement. Rate schedules are worked out which specify the rates to be charged different classes of customers and for various types of service. These schedules must, of course, be reviewed by the regulatory body which may insist upon some changes. Such changes may be required in the interests of fairness among 114 Part 1. Problems of the Industry consumers and of other regulatory objectives or simply to make it more likely that the revenue requirement will be met. The Controversy over the Allowable Return: The Rate Base The first of the two steps noted above-the determination of the rate base-has been the center of a half century and more of controversy in the United States. This controversy has been waged by commissions, courts, and commentators. Agreement has not yet been achieved. Moreover, it is reflected in the dif- ferences among the five Latin American laws under consider- ation. Indeed, economic conditions in Latin America have made the issue much more important there than it now is in the United States. The controversy centers on the question whether the rate base should reflect the "present value" of the utility's property or the amount of money invested in the property. These al- ternatives are generally stated in terms of cost less depreciation. Under the first view, the cost of reproducing the utility plant is taken as a measure of its value. Under the second, the total amount actually expended over the years in constructing and expanding the plant-its "original cost"-is the measure of the investment. Until 1944, the United States Supreme Court had specified the "present fair value" of the property as the standard re- quired by the Constitution.3 The Court had required that state commissions give weight to reproduction cost in arriving at that value. However, in an analytically illogical compromise, the Court allowed the Commissions to give weight to original cost as well. The resulting uncertainty led to countless at- 3 The leading case was Smyth v. Ames, 169 U. S. 466 (1898). The doctrine reached full development in McCardle v. Indianapolis Water Co., 272 U. S. 400 (1926). Rate Regulation in the Electric Industry 115 tempts to set aside commission rate decisions in the courts. The endless litigation that resulted led to about as much criti- cism of the Court's position as did the inclusion of reproduc- tion cost in the formula. Much of the complaint against the use of reproduction cost as a factor was based on the great ad- ministrative difficulty in ascertaining it. This difficulty has been mitigated in recent years by the development of index numbers to measure changes in construction costs. Ascertain- ment of the original cost factor was also a source of difficulty until the gradual improvement of accounting practices and the rectification of past accounts rendered it a reliable instrument. The basic issue has not related to administrative considera- tions, however, but to the stability of the price level. Use of reproduction cost as a measure of value provides a means of adjusting the investor's monetary return to the rising prices of goods and services and to the concomitant falling purchas- ing power of money. Resort to the original cost measure, oni the other hand, permits only past price levels to be reflected in the return. Naturally, the more recently the plant has been constructed, the more closely an original cost rate base will reflect the current purchasing power of money. The protagonists of reproduction cost have argued that the tendency of the price level to rise will persist. The protago- nists of original cost have emphasized the fact that prices have fallen as well as risen in the past. They have stressed the im- portance of avoiding the injection of price instability into the return to holders of utility common stock. It is significant, perhaps, that the champions of original cost won from the United States Supreme Court the right to employ their method in the milestone case, Federal Power Commission v. Hop.e Natural Gas Co.,4 only after the deflation of the '30's had greatly reduced the actual difference between the two views. 4 320 U. S. 591 (1944). The Supreme Court held that, in determining rates, the commission was "not bound to the use of any single formula." The Court re- quired only that the "total effect of the rate order" not be unjust or unreasonable, "from the viewpoint of both investor and consumer interests." Id. 116 Part 1. Problems of the Industry Now that prices have been rising for over a decade, many state commissions and courts are refusing to make the shift to the original cost rate base that the Supreme Court now permits.5 The regulatory laws of Brazil (1934, 1941, 1943),6 Chile (1931, 1935) ,7 Colombia (1938) ,8 Costa Rica (1941) ,9 and Mexico (1939) 10, all contemplate that the rate-base x rate-of- return mode of determining the earnings factor would be used in calculating revenue requirements. Chile, having the earliest laws of the five countries, adopted reproduction cost as the measure of the rate base, whereas Brazil and Mexico (Nvith qualifications) adopted an original cost rate base, reflecting the drift in thinking on this issue in the deflationary '30's. In 5 An earlier trend chiefly to the contrary is reported in Rose, "The Hope Case and Public Utility Valuation in the United States," 54 Columbia Law Reviezu (1954) p. 188. More recent decisions adhering to "fair value" include: Re Dia- mond State Tel. Co., 48 Delaware 497, 107 A. 2d 786 (1954); Central Mlfaine Power Co. v. Maine Pub. Util. Comm'n, 150 Maine 257, 109 A. 2d 512 (1954); State ex rel. Utilities Comm'n v. Southern Bell Tel. & Tel. Co., 239 North Carolina 333, 80 S.E. 2d 133 (1954) Simms v. Round Valley Lighlt & Power Co., 80 Arizona 145, 294 P. 2d 378 (1956); N. Y. Tel. Co. v. Pub. Service Comm'n, 309 New York 569, 32 N.E. 2d 847 (1956). 6 Brazil, Code of Waters of July 10, 1934, was amended and amplified by Decree Laws 3128 of March 5, 1941 and 5764 of August 19, 1943, and wvas implemented in part by Decree No. 41,019 of February 26, 1957. Subsequent acts have authorized increases in rates, e.g., Decree Law 7524 of May 5, 1945 (implemented in Decree Law 7716 of July 6, 1945), Law No. 27 of February 15, 1947. 7 Chile, Decree (with Force of Law) No. 244 of May 15, 1931, implementedl by Regulations for the Operation of Electric Lighting and Power Services, Decree 3386 of Aug. 15, 1935. The Cfa. Chilena, by far the largest privately-owned citer- prise in Chile and the largest retail distributor, is subject to a special conversion contract enacted into law two months before the general law in 1931 (Decree No. 29 of March 11, 1931) and to the Ross-Calder Agreement of Nov. 26, 1935, ap- proved by Law No. 5825 of March 10, 1936. 8 Colombia, Law No. 109 of 1936, supplemented by Rcgulatory Dectees 1,606 and 1,717 of 1937, and Decree Law No. 5, authorizing Government contracts with electric utilities on the basis of the standards presented by the above statute and decrees; cf. Decree Law No. 5 of 1940, enacting the contract with Cla. Colomnbiana. 9 Costa Rica, Law of the National Service of Electricity, Law No. 258 of Aug. 18, 1941, as amended by Law No. 50 of June 7, 1948, and as implementedl by Decree No. 14 of Sept. 17, 1942. 10 Mexico, Diario Oficial ("D.O."), Feb. 11, 1939, as amended, D.O., Jan. 14, 1942, implemented by Regulations under the Law of the Electric Industry, D.O., Oct. 4, 1945, as amended D.O., July 13, 1948, Nov. 11, 1949, Nov. 29, 1952. Rate Regulation in the Electric Industry 117 Colombia the statutory standard of "actual investment" also appears to call for an original cost base, and Costa Rica pro- vides for a return on "the capital invested." In both the latter two countries, the regulatory law provides the framework for concession contracts to investor-owned utilities. In November 1958, Brazil passed a lawloa allowing a re- valuation of corporate assets, including those of electric utili- ties, in accordance with official coefficients. A tax of 107%, payable over the period of a year, was imposed on any result- ing increase in the value of corporate equities. A provision of the bill, as passed, forbade the revalued assets of utilities to be used in computing electric rates. This provision was vetoed by the President, and his veto was sustained. However, the underlying law requiring an original cost rate base has been changed only to the extent advantage is taken of the particu- lar occasion afforded by the 1958 law, and this is awaiting the promulgation of a regulatory decree prescribing the procedure to be followed. The rate-base x rate-of-return formula assumed both that price movements would be gradual and- that the burdensome task of ascertaining a rate base would be required only at long intervals. If, in the interim, additional investment were made in plant, the same rate level that yielded enough on the pre- existing plant would be satisfactory for the new plant. The risk that rising construction costs might result in a higher cost per kilowatt-hour for electricity generated at a new plant was not a source of much concern, especially in view of continuing technological progress in the industry. The Rate of Return The risk that a rising price level would make an original cost rate base inadequate was dismissed by many supporters of lOn Law No. 3470 of Nov. 28, 1958, Art. 57. 118 Part 1. Problems of the Industry the original cost theory in the United States. They argued that any deficiency could be compensated by the other factor in the equation: the rate of return. A 20% rise in rate of return will, of course, increase the revenue requirement just as much as a 20% rise in the rate base. In the United States, no statutory limitations were placed on the allowable rate of return. This was left to the informed judgment of the commissions, subject to judicial review. In the Latin American countries studied, less freedom has been given to the commissioners. Thus, in Brazil, the relevant decree prescribes 10% as the maximum allowable return (on historical cost less depreciation) with a proviso permitting increases in the event of major monetary changes up to the yield on internal government bonds plus 3%. The National Economic Council of Brazil, in reporting in 1952 on the electric supply situation and criticizing the level of rates as too low, pointed out the connection between the rate of return and the rate base to which it was to be applied:"11 It is not the limitation of profits which dissuades private capital from being invested in these services, not the 10% dividend which is manifestly low in comparison with the re- turns from capital invested elsewhere. The fundamental point lies in the unchangeable basis from which this per- centage is reckoned and which reduces the value of the re- turns all the time, as the currency depreciates. Chile has prescribed a maximum return of 16% but, as has been noted above, no revaluation of the rate base can be ob- tained until there has been a failure to earn at least 10% on the existing base for three successive years. In a period of rapidly rising prices, this delay can work grave damage to the earnings position of utilities subject to it, especially when com- pared to that of unregulated industries with which utilities should be competing for capital. In Mexico, no statutory ceiling has been placed on the per- mitted return; instead, a floor has been prescribed: it must be 11 Revista do Conselho Nacional da Economia (Aug.-Sept., 1952) p. 7. Rate Regulation in the Electric Industry 119 no lower than the highest rate for government bonds. How- ever, the rate of return permitted by the regulatory body, has until recently been close to the minimum. Thus for a time, 8% was allowed before taxes; lately this has been raised to 8% after taxes, the usual mode of measuring return. By 1954, Mexican Light & Power Co. had succeeded in getting its target return raised to 11.5% after taxes. In Colombia, the picture is unclear. Many of the electrical companies are taken over by municipalities. One standard recognizes a return of 8% as interest and an additional 8% as profit, subject in the latter case to some adjustments downward. In the case of the largest private company, Compafiia Colom- biana, an American & Foreign Power subsidiary, the contract governing its rates sets a maximum return of 3%/0 above the level at which United States loans would be made to utilities in Colombia. In Costa Rica, a 10% maximum prescribed by the franchise contract is limited to the return on the company's equity. This equity is restricted, surplus excluded, to 35% of the company's capital. These two examples call for a few words of comment. A rate base includes property acquired by the issuance of both debt securities and common stock. Therefore, if the return earned on it is to cover the cost of capital as an element in the utility's cost of service, the return must be available to meet the require- ments of both bondholders and shareholders. If the interest rates payable on bonds are low relative to the rate of return earned on common shares, a single rate of return on the rate base may conceal a wide spread in the respective returns on the two classes of securities. Thus, in the United States, where interest rates have been low for many years, a company may have a capitalization of 50% in 4% bonds and 50% in com- mon stock. If this company is allowed a rate of return of 61/2% on its rate base, it will have 9% per share available for its common stock. In recent years, the growing awareness of the importance of taking these quite different levels of return into account in calculating the cost of capital has been due in large 120 Part 1. Problems of the Industry measure to the recognition of the "leverage" factor discussed in the section following. The Significance of "Leverage" It once was customary for commissions in the United States to fix rates of return without regard to the capital structures of the utilities before them. If two utility corporations were allowed exactly the same return on their respective rate bases, then the actual returns enjoyed by their common shareholders would differ if the corporations' capital structures differed. The return to the common would be higher in a corporation with substantial debt outstanding than in a corporation with little or no debt. This advantage was generally known as "le- verage." It may still be enjoyed but not because capital struc- tures are ignored in fixing rates of return. In practice, the rate of return on the common shares is usually determined sepa- rately from the returns already established by contract on the fixed-income securities, such as bonds and preferred stock. The common is allowed whatever return is found necessary to at- tract new investment in common stock without unfairness to existing shareholders. (They must be protected against "dilu- tion" of their interest that would result if new shareholders were allowed to pay proportionately less for the shares they acquired.) Today, in the United States, the present cost of equity capital is ascertained on the basis of extensive studies of stock market behavior of utilities stocks. Special attention is paid to the price-earnings ratios of the common stocks of the utility whose rates are being determined and of other com- parable utilities. Extensive testimony is provided by financial experts. Their opinions frequently conflict. Once the rates of return have been ascertained for each of the utility's outstanding securities, then a weighted average of Rate Regulation in the Electric Industry 121 these returns is generally stated as the return allowed on the company's rate base. This practice has created a misleading ap- pearance of similarity between the levels of return permitted to utility investors in the United States and in Latin America. The cause of this misleading impression lies in the fact that Latin American utilities, if they are able to sell fixed-interest rate debt securities at all, must do so at rates of interest that are nearly as high or even higher than the rate of return al- lowed by law on the utility's rate base. Suppose such a company is allowed a return of 8%/. after taxes. This return would be 1.5% higher than the 6.5%/ return received by the hypothetical United States utility mentioned above which had 50% of its capital in 4%0 bonds and earned 9% on its common shares. Suppose that the Latin American company has been successful in selling 8% bonds to provide half its capital, a prodigy of financing under current conditions. In that event, after bond interest had been paid, there wvould remain from the 8% re- turn on the rate base only 8% per share for the company's common stockholders. This 8% return on common is 1% per share lower than the return enjoyed by the stockholders in the United States com- pany, yet, ex hypothesi, the latter's credit is much better, and its shares represent a correspondingly more attractive invest- ment. Not only are the common shareholders in Latin American corporations denied the benefit of leverage by reason of the combination of high interest rates with fixed or low rates of return on the rate base, but they may even suffer from nega- tive leverage. This would occur whenever the requirements of the company's bonds and preferred stock average higher than the rate of return it is allowed by law on its rate base. The re- sult, of course, is to destroy the market for new issues of com- mon stock, which could offer only a lower return than that received on securities senior to the common. 122 Part 1. Problems of the Industry The "Dividend Pay-out Ratio" and Retained Earnings In the preceding discussion, reference has been made to the rate of return allowed an electric utility and the earnings avail- able to its common shareholders. These earnings are not to be identified with the dividends actually paid by the utility on its common shares. A well-managed electric utility enterprise will normally withhold some of its earnings for reinvestment in utility plant. This retention of earnings would, of course, have to be at the expense of the dividend income expected by the common shareholders on the common stock in case no substantial margin of earnings had been allowed over and above that level. To cut into the dividend income expected by the common shareholders would further reduce the attractive- ness of the common shares. If, on the other hand, the enter- prise reinvested little or none of its earnings but paid out all or nearly all its net income in dividends, it would deprive it- self of an otherwise ready source of additional capital and might find itself compelled later on to seek outside capital un- der very adverse, costly market conditions. Moreover, it would leave the shareholders' income expectations unprotected by any earned surplus. This dilemma has chronically confronted the shareholders in investor-owned electric companies in the Latin American coun- tries studied. The situation contrasts sharply with the regula- tory practice in the United States. In the United States, regu- latory commissions, in calculating the return allowable to common shares, pass on what is an appropriate "dividend pay- out ratio," namely, the ratio of dividends actually paid to the total earnings available for the common stock. Once the level of dividends necessary to attract capital has been determined, a dividend pay-out ratio of about 70%, is generally approved. Rate Regulation in the Electric Industry 123 This means that of every dollar earned, thirty cents will be available for plant expansion.12 Moreover, this margin is prop- erly regarded as shareholders' funds which they are investing in the company's property in this fashion. Consequently the plant thus acquired is added to the rate base. An element of growth is thus built into the shareholders' investment. It adds to the shares' attractiveness for long-term investment. Income Taxes Before turning to the element of operating expenses in the determination of the revenue requirement, a word should be added concerning the relation of income taxes to the permitted return. No one would question the need to allow rates suffi- cient to cover direct taxes on property, but the question whether the return allowed an electric company should be be- fore or after income taxes was once a subject of some contro- versy in the United States. This was largely ended as a matter of law in 1922 by a decision of the United States Supreme Court which, in an opinion by Mr. Justice Brandeis, declared that, in calculating a fair return, "all taxes which would be payable if a fair return were earned are proper deductions."'3 This view has now been generally accepted. In the Latin American countries studied, taxes appear to be treated the same as they are in the United States, except in Mexico. There the rate of return is calculated before taxes. However, as the recent Mexican rate actions reported above suggest, the rate of return before taxes has to be set with a view to what will be left for the shareholders after the com- pany's tax liability has been satisfied. A frank recognition that 12 In 1955, dividends on all United States Class A and B electric utilities con- stituted 72.4% of earnings available for common stock. Federal Power Comm'n, Statistics of Electric Utilities in the United States, 1955, Privately Owned Com- panies (Washington, D. C.), Table 5, p. xiv. 13 Galveston Electric Co. v. Galveston, 258 U. S. 388, 399 (1922). 124 Part 1. Problems of the Industry this must be done is provided by the rule requiring the return to be calculated after all taxes payable by the company. Operating Expenses and "Regulatory Lag" The ratio of operating expense to income that is projected in establishing the general level of a utility's electric rates may actually be exceeded in experience. This is usually the conse- quence of price rises unless there can be off-setting economies. The effect will be to reduce the projected return to the in- vestors. Though the company has fixed income securities (bonds and preferred stock) outstanding, the entire excess of the actual over the projected ratio of expenses will fall on the return available to the common shares. If, for example, the ratio of operating expenses to revenues has been 80% and the operating expenses then increase by 4 percentage points, the effect will be to reduce the total net return by a fifth. If the common shareholders are entitled to 50% of this return, their portion will be reduced by 40%. This drastic blow to the expectations of the common share- holders (expectations that the regulatory body had approved) is not the result of some unforeseeable upheaval. Indeed, cost increases on the scale assumed are particularly likely in an in- flationary economy. Moreover, the 80%, operating ratio, though high for an enterprise which relies predominantly on hydro- electric generation, is not an unrealistic figure for a Latin American enterprise employing chiefly thermal plants or op- erating mainly as a distributing company. The harmful effect of upward movements in operating costs on the return actually received by electric companies is accen- tuated if the intervals between rate proceedings are long, if the duration of the proceedings themselves is protracted, and if getting interim relief is difficult pending a general re-examina- Rate Regulation in the Electric Industry 125 tion of rate levels. The problem caused by these time factors has been the source of increasing concern in the United States. There it has come to be termed the problem of "regulatory lag." In Latin America, where prices tend to move more rapidly than in the United States, the problem is even more serious. One method of combating regulatory lag is to authorize in advance the making of increases in electric rates whenever in- creases of specified kinds of costs take place. The first and most common form that this method has taken in the United States has been the "fuel clause." If the price of fuel used by a utility increases by a specified amount, then the clause allows the utility to make a specified corresponding increase in its electric rates without obtaining the commission's permission. Origi- nally used only in contracts to supply electric power to indus- trial consumers, fuel clauses have come to be customary for all classes of customers in a number of the United States. The principle has been extended to other costs in some jurisdic- tions.14 In Latin America, the approach to this mode of escalating electric rates has been hesitant. Brazil has allowed fuel clauses for some time and, in regulations issued in February, 1957, to implement the electric power provisions in the Code of Waters, has provided a carefully developed scheme of escalation for three factors in electric utility costs: (a) changes in the cost of fuel or purchased power, where applicable; (b) compulsory increases in salaries or social welfare benefits; and (c) changes in payments of interest or amortization on loans where foreign currencies must be purchased or where indexed debentures have been issued by the National Bank for Economic Develop- ment. The plan calls for adjustments in monthly billings to cover the first two purposes and for semi-annual adjustments 14 Escalation clauses for tax increases have been accepted in a few states. In Florida, escalation of rates is also keyed to a construction cost index. For a de- scription of the Florida practice, see W. E. Wright, "Florida's Rate Adjustment Plan," 53 Public Utilities Fortnightly (1954) p. 157. Sec note 2, p. 157, infra. 126 Part 1. Problems of the Industry in billings for the third. Retrospective studies are prescribed at the end of each six-months period during which adjustments are in effect, with power in the supervisory staff to cancel any excessive adjustment and require return of the excess to the consumers.15 In Mexico, escalation of rates is authorized by statute for increases in fuel, purchased power, labor and taxes, but appli- cation to the Rate Commission has always been required for the first two types of cost increases, and, since 1949, for the latter two as well. Chile, Colombia, and Costa Rica appear in general to require commission approval for increases reflecting increased costs. Thus, when the Chilean regulatory agency authorizes an increase in the price of power sold by Endesa to the Compafiia Chilena for resale, the latter company must pay the higher rates and apply to the Commission for an increase in its resale price. The costly interlude pending approval can only mean a reduction in the return that had been projected for Compafiia Chilena when its rate structure was approved. In both the United States and the Latin American countries studied, the regulatory body may disallow expenditures which it regards as imprudent or as not benefiting the public served. If disallowed, these expenses will not be taken into account in the calculation of future rates. In exercising a power of this character, a commission is obliged to reach a nice judgment as to the permissible limits of managerial discretion. Observation of the workings of regulation in Latin America leads us to be- lieve that greater freedom is allowed managerial discretion in the United States than is typical in the countries studied. In some instances, rather serious disallowances were made from the actual expenses of the regulated companies on the ground that they were not adequately justified. The regulatory body is always, of course, under a great temptation to take advantage of opportunities to deny rate increases which the disallowance of various categories of expense can provide. The response to 15 Brazil: Decree No. 41,019 of Feb. 26, 1957, Art. 176. Rate Regulation in the Electric Industry 127 such pressure may be greater if action in granting rate increases must ultimately be approved by political officials. Depreciation As has already been noted, depreciation is a major factor in the cost of an electric utility because of the unusually high amount of capital invested. Unfortunately, however, the public understanding of depreciation is limited. Since depreciation seldom requires the immediate outlay of cash, it is always tempting to disregard its demands when the effect of doing so will be to postpone or to minimize an otherwise necessary rate increase. The investor in a utility is entitled to the return of his investment as well as to a return upon it. Since a utility must continue to perform service, the investment returned is ordinarily reinvested in new plant. Moreover, the consumer's stake in the problem is a real one. Ultimately, it will become necessary to retire plant and to replace it with more costly plant. Should not the cost of the retirement at least be spread over the customers who have benefited by the plant which, having been consumed in service, must now be retired? At the time when a number of Latin American regulatory laws were enacted, the United States had not accepted the necessity for charging a proportionate part of the cost of utility plant to the consumers benefiting by its use each year. This is now done by straight-line depreciation and its variants. Then, however, the retirement reserve system was in wide use.16 Under it only enough depreciation expense was charged to provide a sum sufficient to even off the cost of retirements from year to year. If large retirements and replacements were 161t was included in the 1922 classification of accounts recommended by the National Association of Railroad and Utilities Commissioners (NARUC). See Foster & Rodey, Public Utility Accounting (New York, 1951) p. 355. 128 Part 1. Problems of the Industry in order, they had to be handled by the issuance of new capital. The result was that annual depreciation expense and accumu- lated depreciation reserve both were low. Beginning in the late thirties, commissions in the United States under the leadership of the Federal Power Commission threw over this system for straight-line depreciation.17 How- ever, in Latin America the change has not been uniform. In Mexico, for example, the retirement reserve system is still in effect. In Chile, the contract governing the large Chilena en- terprise prescribes an annual expense equal to 20% of net operating income. Since net operating income has represented only a small fraction of gross-say 10%-the allowance of de- preciation expense reflected in the company's electric rates is only about 2% of gToss-a sixth or a seventh of the amount that might be found in the accounts of a utility in the United States. In Colombia depreciation practice is very irregular, despite the recognition in the Colombia law of the need to re- flect depreciation in the rates. The Costa Rican contract with the Compafifa Nacional de Fuerza y Luz provides for a rate of 21/20%,, of the cost of the property but the reserve built up by this role is subject to an over-all limitation of 30% of the depreciable assets. In a com- pany that is growing steadily such a reserve is likely to be ade- quate. Even though a company is not allowed to include an ade- quate allowance for depreciation in calculating its rates, it may still charge an appropriate amount for depreciation on its books. If, for example, it is allowed only 1% of plant for de- preciation expense, yet allocates 3%, to this purpose, the 2% allocated in excess of the permitted allocation must come from 17 The widespread adoption of straight-line depreciation (and, in "original cost" states, the deduction of the depredation reserve required by this method from the rate base) stems from a notable report of the NARUC Special Com- mittee on Depreciation Principles and Methods in 1938. For its text, see NARUC. Proceedings, 1938, p. 438. Today, Federal tax advantages are causing a swing to variants of the straight-line method: the declining balance and the sum-of-the- digits methods. Rate Regulation in the Electric Industry 129 the company's income available for dividends. Sometimes such a policy has in fact been followed, notably in the case of Mex- light.'8 This accounting practice makes clear the fact that the earnings actually realized under the law amount to a net figure considerably below the legal target. Foreign Exchange The fact that a considerable share of the capital invested in electric enterprises in Latin America is borrowed from abroad and is payable in foreign money injects another element of uncertainty into electric rate-making. If the money of the country in which the enterprise is located depreciates in terms of the money in which the loan must be repaid, a form of ex- pense is incurred which could scarcely have been provided for when the rates were most recently established. Either a rate increase must be allowed or the projected return to the utility will not be realized, with the loss falling first upon the common shareholders or, in the case of a public enterprise, on the gov- ernment's proprietary interest. In nations which have a multi- rate foreign exchange system, another alternative exists. The company can be given a much more favorable rate of exchange than the free market (if any) had established. This, of course, represents a subsidy to the company wvith foreign, as compared with domestic, bondholders. As a subsidy it is subject to the criticism that it is concealed, uncertain, and not exacted of the persons benefited by the service subsidized. However, this form of subsidy has been granted by both Brazil and Chile to com- panies operating within them and has offset in some degree the deficiency in the electric rates they have been permitted to charge. Where the proprietary investment in a utility has been pro- vided by foreign capital, the utility is under no obligation to transmit earnings to its investors in foreign funds. However, 18 This case is reported in Chapter 111 at p. 35, supra. 130 Part 1. Problems of the Industry from the standpoint of the foreign investor, a deterioration in the exchange rate has much the same effect on the adequacy of an original cost rate base as does a rise in the domestic price level from the standpoint of the local investor. This can be illustrated by an example drawn from Mexico where rate bases were established as of 1941 and subsequent additions to them are expressed in units of the national currency at the time of investment. The Impulsora group of companies (American 8c Foreign Power subsidiaries) had a rate base as of December 31, 1941, valued in dollars at $63,076,000. They added $15,200,000 to their property accounts between that date and 1953, yet, at the end of 1953, the Impulsora rate base was only $41,188,000 and, at the exchange rate prevailing during most of 1954, was down to $28,502,000. The declines are due, of course, to the succes- sive devaluations of the Mexican peso. Yet Mexico has not experienced as serious an inflation as some of the other coun- tries studied. Since the studies on which this report is based were com- pleted, recognition that the existing mode of regulating the electric industry is unsatisfactory under current conditions has led to extensive study of the need for new regulatory laws, notably in Brazil, Chile, and Mexico. However, despite the seriousness of the situation in all three countries, no basic re- vision of existing laws has yet been enacted, a few amendatory laws having been adopted to ameliorate some specific problems. Indeed, the leadership in remodeling electric energy laws to adapt them to modern economic developments appears to have been taken by a Latin American country not included in the study: Peru.19 Since we believe that the electric power industry, both pub- lic and private, can expand in Latin America without com- peting for scarce government resources with other socially im- 19 Law No. 12,378, D.O. July 14, 1955. Rate Regulation in the Electric Industry 131 portant needs if it can attract domestic savings for investment in electric enterprises, we have presented in Part II of this re- port a number of suggestions that look to regulatory principles and methods which, if put into practice, would render the in- dustry self-sustaining. We hope these ideas may prove of some value to the designers of the new regulatory legislation that is now long overdue.20 20 For convenience, the following over-simplified tabular view is presented of some main features of the regulatory systems in the five countries studied: Costa Brazil Mexico Chile Colombia Rica Rates controlled: By regulatory commission? Yes Yes Yes Yes Yes By franchise? Yes Yes Yes Federal, state or municipal control? Fed. Fed. Fed. Fed. Fed. Is rate-base x rate-of-return method used? Yes Yes Yes Yes Yes Type of rate base Orig, Orig. Reprod. Orig. Fair Cost Cost Cost Cost Value Is rate base adjustable? No No Yes Yes Yes Rate of return allowed 10% Reason. 10-15% Reason. Is rate of return after income taxes? Yes No Yes Yes Yes Is depreciation or retirement accounting followed? Dep. Ret. Dep. Dep. Dep. Annual rate of depreciation 5-8% M of 2-5% Rates "Crucial or retirement (as % of next 5 based point" gross plant) allowed in antnual on U. S. theory fixing rates retire- age-life nients( data Is depreciation rate adjust- able for higher prices? No No No No No Are automatic adjustments used: For higher fuel costs? Yes No No Yes Yes For foreign exchange costs? Yes No No No Contract For higher labor costs? Yes No No No No For purchased power increases? Yes No No No No For higher taxes? No No No No Are government-owned suppliers regulated? Yes Yes Yes Yes Yes PART 2 Suggestions for Corrective Action CHAPTER VI General Considerations The Problem of Capital Expansion The review in Chapters II through V of the situation pre- vailing in the electric power industry in the five Latin Amer- ican countries studied reveals an extraordinary opportunity for bold and constructive action by regulatory agencies and electric authorities, especially if legislatures can be induced to provide, where necessary, the statutory bases for the new poli- cies that seem clearly indicated. Even the brief consideration in Chapter II of the forces now at work in the rapidly growing economies and populations of the Latin American countries serves to disclose the persistent pres- sures for expansion that are being exerted on the electric power industry. The industry must meet the demands for service springing from mushrooming urban populations (even in the medium-sized cities) and from the concomitant expansion of power-consuming industries. At the same time there remain almost wholly unserved the great rural populations of the Latin American countries, all of which are still predominantly agri- cultural. The pressure of demand on electric supply is seldom being 135 136 Part 2. Suggestions for Corrective Action eased for more than brief intervals by the new construction that has taken place. Brown-outs and other forms of rationing have been frequent. Poor service resulting from over-worked and under-maintained equipment has also been common-and has exacted its toll, especially of industrial users. Often new industrial firms have been forced to assure themselves of ade- quate electric service by investing in diesel generating plants. This alternative to the expansion of electric enterprises that serve the public is, from the standpoint of the nation as a whole, almost always uneconomic. While inadequate capacity is thus imposing concealed costs, the demand for electricity is being stimulated everywhere by the surprisingly low level of rates. This, as Chapter IV reveals, is well below that prevailing in the United States despite the numerous cost advantages that the United States electric in- dustry enjoys. In real terms, moreover, the level of rates in the countries studied has been steadily falling despite increases in rates expressed in money terms. Where inflation has hit hardest, this drop in the price of electricity relative to other goods and services has been precipitous. The low rate levels which prevail are usually embodied in rate structures that still reflect, despite modifications in certain cases, the objective common in the 1930's, namely, to induce greater electric consumption by promotional rates and so to achieve lower average costs by the greater use of excess capacity. With present capacity already overburdened, this objective has become an anachronism. Indeed, in some countries the effect has been to provide especially attractive low rates to those users best able to pay high rates and, as a consequence, to stimulate the use of limited foreign exchange resources in the importation of electric appliances. As has been observed in Part 1, the main impediment to the expansion of the industry's capacity has been its inability to command sufficient capital funds, foreign and domestic. This has been due to the impairment of existing investment and the low earnings that have been actually realized. Clearly General Considerations 137 the industry's opportunity to attract fresh foreign capital from private investors abroad must await a restoration of the confi- dence that was badly shaken by the experience in the depres- sion of the 1930's. It can only be won back by a continuing demonstration that sound investment opportunities actually exist. The ability that the industry should possess to attract savings from both domestic and foreign investors, individual as well as institutional, has been seriously curtailed by the in- flationary pressures that stifle investor interest in fixed-return securities and by regulatory practices that leave the return to stockholders far below the levels indicated by alternative em- ployments of capital. The earnings records of the private com- panies are uninviting and often irregular. The attraction for conservative capital which low-risk public utility securities are expected to exert-and do exert in the United States-has been. denied to most Latin American private companies by prevail- ing regulatory policies in economies ridden by, or susceptible to monetary inflation.' Nonetheless, the larger private companies, by the reinvest- ment of earnings and depreciation accruals, have succeeded in expanding capacity, though not to the extent that either they or the publics they serve would have desired. This expansion, however, has been at the expense of the dividends payable to shareholders. Naturally, resort to this policy does not spur them to the investment of fresh funds. As has been noted in Part I, the enforced inability of private corporations to expand adequately to meet rising demands for electricity has led inevitably to government investment. This, 1 'Ihe problem also exists in certain European countries. See OEEC, The Trend of the Selling Price of Electricity and its Relation to the Financing of New Plant (Paris, Nov. 1958) p. 20: "76. . . . the depreciation of the currency which has taken place in many countries since the war has, to a certain extent, induced in- vestors which formerly regarded certain sectors, such as the electricity supply in- dustry as a safe investment to switch to other industries which can provide them with an immediate and higher return." "14. In some countries, moreover, the private investor sees no point in buying electricity supply stock, because of govern- ment policy, which keeps electric rates down to such levels as to reduce the return on the capital already invested in the industry." (p. 10.) 138 Part 2. Suggestions for Corrective Action of course, does not compel subsidization-as the British and a number of other nationally-owned systems demonstrate-but subsidization has in fact occurred. Typically, in the countries studied, the government-owned electric producer, whether operated by an autonomous corporation or by a government agency, has not obtained revenues sufficient to cover the cost of its service, including a realistic allowance for the cost of capital and for the depreciation of plant in service. Indeed, one of the ways in which subsidization has been made necessary has been by denying to the government-owned utility the priv- ilege of earning a return sufficient to enable it to appeal directly to the investing public for the investment of their funds. The resulting government financing of electric expansion by means which do not draw on existing savings has a double-bar- reled effect on the industry's problems. First, the new financing is likely to be inflationary in its impact on the national econ- omy and so to accentuate the industry's difficulties in meeting rising costs and in arranging future financing. Second, the sav- ings that might have been tapped by the industry for its ex- pansion are left free to be invested in more profitable enter- prises which swell still more the unsatisfied demand for electricity. The Latin American government which is prepared to allow the industry rates enabling it to cover the cost of the serv- ice it provides-and to do so even in a period of considerable inflation-will, it is submitted, find that the industry (whether publicly- or privately-owned) will become able to finance its own expansion-and by ways that are not inflationary. At the same time, both the correction of uneconomic rate levels and structures and the effect of drawing on savings for investment in the industry would operate as counter-inflationary influ- ences. What does this entail in terms of regulatory policies? First, the industry should be able to adjust its rates rapidly enough so that operating cost increases will not prevent it from realizing an adequate level of earnings. Second, as a means of enabling the industry to compete with General Considerations 139 unregulated investment opportunities for national savings and foreign investments, provision should be made for regular ad- justments in the capital accounts of electric enterprises, public as wvell as private, to reflect significant changes in the domestic price level and, when relevant to outstanding securities, in the exchange value of the national currency. Third, the regulatory authority (and any financial authori- ties having jurisdiction) should be prepared to approve forms of securities that are adapted to the industry's needs in an inflationary economy. They should adopt measures, where necessary, to open up markets for the distribution of those se- curities among institutional investors and perhaps should re- quire investment by such of the electric industry's business and even residential customers as make especially heavy demands upon its capacity. As will be seen in the discussion of possible ways in which these policies, if adopted, might be implemented, their inescap- able initial effect would be to raise most electric rates. Rate increases bringing the level of electric rates more nearly into line with the level of prices generally would not compensate the industry for the lean years it has been experiencing, but they would bring it to a point where fresh investment would be adequately rewarded. Subsequent rate increases would come only in response to subsequent price rises (including rises in the cost of capital and wages) and would, of course, take place only as and if such rises occurred. To make provision against inflation does not bring on inflation, but to leave an industry exposed to its ravages is to deny the industry the power to grow by other than inflationary means. Fortunately, as the data in Part 1 reveal, few nations are bet- ter able than the Latin American to absorb corrective electric rate increases. Since the very small consumers contribute only a small fraction of most suppliers' total revenues, a decision on grounds of social policy to continue their subsidization by not extending the indicated rate increases to them fully would not cast an oppressive burden on the other categories of users. 140 Part 2. Suggestions for Corrective Action To offer below-cost rates to other residential users is to en- courage uneconomic uses of electricity and stimulates a larger demand than overloaded facilities can meet. Certainly com- mercial and industrial users have no claim to be subsidized, especially in the light of profit rates that have been common in Latin American trade and industry. And any claim by municipalities as users to continue to enjoy below-cost rates would be a claim to have electricity users subsidize taxpayers, a fiscal policy that can scarcely be defended. With these considerations in mind, the prospect of examin- ing the case for regulatory policies that would for a time have the effect of raising Latin America's phenomenally low rates does not seem as forbidding as might otherwise have been the case. To repeat the view expressed in the opening chapter of this report, today emphasis needs to be placed on plentiful power, not on cheap power. Yet electricity which is priced at levels adequate to meet its cost is still likely in most Latin American countries to be cheaper than alternative sources of power. It is likely to grow cheaper still as the industry's growth makes possible the installation of economical modern facilities for generation and distribution. A prosperous, adequately fi- nanced electric power industry, capable of taking continuing advantage of the tremendous technological progress of the in- dustry (to which nuclear fission may come to contribute), would before many years deliver power that would be both plentiful and cheap. In the ensuing chapters, a number of suggested measures for advancing the three policies listed above will be considered. They are presented, not as prescriptions for particular coun- tries, an undertaking for which this overall survey can scarcely have qualified us, but as a generalized approach to the prob- lem. We view them as means of implementing the basic prin- ciple presented in the succeeding section, and it is the realiza- tion of that principle that is of basic importance, not the particular methods by which this is done. We do not insist on the exclusive merit of the suggestions we are advancing, but we General Considerations 141 believe they are worthy of exploration by regulatory and finan- cial authorities. Moreover, we think they serve to bring the issues of policy confronting the industry and its regulators into sharper focus than do more conventional formulations. We are confident that, if the authorities in any nation decided to adopt the three policies we have put forward, they would find it prac- ticable to adapt our more specific suggestions to their nation's special circumstances. Before the implementing suggestions are put forward, the principle on which they are based will be discussed in the re- mainder of this chapter, together with certain general problems relating to their application, timing, and administration. The Cost-of-Service Principle and Factors Impeding Its Application The basic principle upon which the measures suggested in this study will be rested is a simple one: that the aggregate of the revenues derived from the provision of electricity should cover all the elements of cost incurred in providing it. This cost-of-service principle is equally applicable to the govern- ment-owned power installation and to the privately-owned electric company. It is not concerned, however, with the allo- cation of the cost of service among classes of electricity consum- ers; it seeks simply to assure that the cost be borne only by con- sumers. It would allow no part of the cost to be cast upon taxpayers by subsidizing public power authorities or privately- owned companies; upon security holders in electric enterprises by denying them returns sufficient to maintain the integrity of their investments; or upon the entire economy, as by the resort to inflationary borrowing to cover the consequences of a failure to provide sufficient revenues to meet the industry's cost of capital. 142 Part 2. Suggestions for Corrective Action Conditions may, of course, be encountered in the develop- ment of a region within a nation that for a time will preclude the application of the cost-of-service principle there. The elec- tric service required for the region's development may cost more than its consumers can hope to pay until the region reaches a more advanced stage in its development. The alloca- tion of a part of the cost of electric service to the taxpayers of the nation may therefore be inescapable. However, in such a case preservation of the principle would require that the nature and purpose of the government's subvention be made clear and not concealed either by an understatement of the service's costs or by a failure to provide adequately for them. Such special situations apart, our study has persuaded us that the electric industry in Latin America has the intrinsic capacity to cover its cost of service. We recognize, however, that the legal and institutional impediments to this achievement are many. Even the task of defining "cost-of-service" and, particularly, of applying the concept to actual operations under the economic conditions prevalent in many Latin American countries today is a difficult one; indeed, the analogous task in the more ma- ture and stable economy of the United States has led to much controversy. Yet general acceptance of the broad outlines of the concept may be anticipated: "cost-of-service" should include not only operating expenses and maintenance but also ade- quate provision against the depreciation of the property and the annual cost of providing the capital employed in the enter- prise. The latter cost will, from the standpoint of public and private investors, be viewed as interest and profit, but, from the standpoint of those who depend on the continuance and expansion of the industry's service, it is a cost that must be met. The application of the cost-of-service principle in Latin American rate regulation involves no major departure from the theory of rate-making that Latin American governments have generally recognized as applicable both to privately-owned electric companies and, where subsidization has not been an avowed policy, to governmentally-owned installations. Rates General Considerations 143 have always been expected to cover operating costs and main- tenance. The common inadequacy of the provision for depreci- ation has been due chiefly to continued adherence to theories of depreciation that are now widely viewed as mistaken. It does, not represent a denial of the appropriateness of taking depreci- ation into account in determining rates. Finally, the calculation of a return to investors by reference to a rate base reflecting either the "fair value" or the "original cost" of the property has represented an attempt to determine a reward for capital that would compensate its contribution to the total product. If the theory of rate-making embodied in existing Latin American laws and regulations could have been faithfully applied in an economic environment of the kind for which it was originally conceived, the results would not have differed widely from those which would be Teached by the application in the same environment of the suggestions advanced in this study. What then have been the economic environmental factors that have caused the aggregate revenues yielded by the rates allowed private and governmental electric enterprises to fall short of the cost of the service rendered by those enterprises? Without constituting an exhaustive catalogue, the following appear to have been the most consequential: (1) Persistent and often rapid rises in operating costs, ac- companying inflationary rises in the general price level. These have meant that calculations of the costs of service on the basis of which rate increases have been authorized are often out- dated before the new rates can even be put into effect. This is the problem of "regulatory lag." It is not confined to Latin America; it is perplexing both government and industry in the United States today. (2) The inability to achieve compensating economies by increasing capacity and load factors. Capacity and load factors for some Latin American installations are so high already as to endanger the quality of service and require higher main- tenance. Fluctuations in supply and difficulties in achieving interconnections impair the efficient utilization of other plants. 144 Part 2. Suggestions for Corrective Action (3) The inability to achieve compensating economies by the large-scale replacement of old, high-cost plant by new, low- cost plant. It has been necessary to use such new plant as has been built largely to supplement rather than to replace existing installations, however inefficient, and, in many areas, new con- struction has not been on a scale sufficient to bring down over- all costs materially. (4) The rising cost of capital which has seriously impaired and frequently destroyed the market for securities having a fixed or stable monetary return of the sort that traditionally utility enterprises have issued. This condition has been accen- tuated by two factors (themselves a reflection of the inflationary pressures on the capital market): (i) The relatively small flow of funds to institutional in- vestors (savings banks, insurance companies, and pen- sion funds) which are primarily interested in stable monetary return as distinguished from stable purchas- ing power. (ii) The competing attraction of "equity" investments in industry, commerce, and real estate, the return on which can rise with the rising price level and which afford the prospect of capital gains. (5) The inability to reduce to fixed amounts in the national currency the cost of servicing obligations payable in foreign currencies even when these obligations are contractually fixed. This difficulty springs from the depreciation-or the risk of depreciation-in the national currency in terms of the cur- rencies in which the obligations are payable. The first of these environmental factors has been acute in most of the countries studied. But even where it does not exist in acute form-where, for example, the rise in the price level is no greater and no more rapid than in the United States-it can still bring about a serious,failure to cover the cost of service in the absence of compensating economies of the types noted in the second and third of the factors listed above. These econo- mies have done much to make slow-moving regulatory ma- General Considerations 145 chinery tolerable in the United States. Their absence means that any price rise that does take place will have a direct and immediate effect upon the adequacy of the rates to cover the cost of service. Moreover, the full impact of this deficiency must fall upon the residual claimants to revenues-the shareholders or the government's proprietary interest. In other words, it is the cost of capital that is most likely to remain uncovered, if regulatory machinery cannot respond promptly to rises in operating costs. The consideration just noted is a constant threat to the sta- bility even of the monetary return on electric investment, and a stable monetary return has been the traditional attraction of investments in regulated industries enjoying monopoly mar- kets. But, as the fourth factor indicates, even where this sta- bility of monetary return is present, its effectiveness in attract- ing capital to the electric industry is diminished in national economies that have not developed large-scale investment insti- tutions interested primarily in assuring a fixed money income. In such economies-and the Latin American are of this type- fixed-return securities must compete with investments in real estate and in industries and commercial securities that respond to changes in the price level. Moreover, in rapidly developing economies like the Latin American, even those countries that have avoided serious price inflation may find that capital short- ages produce rising rates of return for capital investments in industry, commerce, and rental real estate. The result, of course, is that the fixed-return utility securities steadily lose attractiveness. IJtility rate-making theory and machinery have always as- sumed a pattern of a cost of capital that remains relatively fixed over long periods of time between rate adjustments. For the reasons noted, this is not likely to correspond to the economic actualities in the Latin American environment. Moreover, as is indicated by the fifth factor listed above, constancy of capital costs is not to be assured by casting the obligations of the elec- tric enterprise into a more stable foreign currency. That tactic 146 Part 2. Suggestions for Corrective Action simply exposes the actual cost of capital to the fluctuations of the exchange rate. The regulatory theory and rate-making machinery that the Latin American countries studied have adopted, largely on the basis of United States experience, presuppose an economic environment that is in sharp contrast with the picture broadly sketched above. The theory and machinery assume stable eco- nomic conditions that would require changes in electric (and other regulated utility) rates to take place only at substantial intervals of years. The regulatory system has assumed, as does that in the United States, a long-continued stability in utility costs achieved either as a consequence of the stability of the general price level or as a consequence of compensating econo- mies. As in the United States, it has been concerned much more with preventing electric rates from rising too high than with preventing their falling too low. When such a maladjustment is found between the conditions under which regulatory principles and machinery are designed to function and the actual conditions that confront them, then either the principles and machinery must be altered or the conditions. Obviously it is to the potentialities for helpful change in the theory and practice of regulation that this study's proposals must largely be confined. With a view to the basic principle stated earlier in this chapter, the changes suggested are designed to adapt the machinery of regulation to its eco- nomic environment so that the rates established under regula- tion will yield revenue equal to the cost of service. The ancillary measures that are also suggested go beyond the direct regulation of electric rates to the problem of devising securities and sources of financing that will aid the electric industry in taking full advantage of the revised scheme of regulation to obtain the capital it needs for expansion. General Considerations 147 The Corrective Measures Proposed and the Problem of Timing In considering measures that could achieve the three policy objectives listed on page 138, we have given consideration not merely to their substance but also to their timing. To devise and put into operation a new system of rate regulation is time- consuming. While disagreements among the architects of the new system are being ironed out and the accounts of the private companies and government agencies are being restated in con- formity with its requirements, continued inflation may be wors- ening the industry's situation. The temptation for the govern- ment to cut through the entanglement by resort to inflationary deficit financing will be growing. Moreover, delays in setting up the new system would augment the initial rate increases it called for and so might excite political opposition to the whole scheme. We have therefore sought, in formulating our sug- gestions, to propose an initial plan that could be put into effect rapidly and halt any further spread of the gap between revenues and cost of service. Moreover, this plan contemplates the pos- sibility of a succession of small rate increases at relatively short intervals. These would reduce the need for a large increase at a later stage in the process of change. At the same time they would require the consumers of electricity to face the fact that the price of electricity is not exempt from the operation of economic forces. With such a plan of rate adjustment in operation, the development of more definitive measures could proceed with due deliberation. This plan, described in the succeeding chapter, Chapter VII, may aptly be termed a plan for the automatic compensatory adjustment of electric rates. It represents only an extension of rate adjustment methods that have been tested in operation 148 Part 2. Suggestions for Corrective Action both in Latin America and in the United States. Moreover, a plan of compensatory adjustment has the happy feature of lead- ing to rate increases only in the event increases in the cost of service occur. And it provides a two-way process. When the cost of service declines, so do the rates. A plan of automatic adjustment builds on whatever may be the existing rates. But our study has indicated that the existing rates not only are often deficient in failing to reflect increases in operating costs where these have occurred but also depend, for calculating the return to the investor, on rate-base x rate-of- return formulas that are unsuited to economies which may ex- perience periods of rapid or persistent increases in the price level. These formulas either call for the time-consuming, con- troversy-breeding calculation of the reproduction cost of the existing plant or instead insist on the use of its original cost as the rate base. The former alternative invites the worst evils of regulatory lag; the latter exposes the investor to the impact of inflation and operates to discourage investment in utility enter- prises. Specific suggestions to correct this situation cannot be made in this report since, if the problem is to be attacked, it must be done on the basis of conditions as they exist in each country and with respect to each enterprise. However, whatever pro- cedures may be adopted for adjustments to meet past deficien- cies, proposals for procedures to be followed in making future adjustments can be advanced with more assurance. This has been done in Chapters VIII and IX. In Chapter X, considera- tion is given to the adjustment of depreciation reserves. If plans for the restatement and subsequent adjustment, when necessary, of the capital accounts of electric enterprises were adopted as suggested in Chapters VIII, IX, and X, and if these had been preceded by the adoption of a plan for the automatic compensatory adjustment of rates, the attractiveness of investment in electric enterprises would be greatly enhanced. However, the hoped-for benefits would not be fully realized in those countries where the actuality or the risk of inflation-or General Considerations 149 the attractiveness of investments in equity securities of non- regulated industries-had already alienated investors fromn fixed-income securities. If this difficulty is to be avoided, suit- able securities must be made available for investors in electric enterprises. Their returns cannot be fixed; otherwise they could not pass on to investors any higher level of earnings that might be yielded as a result of rate adjustments to reflect increases in the price level. Measures to provide suitable securities are considered in Chapter XI. In that chapter are also considered possible steps by which electric enterprises might obtain access to some of the savings now invested in electricity-consuming industries or construction. These measures could fairly be contemplated only if electric rates were placed on a cost-of-service basis. If suc- cessful, they would increase the interest of Latin American in- vestors in the electric industry. The industry has not been starved for capital because Latin America lacks the capital to invest in it. The capital has been invested elsewhere because the electric industry, both private and governmental, has not been able to offer the inducements necessary to divert capital from its rivals for investors' funds. CHIAPTER VII Automatic Compensatory Adjustments Automatic compensatory adjustments in electric rates to reflect changes in the cost of electric service represent the first of the three categories of measures suggested in Chapter VI for change in the regulatory laws and practices now governing the electric industry in Latin America. In unregulated industry, compensatory adjustments in price are taken for granted when- ever any major cost increase occurs, at least in the absence of countervailing market forces. Why should not an equivalent opportunity be accorded the regulated electric industry? And should not adjustments downwards be required whenever im- portant cost reductions occur? Since rate adjustments by a regulated industry must keep within the framework of regulation, the amount of an "auto- matic" increase cannot be left to the discretion of the regulated enterprise. This does not mean, however, that the regulatory agency's prior approval must be obtained before rate increases are made in response to cost increases-or for rate decreases, in response to cost decreases. Instead, the amount of the rate in- crease (or decrease) required to reflect an increase (or de- crease) in one of the elements of the cost of service should be determined in advance. This can be done by means of a for- mula embodied in an adjustment clause in the supplier's rate schedule and approved by the regulatory body. 150 Automatic Compensatory Adjustments 151 The proportion that any one element of cost bears to the total cost of the electric service rendered by any particular supplier is reasonably stable. Hence, the extent to which an increase in a given cost element would increase the total cost of the service can be predicted with substantial assurance. Minor fluctuations in cost can be ignored; the formula can provide a margin of tolerance for cost changes too slight to justify corresponding rate changes. The workability of formulas of this sort has been thoroughly tested in the so-called "fuel clauses" long employed in industrial power contracts in the United States and now being extended to all classes of customers in an increasing number of states. Extensive use is already being made of such clauses in Latin America. The chief objection to automatic compensatory adjustments that must be anticipated goes to their automatic character. Com- pensation for cost increases may be proper, it will be argued, but only after approval by the regulatory agency. The ques- tion, though going to procedure, is crucial. The real danger against which the proposal is aimed is "regulatory lag," the in- ability of regulatory commissions to act promptly on applica- tions for compensatory rate adjustments. Experience shows that a commission cannot act quickly enough to prevent the appli- cant electric enterprise from having to absorb the cost increase out of the return that it has been allowed. Even though the rate increase requested is ultimately granted in full, the loss sus- tained in the interim cannot be recovered if the increase is to operate prospectively. Experience in the Latin American countries studied indi- cates that regulatory lag is an even greater problem than in the United States: the cost increases tend to be sharper; the regu- latory machinery moves more slowly-or not at all. Since the authorization of an automatic adjustment would relieve the electric company of this risk entirely-and also relieve over- burdened regulatory bodies of a task that they may be unable to execute properly-the arguments against permitting adjust- ments to be automatic would have to be very weighty to justify 152 Part 2. Suggestions for Corrective Action continued denial. Let us examine the adverse arguments to see whether they measure up to this standard. (1) It may be contended that electric companies will tend to overestimate the rate adjustments needed for particular cost increases if the adjustments do not have to get prior approval. This can be guarded against effectively by requiring the ap- proval of each formula for the reflection of cost increases in rate increases. If there were substantial doubt as to the ade- quacy of a formula, it could be approved on a contingent basis and rate increases authorized subject to refund, as will be more fully explained below. Approval of a formula, moreover, would always be subject to withdrawal if, in the light of operating experience employing it, the formula was found to be inac- curate. (2) It may be proposed that, instead of authorizing auto- matic compensatory rate adjustments, the regulatory law should empower the regulatory body to include a temporary surcharge in any rate increase designed to compensate for a previous cost increase. This surcharge could be sufficiently high to repay over, say, the period of a year the increased cost experienced by the electric enterprise before the issuance of the rate order. As a substitute for automatic compensatory adjustments, this sur- charge procedure is unsatisfactory. It is burdensome to the consumers who during the surcharge year must bear twice the amount of the cost increase (and those thus burdened may not always be the same as those who had benefited from the unduly low rates during the period of regulatory lag). Moreover, in the interim pending the allowance of the surcharge, the com- pany's financial position may have been jeopardized. And if the purchasing power of the country's money has been falling, the surcharge will not yield as much in purchasing power as the uncompensated cost increase had cost the utility. (3) It may be argued that, even though the formula for a given cost element is correct, the increase in total cost due to the increase in that element may be offset either by some other economies not covered by an adjustment clause or by some new Automatic Compensatory Adjustments 153 source of revenue. It will be stated, quite correctly, that an electric supplier is not entitled to a price increase simply be- cause one of its costs has increased but only if its total costs have increased more than its total revenues. Hence the regula- tory body should be free to examine the supplier's entire pic- ture before permitting its rates to be increased. The answer to this argument lies in the actualities of the electric industry, especially in Latin America. The pressure of rising prices on costs has not been sporadic; it has tended to be "across the board" so that cost increases have not been counter- balanced by decreases caused by concurrently falling prices for other elements of expense. Moreover, as has already been noted in the previous chapter, the reduction in average costs that comes from increasing capacity and load factors is not avail- able in most Latin American countries. The argument has point only when the introduction of a big new installation, operating on a lower cost basis, offsets for a time cost increases in whole or in part. But this is an opportunity that can be seized when it occurs-and such installations come at unhappily long intervals. When a big new installation was put into service, then there would be an apt occasion-and ample justification- for the regulatory body to require the electric supplier to re- calculate and resubmit all its adjustment formulas, taking ac- count of the effect of the new installation on costs. (4) If an electric supplier is assured that its cost increases will be promptly reflected in rate increases, then, it is con- tended, the supplier will be on a cost-plus basis and have no incentive to keep costs down. This argument, like the preceding one, ignores the realities of the industry's operations, in par- ticular, the relatively small control it has over the most im- portant elements in its costs. In some instances the cost increase flows directly from price or wage increases approved by the government or by industries controlled by it; in other cases, it is the inevitable, if indirect, consequence of increases author- ized by the government. In still other cases, cost to the electric supplier is controlled by a world market. 154 Part 2. Suggestions for Corrective Action Whatever the source of the cost increase that was compen- sated for, however, the management of the electric enterprise would continue to have an incentive to hold down costs by efficient operation. The setting of a utility's rates on the basis of projected costs and earnings neither assures the utility that it will actually cover its costs and earn the return nor limits its earnings to those projected in the rate-making process. The actual result will depend in part on management efficiency. If the management succeeds in reducing the costs which it can control and if it receives timely compensation for the costs it cannot control, the enterprise will have a greater chance of actually realizing the returns that were projected in setting the rates. It may even, on occasion, exceed them. (5) It may be argued that the escalation of electric rates in a period of inflation will simply accentuate the inflationary spiral. This contains only a grain of truth. Increases in electric rates can, to be sure, lead to some price increases for other com- modities and services, but, as is shown in Part I, their signifi- cance in living costs and in most industrial costs is so small that the inflationary impact will be minuscule. What the argument ignores is the inflationary effect of a rate policy that results, through regulatory lag, in denying in fact to the electric sup- plier the return that the supplier was in law entitled to earn. With its ability to command new capital thus impaired, the supplier cannot satisfy needs for new capacity. Industrial cus- tomers must either go without or buy diesel generators (both steps being inflationary in effect) or the government must turn to inflationary borrowing to construct new capacity. Moreover, one may question the justice of a policy that casts on a regu- lated industry the burden of absorbing the price increases oc- curring in unregulated industries in order to protect the latter industries from a small increment in their costs. (6) A final argument is that automatic adjustment clauses would lead to frequent changes in electric rates and that this would be resented by users. In our judgment, frequent small changes in rates, if required by cost increases, are to be pre- Automatic Compensatory Adjustments 155 ferred to less frequent but proportionately greater increases. Once established, the former policy would work to the advan- tage of the electric industry in its relations with its customers and would diminish the intrusions of political considerations into the regulatory process. The long intervals between rate changes that have been customary have tended to create the expectation that electric rates will remain stable indefinitely. This leads to surprise and customer dissatisfaction when the belated rate increase does come along, with a necessarily sub- stantial jump in rates to meet the cumulation of cost increases in the interval. Rate schedules with adjustment clauses would prevent the expectation of stability from taking root. The ad- justments themselves would be small since they would not re- flect a long build-up of cost increases. As a result, they would cause no dislocations in customers' businesses and would, we predict, be accepted quite ungrudgingly. It should be recalled, moreover, that the adjustment clauses we suggest would be two-way in their operation. This should aid materially in securing their acceptance. The customer would have the assurance that, if ever the tide should turn in the direction of lower costs, he would have the benefit of that change in the form of lower rates-and that without regulatory lag. Where automatic adjustments are in effect, their operation should, of course, be subject to scrutiny by the regulatory body and appropriate reports required of all electric suppliers using adjustment clauses. Suppose it became apparent that an adjust- ment figure was wrong and that the change had resulted in over-compensation. The regulatory body could require the supplier to ascertain the amount of the overcharge to each cus- tomer and to reimburse him by appropriate credits on his next electric bill.' Since not many customers of any given supplier chiange over a year's span, it should not be hard to reach most of them in this way. Direct cash refunds could be made to those 1 The Brazilian procedure for this purpose is described on p. 125, sutpra. 156 Part 2. Suggestions for Corrective Action who had ceased to be served. A few might be missed, but their number would be inconsequential. In most countries, the regulatory bodies could authorize the systematic use of automatic adjustments clauses without statu- tory change. Since a clause of this character is a part of the rate schedule, it is approved when the schedule is approved. The fact that a schedule takes effect subject to specified conditions or may be varied in response to specified conditions should not require a new proceeding to pass on rates which become oper- ative in accordance with the schedule's terms. In the sections that follow, the main categories of cost that might be made the subject of compensatory adjustments will be noted. 1. Operation and Maintenance Expense The most important items of expense in this category are fuel, purchased power, and wages and salaries. As has been noted in Chapter V considerable experience has accrued in the United States and some in Latin America, especially in Brazil, which points to the feasibility of fuel clauses. Clauses covering purchased power have also been used. Of course, there must be some element of regularity in power purchases before changes in the price of the power purchased can readily be translated into formulas to govern changes in electric rates. Where a thermal plant has the capability of using alternative forms of fuel, naturally its rates would be adjustable only with movements in the price of the lower-cost fuel. And like condi- tions should be stipulated where the alternatives arose from options to use purchased power or to draw on hydro. However, the Latin American electric systems we have examined charac- teristically present fewer such problems than might be antici- pated in larger, more fully integrated systems. And, though the Automatic Compensatory Adjustments 157 availability of escalation for fuel costs might be a factor in the choice between a thermal and a hydro installation, this would not put a premium on an uneconomic choice. A thermal in- stallation that would otherwise be economically desirable should not have to be rejected for hydro on the ground that regulatory lag following rises in fuel costs might cause the thermal plant's earnings to fall short of permitted rates of return. The case for covering increases in wages and salaries by esca- lator clauses is reinforced in Latin America by the fact that labor rates and benefits are frequently embodied in contracts between the electric enterprises and the unions which must have government approval. This approval protects against the risk that the employer's ability to pass the increased labor cost on to the consumer by escalation would lead to improvident concessions to union demands. Moreover, in some instances, wage increases are the direct result of government action, as in the case of the minimum wage law in Brazil in 1954. An alternative to the use of a series of escalation clauses covering various elements of operating costs would be to pro- vide for escalation on the basis of movements in an index con- structed to reflect certain important cost factors in the opera- tion of electric systems. An example of escalation of this nature is provided by the use of a commodity clause which has been approved by the regulatory agency of the State of Florida.2 This is based on movements in an index which moves in the same direction as electricity costs. Some administrative convenience can doubtless result for both the regulatory bodies and the electric enterprises by the employment of such an index but at 2W. E. Wright. "Florida's Rate Adjustment Plan," 53 Public Utilities Fort- nightly (1954) p. 157. Escalation is granted for fuel, taxes, and change in an index of construction costs. For a comprehensive current report on escalation in the United States, see R. S. Trigg, "Escalation Clauses in Public Utility Rate Schedules," 106 Univ. of Penna. Law Review (1958) p. 964. Fuel clauses are found in 44 states, extending to resi- dential users in 17 states; tax clauses, in 84 states; and clauses covering various other costs, in 10 states. Express statutory authority is rarely found. 158 Part 2. Suggestions for Corrective Action the price of some loss in the precision afforded by the use of several specific adjustment factors. Moreover, the use of a single index raises questions of the allocation of costs as between de- mand charges and energy charges. This might be largely met, as in France and Belgium, by the use of separate indexes for high and low tension sources. An increase in labor costs could be applied to demand as well as energy charges while increases in fuel and purchased power costs could be applied to energy charges alone. 2. Taxes As we pointed out in Chapter V not only do direct taxes on utility property have to be taken into account in calculating a utility's revenue requirement but so too do taxes on its cor- porate income. All the Latin American countries studied, save Mexico, determine the return to the equity interest after income taxes. Such being the case, it becomes necessary for the regula- tory body, in passing upon the utility's projection of its revenue requirement, to assume taxes at given rates, presumably the then existing rates. What happens if thereafter the tax rates are increased? If the utility takes no action, the projected return available to the common shareholders will be reduced by the amount of the tax increase. No question of the enterprise's efficiency is in- volved. Moreover, once the tax rate increase is announced, the amount of the increase in the tax payment can be forecast with reasonable certainty. There should be no difficulty, therefore, in allocating it among the electric rates in the utility's schedules. Why then should the company be obliged to absorb the tax increase pending action by the regulatory commission? Here again, an appropriate formula in the rate schedule would enable the increase to be reflected immediately and accurately in the rates. Automatic Compensatory Adjustments 159 It is equally desirable that automatic adjustment be provided if the tax rate is lowered. If provision for this is not made in the rate schedule, the risk of delay before official action is taken is enhanced by the fact that the utility would have no economic incentive to move the regulatory body to act. 3. The Cost of Capital: Adjustments for Changes in Foreign Exchange Rates Provision for adjusting shareholders' claims to changes in the price level or in the rates of return required by capital can be made at longer intervals than can be safely indulged in the case of operating costs. Measures to this end are considered in the next chapter. Ordinarily changes in the interest charges payable by an electric enterprise occur only at such extensive intervals that adjustments in electric rates in response to these need not be automatic but can be allowed to await action by the regula- tory authority. A different situation, however, may be pre- sented if the utility has fixed obligation securities outstanding which are payable in foreign moneys or are payable in national money in amounts fixed in terms of foreign moneys. If the ex- change rate applicable to these obligations moves adversely to the national money, the utility will have to use more of its revenues to meet its interest payments than had been assumed in projecting its revenue requirement. Again the common shareholders-or, in the case of a public enterprise, the gov- ernment's proprietary interest-must absorb the loss. On the other hand, a movement of the exchange rate in the opposite direction would present them with a windfall. There seems no insuperable technical obstacle to the use of escalator clauses in this situation.3 If the exchange rate changes 3 The progressive new Brazilian regulations authorize semi-annual adjustments for this purpose. These regulations are described on p. 125, supra. 160 Part 2. Suggestions for Corrective Action were to come at infrequent intervals, as in Mexico, the escalator could be set to operate only at corresponding periods, say, semi- annually, even though this would not compensate wholly for a change taking place between interest payment dates. Where the exchange rate was allowed to fluctuate freely, the escalator might cause many small changes, up or down, to be made in electric rates, perhaps with each monthly billing, even though minor movements in exchange rates within prescribed limits could be disregarded. In summary, we suggest the prompt inclusion in rate sched- ules and contracts (except those for small residential con- sumers, who, as we have seen, contribute little to total revenue) of clauses requiring the automatic adjustment of rates for changes in costs resulting from changes in the prices of fuel, purchased power, and labor, in taxes, and to the extent needed to meet contractually-determined obligations, for changes in foreign exchange rates. CHAPTER VIIl Measures for Meeting the Cost of Capital: THE INVESTMENT BASE The preceding chapter discussed the need for automatic ad- justments in rates to compensate electric enterprises for in- creases in various elements of their cost of service. The pro- posals advanced in that chapter, however, looked to the adjustment of the existing level of rates in response to future cost changes; they were not concerned with the question whether that level was correct. But our study indicates that, in many instances, legal and other impediments to full-scale re- examination of rate levels have led to the perpetuation of rates at levels not only far below the level reached by prices gener- ally but also below the cost of service. The correction of deficiencies of this character calls for time- consuming studies and for extended discussions between the regulatory body and the regulated enterprise, perhaps for full- scale proceedings. In so far as these are concerned with operat- ing costs, no substantial difficulties need be anticipated. How- ever, the problem becomes one of great difficulty when it reaches the cost of-capital, including provision for depreciation. The process of achieving a solution in some countries may re- quire the enactment of new legislation and, in virtually all, a 161 162 Part 2. Suggestions for Corrective Action redetermination of the capital accounts. It is partly for this reason that our suggestions have stressed the importance of authorizing automatic adjustments to compensate for cost in- creases as a measure which could be put into effect without awaiting more extensive changes. Before advancing proposals for the periodic revaluation of the capital accounts of electric enterprises, we wish first to question, in the context of Latin American conditions, the methods that have been traditionally employed for determin- ing the cost-of-capital element in electric rates. In proposing an alternative to that traditional approach, however, we shall not be advocating a radical departure from past practice. What we shall present for consideration is a theory with respect to capital revaluation that we believe permits the realization of the same ends sought by the traditional approaches but more directly and more equitably than has been possible when those approaches have been employed. 1. A Critique of the Traditional Approaches to Determining the Cost of Capital The United States experience has served as a model for most of the regulatory legislation in the countries studied, and therefore it has been drawn on extensively in the description of the processes for the regulation of rates appearing in Chapter V. As was pointed out in that chapter, regulatory commissions in the United States have long resorted to the rate-base x rate- of-return formula for determining that part of the revenue requirement which is needed to meet the claims of the long- term creditors of the enterprise and also the "equity" interest.1 The use of this formula was due more to the tortuous course 1 The term "equity" is used so frequently in any discussion of the cost of capital for electric utilities in the United States today that we have yielded to the temp- tation to use it here. It represents all a company's capital and surplus that is junior to its debt capital. The term derives from the protection granted by courts Measures for Meeting the Cost of Capital. I 163 of United States Constitutional development relating to regu- latory law than to the considered conclusions of economists in the industry or in government. Moreover, in large measure, a commission's preoccupation with a property rate base, i.e. with the asset side of the balance sheet, reflected the unrelia- bility of the capital accounts which, before financial regulation became comprehensive, were diluted by "water" and inflated by arbitrary "write-ups." The asset side of the balance sheet appeared to provide a definite foundation on which income calculations could be based. But the division was deep between those who argued for the use of "present value" of the assets as a base and those who insisted on ascertaining the assets' original cost. As was noted in Chapter V, this was resolved by the thoroughly illogical compromise dictated by the United States Supreme Court in vhat stood for nearly fifty years as the controlling precedent, the case of Smyth v. Ames.2 It called for the determination of "fair value" by giving weight to both present value and orig- inal cost, a feat that, in any meaningful economic terms, was an impossibility. Neither side to this controversy was successful in defending of equity in England and the United States to debtors whose property was subject to mortgage-frequently the situation of the common shareholders. Strictly, pre- ferred shares represent an "equity" interest, but, for our purposes, we have treated preferred shares having contractually fixed claims as in the same category as debt. A participating preferred stock issue, having a claim to share earnings above its contract dividend rate, might, however, be included as within the equity interest. Since government bodies organized to operate electric utilities do not neces- sarily have share capital, it may be even more of a stretch to extend the concept of the "equity" to cover a power authority's proprietary interest. Nevertheless, on occasion for convenience, we have used "equity" or "equity interest" as terms sufficiently broad to include the government's proprietary interest in an elec- tric enterprise not having share capital. 2 169 U. S. 466 (1898) . The Smyth case was overruled in Federal Power Comm'n v. Hope Natural Gas Co, 320 U. S. 591 (1944), which left the commissions free to use any reasonable method of rate determination which was fair to both the consumer and the investor interests. The Hope case has not been followed by a number of state supreme courts. See Chapter V, note 5, at p. 116, supra. 164 Part 2. Suggestions for Corrective Action the proposition that, as a matter of principle, the assets of the utility should be valued. The protagonists of "original cost" were quite open in recognizing that what they were really try- ing to find out was not the value of the property but the dollar amount of the original investment in the utility enterprise, imprudent investments excluded. In the state of utility ac- counting then prevailing, that investment could best be identi- fied by finding out what had been spent, over the years, in building and equipping the utility's plant. The protagonists of "present value" were unable to value the utility by the economic tests of value, its present market value or its present earning power, since neither the property nor similar properties were for sale and its earning power would be the very objective of a rate proceeding. They there- fore fell back on the device of asking what would be the cost of reproducing the property being valued, however improb- able it would be that the particular property would ever be reproduced in the form in which it had actually been con- structed. What they were really seeking was not "the value" of the property but a way to adjust the claims of investors to changes in the price level. The mechanism that the "present value" protagonists de- vised in order to adjust the investor's claim to the rise or fall in the purchasing power of the dollar he invested was both unsound as a matter of economic theory and cumbersome as a matter of administration. The cost of reproducing an old plant has little relation to the cost of replacing the service it is rendering. (The latter is an economically significant con- cept but one that defies practical application.) The trend in construction costs has no direct bearing on the sacrifice re- quired of the investor in making or retaining his investment in the enterprise. Nor does it bear on the amount of money he should receive as income. Perhaps the most striking anomaly in the use of a reproduc- tion cost or "fair value" rate base has been the fact that, though all the assets of the enterprise were thus revalued upwards, the Measures for Meeting the Cost of Capital. 1 165 resulting change redounded wholly to the benefit of those security holders whose claims were not contractually fixed in amount. To illustrate the working of the principle, suppose that an electric company in the United States with $1,000,000 in mortgage bonds and $1,000,000 in common stock outstand- ing had assets whose original cost (including reinvested earn- ings), less depreciation, was $2,000,000. Suppose that, in a sub- sequent rate proceeding, the reproduction cost of its property, less depreciation, was found to be $2,500,000 and its "fair value" $2,400,000. None of the $400,000 increment of "fair value" over the original cost of $2,000,000 would enure to the benefit of the bondholders save as the common shareholders' larger equity would reduce the risk of default on the bonds. In contrast, the common shares would have a claim to a return on all the residue of the fair value after the value of the bonds had been deducted, or $1,400,000. The effect would be to give the common shareholders a windfall, an increment far greater than the loss in the purchasing power of their investment. The cumbersome character of the formulas employed to reach "fair value" was attested in countless cases. Long, costly proceedings before a regulatory commission were required to reach a reasonable opinion as to what the hypothetical repro- duction might cost. This determination was followed in most states by an essentially arbitrary reduction to "give weight" to the original cost of the property in order to arrive at "fair value." In the circumstances, both commission and utility not only hesitated to institute rate cases but, once embarked upon them, were unable to conclude them for months or, often, for years. By the time decisions were reached and appeals to the courts taken and decided, intervening price increases had fre- quently rendered inadequate the rate increases that had been authorized. The "original cost" theory, it will be recalled, is premised on the proposition that the utility investor's original dollar invest- ment should be used as the basis on which a fair return should be calculated. This has reflected confidence in the long-run 166 Part 2. Suggestions for Corrective Action stability of the dollar, even though short-run fluctuations have been expected to occur in its purchasing power with the swings in the business cycle. Surely, such a thoughtful student of economic institutions as Mr. Justice Brandeis, the great cham- pion of "original cost,"3 would not have supported a method of rate determination that chained investors to a return meas- ured by their original dollar investment if he had anticipated strong and continuing inflationary trends such as have persisted in the Latin American economies. Such trends whittle down the real value of the original investment. The belief has sometimes been expressed by defenders of the original cost rate base that investors in utilities, if threat- ened by inflation, would demand' higher rates of return and so would prevent any loss in purchasing power of the income from their utility investments. Regulatory agencies allow earn- ings at rates of return based on the cost of capital (which would respond to movements in the market for the stock) and would reflect this demand in approving rate increases. Their action, it is argued, would render any revaluation of the orig- inal cost rate base unnecessary. Unfortunately, this belief does not seem well-founded.4 The prospect of inflation will, of course, lead the investor to demand a higher return than he would require in a stable economy-indeed, it is this very prospect that has been leading investors to shun Latin American electric utility securities for which fixed or maximum rates of return have been prescribed 3 His dissent in Missouri ex rel. S.W. Bell Telephone Co. v. Missouri, 262 U. S. 276 (1923), is a source-book for the "original cost" (or "prudent investment") theory of rate regulation. 4 The argument that increases in the rate of return responding to the demands of equity investors in the market will not suffice to assure a fair return to the equity investor has been developed at length in W. A. Morton, "Rate of Return and Value of Money in Public Utilities," 28 Land Economics (1952) p. 91. Pro- fessor Morton's position has been attacked separately by Professors Clemons and Thatcher in 30 id., pp. 32 and 85 (1954). In United States, it can be argued that special protection against inflation has not been necessary to attract equity in- vestors to the electric industry. The situation in Latin America appears to sub- stantiate Professor Morton's forebodings. Measures for Meeting the Cost of Capital. 1 167 by contract or statute. However, a rate of return that would induce a new investor to invest money which had already de- preciated would not preserve the purchasing power of an in- vestor wvho had invested his more valuable money ten years before. For example, a return of, say, 15% might be required to satisfy the new investor whose money had only 50% of the purchasing power of the investor who had been content with 12% ten years before. But the old investor would require a return of 24% to maintain the real value of his original in- vestment. For this reason the test of the current market return on an original cost base is not likely to protect past investors. This being so, ought not regulatory commissions in countries using original cost rate bases simply to keep increasing the rate of return by fiat? If this were done with sufficient persistence, the rate thus boosted could yield a money income with a stable purchasing power, despite continuing increases in the price level. Though this is mathematically possible, is it practicable? If at the same time the percentage rates of return on all other investments were rising to the same extent, perhaps it might be. This, however, is unlikely. The laws of a number of inflation-ridden countries (Brazil and Chile, for example) have permitted the writing-up of corporate capital for industrial and commercial enterprises to reflect rises in the price level on tax terms that have not been prohibitive.5 The resulting revaluations have held percentage rates of return down while permitting money returns to in- vestors to increase. And, even where revaluations of capital are not permitted, an inflation will enable unregulated industries to boost their prices and money profits to levels far beyond those reached by regulated utilities. At the same time, they will be able to expand their net worth rapidly by reinvestment. The result again is a much more rapid increase in money re- turns than in percentage rates of return. 5The revaluations have usually not been available to utility corporations. In this respect, the Brazilian law described on p. 117, supra, seems exceptional. 168 Part 2. Suggestions for Corrective Action Suppose a regulated electric utility that had been held to a slowly-growing original cost rate base sought to maintain money returns to its investors in their normal relationship to money returns on industrial investments. This would require the utility's percentage of return to be lifted to a level well above the percentage rates of return enjoyed by unregulated indus- trials. This would create the illusion of a level of profits for the regulated enterprise so high as seriously to embarrass both the utility and the body charged with its regulation. When industrial shares were averaging 15 or 20%, on revalued capital or inflated net worth, a return of 25% might have to be allowed on the shares of a regulated electric utility, taken at their original value. As a matter of practical political judgment, this appearance of excessive profits on the part of a regulated in- dustry would be very unpalatable even though, because of their revaluation or inflation, the industrial shares were in fact providing a return of 30 or 40%7 on the original investment. 2. A Suggested Alternative to the Traditional Approaches to the Cost of Capital Since inflationary pressures on the price levels have been chronic in many Latin American countries, effective regulation of the electric industry must meet the problems this creates. If fresh capital is to be obtained from savings to finance ex- pansion, investors cannot be subjected to the almost certain erosion and dilution of their investment that adherence to the "original cost" theory entails. But this should not compel regulatory bodies to encumber their activities with the illogical and clumsy apparatus of reproduction cost determination. A simpler and sounder alternative is called for. This we believe is available. In our search for it, we have taken into account the fact that the cost of capital springs from Measures for Meeting the Cost of Capital. I 169 two quite different sources: on the one hand, the cost imposed by claims contractually fixed in amount for bond interest and preferred dividends; on the other hand, the residual claims to the earnings of the enterprise, the claims of the holders of common and participating preferred shares in the case of cor- porations and of the proprietary interest in the case of govern- ment authorities. One of the disadvantages of the rate-base x rate-of-return formula has been that, by obscuring this dis- tinction, it has confused the problem of meeting the contrac- tually fixed claims of the enterprise with the problem of re- warding the equity capital. It is the latter problem which has been the real source of difficulty. As long as bonds or other evidence of indebtedness are out- standing, the indebted enterprise must obtain revenues to cover debt charges or face bankruptcy. Though failure to pay preferred dividends does not impose as drastic a consequence, nevertheless, a utility having issued preferred shares must be able to earn and pay contractually fixed dividends on them or abandon all hope of attracting further equity capital to the enterprise. If, as we shall suggest at a later point in this report, bonds or preferred shares are issued with a provision for ad- justment of the interest or dividend rate to reflect movement in an appropriate index, the increases in payments that this provision may require are just as inescapable a cost of the is- suing enterprise's capital as the basic rates themselves. With respect to these pre-determined elements in the cost of capital, the regulatory commission which has approved the issuance of the securities should have no difficulty in ascertain- ing how much they have added to the issuing utility's revenue requirement. These securities may pose questions in the future whether a given issue should be called or refunded to permit more economical financing, but at no time should there be any question concerning the amount of revenues currently needed to service the outstanding issues. The nub of the cost of capital problem does not concern these pre-determined claims. It con- cerns instead the claims of the investors, private or public, who 170 Part 2. Suggestions for Corrective Action hold the equity (or proprietary) interest in the enterprise. To secure needed capital for expansion from these investors in electric utilities, what is called for are methods that, without undue administrative burden, will accomplish three objectives: (a) to increase (or reduce) that part of the utility's earnings allocable to these private and public investors' claims so as to reflect changes in the purchasing power of money; and, in so doing, in the case of privately-owned electric companies, (b) to maintain an equitable relationship between the claims of in- vestors whose investments antedated a change in purchasing power and the claims of subsequent investors; and (c) to main- tain a normal relationship between the market value of utility shares and those of industrial companies. Failure to accomplish these objectives will mean, in the case of publicly-owned en- terprises, a concealed subsidy to electric light and power con- sumers; in the case of privately-owned utilities, an inability to attract fresh capital to the enterprise. The main method by which all three objectives may be at- tained is to provide for periodic revaluations of those capital accounts of a utility which represent the equity or proprietary interest in the utility. (Revaluations might also be made in the accounts set up for such senior securities as by their terms provide for adjustments in the amounts payable to their hold- ers.) These revaluations would require only a restatement of the monetary value of the accounts adjusted to the extent necessary to reflect significant changes in the purchasing power of money, changes up as well as down. The rate of return al- lowed by the regulatory body would then be applied to the capital accounts thus adjusted, the rate being ascertained by reference to the requirements of the capital market-the cost of equity or proprietary capital. If that rate of return were to remain stable, the allowable money earnings on the revalued equity or proprietary interest would then automatically rise or decline with the price level. Confidence in the new principle of regulation should result in the re-establishment of an actual market for electric utility Measures for Meeting the Cost of Capital. I 171 securities. The rate of return required by that market should reflect the protection the plan would accord investors. With assurance that revaluations would protect them against infla- tion, investors would not have to strive for protection by de- manding a high rate of return. The revaluation feature might well lead to a bidding up of utility shares. If that carried their market value materially above their adjusted value of the capital accounts, this would be a signal to the commission to lower the rate of return. Revaluation of capital accounts could thus keep rates of return from escalating with a rising price level. Note that the revaluation that would be called for is in the capital accounts, not in the property accounts, of the enter- prise. It should be immaterial what might have been the pre- cise changes that had taken place in the cost of capital assets like the physical assets which, over time, the enterprise had acquired. We submit, therefore, that the measure of the ad- justment should be the percentage of the change-up or down -in the country's cost-of-living index since the previous capital revaluation had been made. Moreover, to avoid the risk of reg- ulatory lag, the period between revaluations should be fixed, should not be more than a year in duration, and should be susceptible of being shortened in face of run-away inflation. Why should the cost-of-living index be used as the measure of revaluation? There are several reasons. In the first place, the purpose of revaluation is to stabilize the real value, the purchasing power, of the investment in the enterprise and not to revalue the property. No reason exists, therefore, to go through the rigamarole of reproduction cost estimates or the application of a construction cost index to a laboriously com- piled inventory of all the items comprising an electric utility system. Secondly, the cost-of-living index provides the best measure of the purchasing power of investments in the enter- prise for the individual investor. Thirdly, the cost-of-living index reflects price movements over a broad spectrum of com- modities and services and so is more likely to move in conso- 172 Part 2. Suggestions for Corrective Action nance with the average price of other securities than would a more specialized index. Fourthly, the rate increases it dictates will help to keep electric rates in line with the prices of other commodities, including alternative fuels, and so reduce the risk that the undervaluing of electricity in relation to these fuels and other consumer prices would result in artificial in- creases in demand. Finally, in most countries in Latin America and, indeed, in other regions, reasonably reliable price indexes can be found only for the cost of living. To translate the foregoing suggestion for the periodic re- valuation of capital accounts into operating terms calls for the drawing of some distinctions among the accounts of the enter- prise and also for an explanation of some of the suggestion's accounting aspects. As has already been noted, the capital accounts to be revalued would be all those that were not contractually determined in amount. Obligations payable in a foreign currency could re- main stated in the balance sheet in that currency (with ex- planatory footnotes to indicate the effect on these obligations of changes in the exchange rate) or could be restated in the national currency on the basis of the prevailing exchange rate. However, changes registered in either way would merely reflect the actualities of the exchange rate and would not be a part of the proposed plan of adjustment. A somewhat similar situation is presented by obligations calling for changes not merely in interest and dividends but also in the amount of the principal of the debt or of the capi- tal value of the shares (as the case may be) in response to changes in the index. (Presumably these obligations would not permit decreases in the index to reduce the amounts payable below their original face values.) The balance sheet could continue to state the original amounts with explanatory foot- notes, or, for each type of security, an adjustment account could be set up to indicate the amount, if any, by which the claims of the holders of these securities were being increased or decreased by the operation of the index. Of course, the Measures for Meeting the Cost of Capital. I 173 amount in this adjustment would vary from accounting period to accounting period with changes in the index. Here again the adjustments would come in response to the securities' own contractual terms rather than to the plan for periodic revaluation of capital accounts we are suggesting. The capital account that would normally be the one to be revalued periodically under the suggested plan would be the enterprise's common share capital or, in the case of public power authorities, the government's proprietary interest, how- ever that interest might be evidenced. To provide a term that will cover the various forms into which the accounts to be re- valued might be cast, we suggest the compendious term, "in- vestment base." Initially, any amount added to the investment base by a revaluation might be entered in a new Capital Sur- plus Adjustment Account, identified as reflecting the revalu- ation of the accounts comprising the investment base for changes in the cost of living (if that is the index used). This account and any like accounts used to reflect changes in indexed senior securities, as suggested above, should be balanced on the asset side by an identical Capital Assets Adjustment Ac- count, not by a change in the plant accounts. The latter should continue to be kept on an original cost basis. After a time, if the changes in the cost of living reflected in the Capital Sur- plus Adjustment Account seemed firmly embedded in the na- tional price structure, the utility might be permitted (or even required) to issue a stock dividend covering all or a major part of the Account and thus bring it into the regular capital struc- ture of the enterprise. Needless to say, this should not be the occasion for the levy of a tax on the amount of the surplus thus capitalized; no profit would be realized by the transaction. Regardless of whether the Capital Surplus Adjustment Ac- count had thus been capitalized, the amount of the revaluation reported in it would be included in the investment base. The rate of return allowed the common shares (or the proprietary interest) would be applied to the investment base in calcu- lating the required net revenue and the rates necessary to pro- 174 Part 2. Suggestions for Corrective Action duce it. In this manner, the effect of the revaluation of the capital accounts would be carried into the electric rates and so into the earnings of the enterprise available for dividends or for reinvestment. If a substantial downward trend in the cost of living in any country were to lead to a persistent negative Capital Surplus Adjustment Account, this would not only have to be deducted in calculating the investment base but in time would have to be reflected in a reduction in the capital of the enterprise. Al- though reversals in the upward trend of prices in particular Latin American countries may be anticipated from time to time, it now seems unlikely that these would ever be so pro- tracted and so deep as to require such recapitalization. The suggestion we have advanced may appear to discrimi- nate against securities contractually fixed as to income and principal in the national currency. Though these senior secu- rities would gain in strength as the suggested revaluations increased the earning power of the securities junior to them, they themselves would not be revalued for the decline in the purchasing power of their fixed monetary claims. It is exposure to this decline which has done much to make such securities unattractive to investors, and we are therefore suggesting later in this report that future issues use protective devices such as index clauses and participation and conversion features. But for a law to make present investors in the fixed obligations of electric utilities the beneficiaries of upward adjustments in income and principal would be to discriminate against all holders not only of securities junior to the adjusted securities but also of all other fixed obligations, government bonds, for example. Conceivably an electric utility corporation or a gov- ernmental power authority might conclude that it would be desirable to refund its senior securities with fixed returns by substituting for them securities with adjustable rates of re- turn on the theory that, by thus enhancing the attractiveness of its senior securities, it would benefit in the long run. If so, perhaps the regulatory agency should not stand in its way. But, Measures for Meeting the Cost of Capital. I 175 over the objection either of the equity interest or of the con- sumers, a regulatory agency could scarcely be deemed to have legal authority to require such action. 3. How the Suggested System Might Be Instituted and Operated If the suggestion presented in the preceding section (or any other form of periodically revaluing the investment base) were adopted, the problem of the proper starting point would have to be faced. The objective of the process we have suggested is to permit future determinations of the cost of capital to be predicated on the revalued capital accounts of an enterprise rather than on its property accounts. But one may well ask: do the existing capital and surplus of the enterprise represent a suitable point of beginning? In their origin, they might have been affected by the loose corporate financial practices that prevailed almost everywhere in the period antedating govern- ment regulation of utility rates and security issues. Since the early investment period, moreover, electric enterprises in many Latin American countries have had to withstand the stresses of successive waves of price inflation and exchange deprecia- tion. These have distorted capital values, depreciation reserves, and the enterprise's ability to remit income to its foreign in- vestors. They have led sometimes to subsidies and special arrangements with the government. Given financial histories of this sort, we believe that ordinarily the most satisfactory approach to a fair starting point in the capital accounts of an enterprise will lie through its property accounts, even though the financial history of the enterprise should be taken into ac- count in appraising them. The initial determination of the enterprise's investment base (i.e., the capital accounts comprising the equity interest) 176 Part 2. Suggestions for Corrective Action would seldom be a problem that could be solved by the appli- cation of a general formula. Rather, it would have to be worked out on a case-by-case basis. The difficulty would not be due simply to unreliability of past accounting. Even if an enterprise had maintained accounts which accurately disclosed the amounts of the original investment, retirements, and depreciation, as well as additions out of depreciation, retained earnings and new security issues, the task of ascertaining the investment base would not be a simple matter of arithmetic. A dilemma like that noted in our discussion of property bases would be encountered here. If the depreciated original costs of the property were adopted as the measure of the investment base the effect in many coun- tries would be to deprive investors of most of the value in real terms of their original investment. But to adjust that original cost figure to reflect the entire intervening change in the cost of living would probably be excessive. It is one thing to under- take to adjust fully for changes in the purchasing power of money for the benefit of future investors who will rely upon the protection that such adjustments give their investments and accept lower rates of return in consequence. It is another thing to give full adjustments retrospectively for the benefit of investors who could not reasonably have counted on such protection when they made their investments. The establish- ment of a fair initial investment base therefore would not ordinarily require the full reflection of the exchange and price movements that had taken place since the date of the origi- nal and all succeeding investments in the electric enterprise's property. Even where the data were all available, therefore, the ini- tial investment base should ordinarily be a frankly compromise figure, a product more of judgment than of accounting for- mulas. Although this "judgment" figure might seem reminis- cent of the "fair value" of Smyth v. Ames, it would be quite different in its effect on the regulatory process, because the initial investment base would have to be determined only once for any one enterprise. Therefore, the fact that it could more Measures for Meeting the Cost of Capital. I 177 readily and wisely be arrived at by round-the-table negotiation than by formal proceedings is not a serious source of objection. In arriving at a fair initial investment base, the regulatory commission would want to take account of changes in capital values in the economy generally, not merely in electric equip- ment and construction costs. The calculation of an equitable figure to serve as the starting point for future capital costs should give weight not only to physical and functional depre- ciation but also to the commission's own refusal to allow the enterprise to accumulate an adequate reserve for depreciation, if such had been the case. Since the investment base would represent the capital value of the enterprise after the deduction of the amount of outstand- ing bonds and other contractually fixed claims to share in the utility's earnings, it would be necessary to take precautions against giving a windfall to the equity interest in making the original determination. Some of the enterprise's capital might have been raised by the issuance of bonds or other forms of contractually fixed obligations. For the reasons noted in the preceding section these fixed obligations could not ordinarily be adjusted above their contractual limits. The benefit of the adjustment denied them should not be passed on to the junior securities, as could occur if the investment base were arrived at initially by revaluing the property accounts. The accounts reflecting the common shares or the proprietary interest should share in an increase in the property accounts only in the pro- portion which the equity accounts bear to the total capital of the enterprise. After the equity interest had been ascertained, it might be found that the enterprise was under- or over-capitalized. Per- haps it would be desirable for new common shares to be issued against part of a capital surplus that otherwise would be ex- cessive, or some shares might well be retired because the origi- nal capital had been impaired. In any event, the enterprise, whether private or public, should be permitted to earn a fair return on the investment base which would ordinarily be 178 Part 2. Suggestions for Corrective Action represented by the common shares and surplus or by the gov- ernment's proprietary interest. In order to determine the full cost of capital of the enterprise, it would be necessary to add to the fair return on the equity interest the prior claims of the holders of debt securities and preferred shares. The revenue requirement that the new electric rates would have to meet would have to cover the cost of capital so computed as well as operating expenses, taxes, and depreciation expense. If the automatic compensatory adjustment procedure pro- posed in the preceding chapter were operative, the new elec- tric rates thus established would be increased or decreased from time to time in response to changes in certain electric industry costs or to movements in exchange rates or in the price level. These electric rate adjustments would not be accompanied by any changes in the capital accounts until the prescribed inter- val between capital revaluations had elapsed. Then, the capi- tal accounts would be adjusted speedily by the application of appropriate adjustment factors, and new electric rates approved to give effect to the new valuations. Earlier it was suggested that the revaluation in the capital accounts be placed on an annual basis. This would enable the company or public enterprise to publish the revalued figures in the balance sheet in its annual statement each year and give investors during the year following an accurate indication of how big was the investment base on which their securities would be allowed to earn money. More frequent revaluations would not ordinarily be necessary, but discretionary authority to permit or require interim action in case of an unusually abrupt and wide price movement should be vested in the regu- latory authority. Violent upsurges in prices in some countries in recent years may have left investors wary of adjustment mechanisms that can operate only on an annual basis. Once established, the system of valuation should work smoothly, speedily, and without requiring any single change in the rate levels to be a substantial one, with the possible exception of that which followed the initial revaluation of Measures for Meeting the Cost of Capital. I 179 capital accounts.6 Moreover, it must always be kept in mind that if there were no changes in the level of prices in a country in the course of a year, the mechanism of revaluation and rate adjustment that we have proposed would be inactive, and, if the changes were small, their impact on the level of electric rates might be negligible. The device we suggest has the virtue of having to be used only when and to the extent it is needed. Before turning to a consideration of that other ingredient in determining the cost of capital in a cost of service system, the rate of return, we wish to append a note to our suggestion for the revaluation of the capital accounts of electric enter- prises as an alternative to the traditional course of adjusting the plant accounts. Though we believe that the former course offers significant advantages over the latter and introduces no important disadvantages, we recognize that its novelty may work against its adoption. We therefore wish to recognize the availability of the latter course-the periodic adjustment of the plant accounts. In a more cumbersome way, this method enables the capital accounts to be adjusted in rough approxi- mation to movements in exchange rates and the cost of living. We therefore do not reject the use of a reproduction cost rate base where this seems the only instrumentality that can be made available for capital adjustments. Where it is used, how- ever, we suggest that, as a starting point, a new rate base be arrived at in the full light of the history of the enterprise, much in the way that we propose that the "investment base" be determined.7 Moreover, we suggest that subsequent changes in that rate base be limited to that proportion of the total ad- 6 For smooth operation, problems of allocation between electric service and other functions of an enterprise must be held to a minimum. Probably separate incorporation should be sought for non-electric functions whenever possible. Where this cannot be provided, administrative convenience suggests that the capital involved in collateral activities not be excluded from the investment base and that the gross and net revenues they yield be taken into account in projecting the revenue requirement and the return to be received. Of course, large opera- tions like traction and telephone service would have to be segregated. 7 See pp. 176-177, supra. 180 Part 2. Suggestions for Corrective Action justment in plant accounts which securities that are not fixed in return bear to the total capital of the enterprise.8 Otherwise a windfall to the equity interest would result. Moreover, if resort is to be had to the reproduction cost rate base as a means of effecting capital adjustments, the design and use of reliable construction cost indexes to effect the revaluation seems essen- tial if adjustments are to be made annually and serious lag is to be avoided. 8 If the enterprise has indexed senior securities outstanding, it would be prefer- able to rely on their contractual terms in computing the revenue requirement than to depend on the redetermination of the reproduction cost of the rate base to provide the revenue needed to satisfy them. CHAPTER IX Measures for Meeting the Cost of Capital: THE RATE OF RETURN The rate of return used to determine the earnings allowed an electric utility has traditionally been expressed in terms of a single percentage applied to the rate base, however that base may have been ascertained. In both the United States and Latin America the determination of this single rate has been left to administrative discretion, but, as was pointed out in Chapter V, this authority was subject in the United States to judicial re- view and, in a number of Latin American countries, to statu- tory ceilings on the return allowable. Under this approach, in neither area was much attention paid to the actual cost of capital to a given company in the light either of its particular capital structure or the rates of return prevailing in the invest- ment market for the types of securities it had issued. The cost-of-service principle dictates a departure from this approach. If an electric enterprise has been able to issue senior securities bearing a fixed interest charge in the national cur- rency, that interest charge is the cost to it of the part of its capital that is represented by those securities. If it has securities issued in a foreign currency, the cost of capital is the contract rate adjusted for changes in the exchange rate. If its senior se- curities provide for an adjustment in their return to reflect 181 182 Part 2. Suggestions for Corrective Action changes in the price index (or in any other index), the cost of capital will be calculated at the contract rate, plus or minus the amount of the adjustment. If, instead, provision has been made for adjustment in the principal amount of such securities to reflect exchange or price movements, then fixed percentage rates of return applied to the adjusted principal amounts could have the same effect as adjusted rates applied to fixed principal amounts. When the cost of capital has been ascertained for all securities having a contractually determined rate (whether sus- ceptible to index adjustment or not), there remains the difficult problem of determining the rate that must be allowed the equity (including the government's proprietary interest in the case of a publicly-owned enterprise) . In the case of the pri- vately-owned enterprise at least, this is a rate that is not fixed by contract but is expected to respond to change in the money market. As the residual claimant to the earnings of the enterprise, the equity is exposed to the risk that the enterprise will fail to earn as much net income as the regulatory body and the investors have anticipated. Because of this risk, moreover, investors in the equity properly expect a higher return on the capital they have invested than is allowed on the enterprise's senior securi- ties. Equity capital is therefore more expensive than senior issues, but it is not for this reason dispensable. The factor of risk will exist and ought to be compensated for whether the enterprise has a single class of security or many. It is advantageous for an electric enterprise to have both senior securities in which the element of risk is minimized but the return is fixed (or adjustable above a fixed minimum) and also equity securities in which the risk is greater but the return more flexible. This diversity in capital structure enables an enterprise to issue the type of security that the market can best digest. For example, in a period when inflationary pressures are serious, fixed income securities are unattractive (unless adjust- able) and only equity securities may have a substantial market. In periods when deflationary pressures are evident or when Measures for Meeting the Cost of Capital. II 183 funds are available only from institutional investors seeking low-risk investment, the availability of senior securities may be essential to the attraction of fresh capital. As was noted in Chapter V, if an electric enterprise has more than one category of security, its cost of capital can be-and often is-expressed in a single rate of return by calculating weighted average return, weighting the rate for each category of security by the amount outstanding. But one would expect this single rate to be quite different from the single rate of re- turn that might be determined by the traditional method of applying to a rate base embracing the entire enterprise a single rate of return calculated without regard to its capital structure. Moreover, the traditional single rate of return allowed on a rate base can serve to conceal the "leverage" enjoyed by the equity when the senior securities have been issued at rates materially lower than the over-all return allowed the enterprise. In conformity to our basic commitment to the cost-of-service principle, we suggest that regulatory bodies in Latin America calculate the cost of capital separately for each category of capital and that, in so doing, they be freed from the straight- jacket of statutory maximum rates of return. If the cost of capital were in fact higher than the statutory maximum, the effect of that maximum would be either to deny to the enter- prise the fresh capital it needed or to require the government to conceal the true cost of capital by subsidization. 1. Determining the Rate of Return on "Equity" for the Privately-Owned Utility When a regulatory body in the United States today seeks to ascertain the cost of capital to an electric utility applying for an increase in rates, it will receive voluminous evidence of cur- rent rates of return and the ratios of the market price of the 184 Part 2. Suggestions for Corrective Action utility's securities to its earnings and to like ratios for other utilities of a similar character, even for non-regulated, high grade industrial corporations. The regulatory commission will try to determine what rate of earnings it would have to allow on the book value of the equity to enable the utility to market its equity shares (if it were to seek to do so) without diluting the value of the shares already outstanding. The latter condi- tion is important; in the United States, unless an issuer's finan- cial position were to appear precarious, its shares could almost always be sold-at a price. But sale at any price can work an injustice to the old investors. Fairness requires that the price at which new capital is raised does not allow the new share- holders to obtain a claim to earnings disproportionate to their contribution to the utility's real capital. In Latin America today, the task of arriving at appropriate rates of return for the various categories of securities issued by an electric enterprise would be seriously handicapped by the lack of free and active markets for utility securities in many countries. Where relative stability in price and exchange levels has permitted a reasonably active market in utility securities, the problem would be very much the same as in the United States, provided there were no statutory ceilings on allowable rates of return. Where, however, inflation has driven investors away from securities with fixed or limited return, no satisfactory market prices exist for guidance. For properly designed securi- ties with returns subject to adjustment by reference to an index, the new feature might render returns on superior industrial securities a reasonably good initial guide. However, in the case of equity securities, little prospect of developing a market test of the cost of capital will exist until provision can be made for automatic compensatory adjustments and for periodic readjust- ments in equity capital accounts. If, on the other hand, these changes could be introduced and securities issued in such form as to enable the investors (public as well as private) to enjoy protection from price instability in the economy generally, there is ample reason, we submit, to Measures for Meeting the Cost of Capital. II 185 expect the electric industry to attract a reasonable share of do- mestic savings.' In Latin America, with its great growth poten- tial, the industry enjoys almost unlimited opportunities for expansion. It is peculiarly well situated to provide investors a type of outlet for investment which they have long needed. Latin American investors have lacked access to securities that combined liquidity, low risk, and protection from inflation in the manner that properly-designed electric securities could achieve once governmental regulation of the industry took account of its inflationary environment. To attract new investment might at the outset require a rate of return that by the artificially low standards set by statutes in the past would appear high. However, as has been noted above, when investors came to appreciate the protection afforded them by the adjustments in the issuing enterprises' earnings and in the returns paid on the securities themselves, they would be content with lower rates of return. This change in appraisal would be reflected in higher prices bid for electric securities. Moreover with respect to equity issues, the return which is not controlled by contract but, under the cost-of-service principle, by the current cost of equity capital, the benefits of this market reappraisal (once clearly demonstrated) could be passed on to the consuming public from time to time in lower rates of re- turn on the equity portion of the industry's capital structure. Similarly, public corporations could be required to take a lower return on their proprietary interests. The prospect of increased rates of return should, we believe, be viewed in the light of the self-defeating low rates of return 1 For a similar expression of optimism concerning European utilities, see OEEC, The Trend in the Selling Price of Electricity and its Relation to the Financing of New Plant (Paris, Nov. 1958) p. 24: "111. On the whole the structure of the elec- tricity supply industry is still basically sound. When both public and private undertakings can be run, as in the past, on sound business principles (this being, incidentally, the method most in keeping with the social systems of the O.E.E.C. countries), the possibility of giving investment a yield which could be considered normal on the money market, and of building up adequate depreciation re- serves, would make it easier to attract the private investor." 186 Part 2. Suggestions for Corrective Action that have long prevailed. Even without taking account of the effects of inflation in reducing fixed-interest securities to nega- tive real returns, the prevailing rates of return are low. Con- sider, for example, a maximum over-all return of 10% on an electric utility's rate base in a country where interest rates on government bonds are at or near that figure. Such a ceiling means that the senior securities of the enterprise must yield about 10%, perhaps even less. But, as has been remarked, in the United States where the single rate of return on an electric utility's rate base is typically set at 6 or 7%, the utility's bonds may have cost only 3 to 4%. The result is that the return on the common may range from 8 to 12% of the common's book value, the 1955 average rate of earnings in Class A and B electric utilities being 10.8% of common equity.2 Note that these earnings of about 10% on book value (cal- culated usually at original cost less depreciation) are being per- mitted by utility regulatory commissions in the United States which ordinarily have sought to ascertain the actual cost of capital by extensive investigations in active securities markets.3 Note, too, that these returns are allowed in a country in which investment funds are relatively plentiful, where the electric industry enjoys a strong economic position, and where the earn- ings on many industrial securities are little greater. Assume that the earnings allowed Latin American electric enterprises were set so that net earnings rates on the equities would enjoy the same relationship to long-term bond yields and to earnings rates on industrials as are enjoyed by United States electric utility equities. On that assumption, Latin American electric equities would have to earn as much as three times their present authorized (if not realized) earnings levels. We do not advocate the use of these relationships as a test; we point to them simply to underscore how unrealistically low are the levels of earnings 2 Federal Power Commission, Statistics of Electric Utilities in the United States, 1955. Privately Owned Companies (Washington, D. C.) p. xiv, Table 5. 3 Most electric utility common stocks in the United States have market prices well above their book values. Measures for Meeting the Cost of Capital. II 187 with which Latin American electric enterprises have been ex- pected to attract capital. Little wonder that they have failed. In considering the earnings available to the common stock, the regulatory authorities should not limit themselves to the dividends that may be paid. For the reasons noted in Chapter V, our discussion has been cast in terms of earnings on common stock rather than of dividends. To provide the electric power industry with vast amounts of capital it requires each year, it must depend on loans, sales of stock, depreciation accruals and retained earnings. Conceivably, where capital markets are highly developed and capital is abundant, a company could pay out nearly all of its earnings as dividends and rely on sales of new securities for additional funds. Even in the United States, however, the electric power enterprises, in the aggregate, raise approximately 35% of expansion funds out of operating reve- nues, namely, by depreciation accruals and retained earnings.4 The privately-owned electric power companies in the United States pay out as dividends about 70% of their annual net profit and reinvest the remainder.5 This is done with the knowledge and approval of the regulatory commissions. The resulting reinvestment adds to the ownership account of the stockholders and increases the base upon which they are entitled to earn. Since such earnings could have been legally paid out as dividends, their withholding and use in purchasing plant and equipment means that the owners of the enterprises are reinvesting their own capital. To the extent that they do so, 4 In 1954, depreciation and retained earnings contributed 33.5%, in 1955, 36.1%, of the funds required for gross plant investment of privately-owned electric utili- ties. Federal Power Commission, op. cit. supra note 2, pp. xi, xvi, Tables 3, 6. Self-financing is comparable as a source of capital for the European electric in- dustry. See OEEC, op. cit. supra note 1, p. 29 (Germany: about 40% of total re- quirements for the next few years); p. 30 (Austria: "about 45% came from firms' own reserves and from capital payments"); p. 34 (United Kingdom: "at present about 40% of capital requirements are met from internal resources") ; p. 36 (Switzerland: "self-financing accounted for about 29% of this new investment in 1956.") . 5 Federal Power Commission, op. cit. supra note 2, Table 5. The 1955 pay-out ratio was 72.4%. 188 Part 2. Suggestions for Corrective Action it becomes unnecessary to raise additional capital from the sale of bonds or shares, as would be the case if all earnings were paid out as dividends and depreciation accruals were low. Where, as in Latin America, utilities lack adequate market outlets for new security issues, the importance of allowing earnings to exceed dividend requirements is especially great. Since Latin American investors are more accustomed to high pay-out ratios than are their counterparts in the United States, specific provision for lower pay-out ratios seems desirable. To aid the accumulation of funds for expansion, legislatures and regulatory bodies might expressly adopt the policy of encour- aging rather than discouraging the reinvestment of earnings. This policy objective would be advanced by provisions: (1) including undistributed profits in the equity capital on which the return to equity is computed-an indispens- able condition; and (2) authorizing the annual issuance of a common stock dividend equal to a percentage of the current year's net income (possibly 30%) and providing that no income tax would be payable on this stock dividend6 as long as funds equal to the amount thus capitalized are to be used for expansion of the public service facilities of the enterprise, as certified by the regulatory commission. Such a provision would not lend itself to the abuses against which are directed the tax laws that forbid evasion of taxes through stock dividends, since it would be controlled in amount and related to necessary expansion of the electric power indus- try, a limited-return industry. The same authorization could be broadened so as to provide also for the capitalization of an appropriately limited part of the Capital Surplus Adjustment Account proposed in the preceding chapter. The allowance of earnings sufficient to permit the reinvest- ment of a part of them may be objected to on the ground that, to the extent the earnings are not required by the shareholders 6 The rule which prevails in the United States that stock dividends are not sub- ject to the income tax does not prevail in all Latin American countries. Measures for Meeting the Cost of Capital. 11 189 for dividends, they must exceed the necessary cost of capital. This view, however, ascribes a shortsighted outlook to the utility investor. He is aware of the industry's need for capital and the difficulty and expense of having to raise all of it by new securities issues. He is therefore prepared to forego a part of the immediate return he might enjoy in order to build up his in- vestment and so assure participation in the growth of the enter- prise. A higher dividend would be needed to compensate for the absence of this growth feature and on top of that would come the cost of larger and more frequent security issues and the uncertainties of new financing. However, countries subject to inflationary pressures cannot stimulate the growth of utilities by encouraging the reinvestment of earnings not required for dividends unless they protect investors against decline in the purchasing power of their investment as, for example, by ar- rangements of the sort we have suggested. Otherwise, the more quickly an investor can withdraw earnings from a utility and invest them in an industry free to respond to price rises, the safer he will be. 2. Determining the Rate of Return on "Equity" for the Government-Owned Utility We have postulated a self-sustaining electric industry as a desideratum, and this objective is not changed by the fact that all or part of the industry is government-owned. An electric enterprise created by a governmental body can attract private capital by the issuance of debt securities at interest rates which are adjustable for changes in the cost-of-living or other suitable index provided, of course, the enterprise is permitted by the regulatory authority to earn a return sufficient to cover the charges on those securities adequately in case the charges rise in response to changes in the index used. 190 Part 2. Suggestions for Corrective Action However, this does not exhaust the problem of return. In all probability the governmental body creating the utility will have contributed capital to it. If not, it will almost certainly provide for building up its proprietary interest by the amortization of debt and the retention of earnings for investment in plant. How the governmental body's proprietary interest (which can aptly be termed its "equity") may be formally evidenced is not of critical importance to the problem of its return. Thus, the electric authority may be set up in corporate form with shares held by the governmental parent; the government may instead have taken junior obligations of the authority in re- turn for its advances, or the government may retain direct ownership of the utility's property. Whatever the form used, the failure of the electric utility to pay an adequate return on the governmental equity will mean subsidization pro tanto of the electric utility. Since, however, the governmental body will not, ex hy- pothesi, have bargained for the securities or the proprietary interest it has obtained, no direct resort can be had to the in- vestment market to determine the cost of this capital. Where there is a genuine market for the securities of the governmental body, it seems clear that the electric enterprise should earn a return at least equal to the interest rate that the governmental body itself would have to pay for capital. This return should be allowed not merely on the capital that the governmental body has advanced but also on any earnings that the utility has been allowed to retain and reinvest. Those retained earnings could have been withdrawn for other governmental purposes; if, instead, they are reinvested in the electric enterprise to aug- ment the service it can render, the cost-of-service principle re- quires that electricity consumers meet the cost of that capital. Moreover, we believe that the return to the governmental proprietor should not merely match the actual cost of capital to the proprietor. The electric rates should be set to provide in the return a factor over and above that cost in order to generate additional capital for use in expanding the plant and service Measures for Meeting the Cost of Capital. II 191 of the electric utility. Absent such a provision, the electric en- terprise, with its recurrent need of capital, would be danger- ously dependent on the vagaries of the investment market. (On occasion this might yield no capital or demand an exorbitant return.) Alternatively, the enterprise would have to draw further on its governmental parent's funds. In doing so, it might be depriving other needed public services having no earnings capability of the only capital to which they could have recourse. Where the governmental body has no genuine access to in- vestment funds and merely pays rates established by fiat on manufactured credit, these rates do not provide a helpful meas- ure for the basic element in the rate of return on the govern- ment's equity. It may be that returns on non-speculative indus- trial investment could be used instead.7 Perhaps, in some situations a guide for electric rates sufficient to yield a return which would cover this element and provide a factor for ex- pansion as well can be found by using the electric rates charged by a privately-owned, regulated public utility. The consumers would then have the assurance that they were paying no more for their electric service than they would if it were in private hands. Any surplus which the rates yielded could be used either for expansion or, if this were not in prospect, for other govern- mental services. No precise formula of general applicability can be developed for determining the allowance for expansion that should be included in the return on the government equity. The pro- spective need for expansion will vary from power authority to power authority. In one case, the prospect may be for continu- ing expansion of generating plant, especially to meet the growth of electricity-using industries. A substantial amount of self- financing here seems both advantageous and equitable. In a 7 Compare OEEC, op. cit. supra note 1, p. 21. After noting at para. 90, that the electricity industry may use funds derived "from fiscal taxes or other revenue sources," the report continues: "91. The interest from investment of this kind must not be less than that obtainable on the capital mnarket if national productive re- sources are not to be wasted." 192 Part 2. Suggestions for Corrective Action second case, the need for generating plant may be less than for distribution facilities, the expansion of which calls for a steady but modest annual outlay of capital funds. Especially if the utility is operating in an economy where capital is costly and difficult to raise, self-financing of most or all of these expan- sion costs may be practicable and economical. In contrast, in a third case, a large installation is indicated to utilize the hydro resources of a newly developed area. Here the possibility of a significant capital contribution derived from consumer rates is slight; the maintenance of a lower rate level, in so far as it helped to build up the utility's load, could prove more useful. Finally, there is the situation, doubtless largely hypothetical in Latin America, of a stabilized community with ample and readily expansible hydro resources at hand. Clearly here no expansion factor would be called for. It is no easy task to project a power authority's capital needs over a period of years and estimate how much internally-gener- ated capital from retained earnings and depreciation would be required to enable the authority to obtain at reasonable cost from outside sources the rest of the capital it would need for that period. The answer to this question, moreover, might in- dicate such a burden for future ratepayers that a smaller scale of capital investment would be called for. Obviously, decisions on issues of the sort cannot be embodied in any mathematical formula. They require the exercise of "a fair and enlightened judgment," but that has long been recognized as a necessary element in determining rates of return for utilities.8 8 It is stressed in what has come to be regarded in the United States as the dassic statement of rate-of-return principles by the Supreme Court in Bluefield Water Works & Imp. Co. v. Public Service Commtssion of West Virginia, 262 U. S. 679 (1923), where the Court said (p. 692): "What annual rate will constitute just compensation depends upon many circumstances and must be determined by the exercise of a fair and enlightened judgment, having regard to all relevant facts . . . The return should be reasonably sufficient to assure confidence in the financial soundness of the utility and should be adequate, under efficient and economical management, to maintain and support its credit and enable it to raise the money necessary for the proper discharge of its public duties. A rate of return may be reasonable at one time and become too high or too low by changes affecting Measures for Meeting the Cost of Capital. II 193 Since inflationary dangers beset government-owned elec- tric enterprises as well as those privately-owned, the investment base to which the rate of return would be applied ought to be adjusted periodically for changes in the cost of living (or other index) if the net revenues thereby yielded are to be adequate to meet the rising costs of expansion. A scarcely acceptable al- ternative in an inflationary situation would be to boost the rate of return on the original capital to a level that would seem ex- orbitant in comparison to the rates of return paid on other securities. Each year that the government-owned utility earned a fair return on its proprietary interest, and reinvested those earnings in plant, its equity would, of course, grow in absolute terms, though whether its proportionate share of the enter- prise's total capital would also grow would depend on the extent of the latter's borrowing from the public. In the United States, gradual accretion in the equities of the 262 government- owned (non-Federal) Class A and B electric utilities had, by 1956, brought the proportion of their total capital represented by the investments of their governmental parents (mostly municipalities) and by retained surplus to a little over one half, 50.2% to be exact.9 The dominant extent to which re- tained earnings had contributed to the building up of the equity is revealed by the fact that, while the municipalities' opportunities for investment, the money market and business conditions gen- erally." The "fair and enlightened judgment" called for in the Bluefield case is, of course, easier to exercise where the standard can be applied by reference to an active securities market, a guide to judgment to which (in the omitted portion of the quotation) the Court refers. 9 Federal Power Commission, Statistics of Electric Utilities in the United States, 1956. Publicly Owned Utility Systems (Washington, D. C.) p. vii, Table 2. The utilities covered comprise about 70%, on the basis of plant and sales, of the entire non-Federal part of the publicly-owned industry (60% on the basis of revenues and number of consumers served) . Id., p. vii. In 1956, these utilities served approxi- mately 3,536,000 residential, 558,000 commercial and industrial, and 148,000 other customers. Id., Table 13. They had total electric revenues of, roundly, $612.2 million. Id., Table 12. This total seems small, however, when compared to electric revenues of $7,780 million in 1956, for Class A and B privately- 194 Part 2. Suggestions for Corrective Action investments totaled $196,000,000, the accumulated surplus amounted to $1,413,000,000.10 Data with respect to the over-all earnings experience of these United States municipal electric utilities may provide some guidance for the formulation of rate policies for publicly- owned utilities in Latin America, although the latter are chiefly wholesalers of power to privately-owned distributing companies whereas the United States municipal power com- panies buy almost as much power as they generate,'1 and retail most of their output.12 In 1956, electric net operating revenues of the publicly- owned utility systems amounted to 7% of average net electric investment. The latter term, which corresponds to a net invest- ment rate base, embraces electric plant less reserves plus work- ing capital.'3 The depreciation reserves, it should be added, were ample: 22.4% of electric utility plant;'4 they were being owned utilities. Federal Power Commission, Statistics of Electric Utilities in the United States, 1956. Privately Owned Companies (WVashington, D. C.) Table 10. Federal power projects are excluded from the compilation of data in the FPC series on publicly-owned electric utilities because of their lack of comparability to the Class A and B electric utilities it covers and to the privately-owned electric utilities covcred in the FPC's other series. The Federal projects are typically huge, multi-purpose hydro installation projects. They are not financially auton- omous, depending on Congressional appropriations for expansion. Though the principal projects (Tennessee Valley, Bonneville, and Grand Coulee) have pursued policies designed to produce a net return on the Government's invest- ment, the question whether they have succeeded in achieving this goal has been the center of much controversy relating to allocations and tax payments. The scale of these enterprises, though large, is, of course, small in comparison to the United States budget. In the light of these facts, the experience of the United States in financing these projects does not seem helpful as a source of guidance for Latin American policy makers. 10 Federal Power Commission, op. cit. supra note 9, p. vii, Table 1. 11 Id., p. xvii, Table 16. In 1956 they generated 27.2 billion kwh (17.1 billion by steam) and purchased 24.6 billion kwh. Id., p. xvii, Table 15. 12 Id., p. xvii, Table 16. 13 The amount of working capital is not specified. The amount that appears to have been used in the calculation equals 5% of average net electric plant. For the data, see Federal Power Commission, op. cit. supra note 9, p. x. 14 Id., p. x, Table 4. Measures for Meeting the Cost of Capital. I1 195 built up by annual depreciation expense equal to 12.6% of electric operating income.15 Earlier in this chapter, wve urged that rates of return should be determined separately for each class of security instead of being stated as an over-all figure typified by the 7% rate of return reported above. To apply this principle to the publicly- owned utility systems of the United States, it would be neces- sary to find what the return had been on the governmental equity in the electric utility plant after provision had been made for payment of interest on the long-term debt. This cal- culation requires adjustment for utility income received from other than electric utility sources and a parallel adjustment in the interest on the debt. The result may not be precise, but even as an approximation, it is impressively high. Our calcula- tions show that, after interest, these utilities in 1956 in the aggregate earned 11.6% on the adjusted equity (average net electric investment less 83.5% of the average long-term debt) .J1 The rate of return taken by the publicly-owned municipal electric utilities in the United States is somewhat distorted for purposes of comparison by the fact that their payment of taxes and tax equivalents is so much lower in relation to revenues than is that of privately-owned companies. In 1956, for exam- ple, these payments by the former group amounted to only 15 Id., p. vii, Table 2. 16 The adjustments in arriving at the figures in the text were made as follows: To adjust for the fact that the long-term debt is an obligation of these systems' total utility operations and not merely their electric utility systems, the per- centage which the average net electric investment ($2,369 million) bore to the average total net utility investment ($2,835 million) was ascertained for 1956. This proved to be 83.5%. The data used for this purpose were derived from Federal Power Commission, op cit. supra note 9, p. xv, Table 11. An amount equal to the same percentage of the 1956 interest payments on the long-term debt, id., p. x, Table 3, was calculated, and this figure ($28.6 million) sub- tracted from the 1956 net electric operating income ($164.3 million) (ibid.). This yielded the (approximate) net electric income available for the equity or $135.7 million. This amount represents a return of 11.6% on the electric utility equity (the average net electric investment less 83.5%,, of the average long-term debt [$1,202 million]) . Id., p. xv, Table 11. 196 Part 2. Suggestions for Corrective Action 3%0 of their electric operating revenues17 whereas the latter's payments were 22.5.%18 Suppose the former group also had had a tax payment of 22.5% but had had no greater electric operating revenues than those the group actually received in 1956. In 1956, the percentage which the publicly-owned sys- tems' electric operating income and electric tax payments to- gether bore to electric operating revenues was 29.8%.19 If the tax component had been 22.5% instead of 3%, the portion of the percentage which represented electric operating income would have been reduced from 26.8% to 7.3%. However, such an adjustment is rendered highly artificial by the fact that half the tax burden of the privately-owned companies is attribut- able to the Federal Corporate Income Tax, and, if municipal electric utilities had not been immune from this tax, that cir- cumstance would certainly have affected their growth and capi- tal structures. Perhaps a more useful comparison for our purposes is pro- vided by the government-owned utilities which distribute power generated by the Tennessee Valley Authority. There the state and local tax or tax equivalent payments of the distribut- ing companies plus payments of state and local tax equivalents paid by TVA itself amounted to 6.7% of the distributors' elec- tric revenues in fiscal year 1957.20 Since the corresponding fig- ure averaged for a group of private utilities in the same area was 8.2 %,21 it appears that the government-owned utilities were not far out of line in this respect. Moreover, after having paid these taxes, directly or indirectly, the municipal distribut- ing companies showed for fiscal 1957 an aggregate net income of $23,854 million before interest and $21,822 million after 17 Id., p. x, Table 4. 18 Federal Power Commission, Statistics of Electric Utilities in the United States, 1956. Privately Owned Companies, p. xv, Table 5. For these companies, the percentage of electric operating income to electric operating revenues in 1956 was 20.7%. Ibid. 19 Federal Power Commission, op. cit. supra note 9, p. x, Table 4. 20 Tennessee Valley Authority, 1957 Annual Report, p. 44. 21 Ibid. Measures for Meeting the Cost of Capital. 11 197 interest and had accumulated earnings of $212.7 million as against capital contributions of $16.5 million and long-term debt of $245.2 million. Net plant was $279.6 million.22 The helpful consequences of making provision for expansion in electric rates are indicated, moreover, by investment data. Thus, in 1954, all TVA's municipal distributees invested an aggregate of nearly $39 million in plant construction, in pay- ments on long-term debt, and in increases in working capital.23 This total was provided by nearly $33 million in capital that had been internally generated (almost $10 million coming by way of depreciation) and by less than $7 million that had been raised by the creation of new long-term debt.24 Government-owned power enterprises in Latin America would doubtless find it prudent to set aside a larger portion of revenues for expansion than their United States counterparts because of the greater difficulty and cost of raising outside capi- tal actually needed. However, in shaping their policies with respect to rate of return, they might find some guidance in the contractual provisions which TVA writes into its contracts with distributing municipal power boards. These provisions control the resale rates the boards may charge and the use to which they may put their revenues. After referring to the scheduled rates and charges agreed upon, Section 5 of the standard con- tract continues as follows: (c) If the rates and charges provided for in said resale schedule do not provide revenues sufficient to provide for the operation and maintenance of the electric system on a self-supporting and financially sound basis, including re- quirements for interest and principal payments on indebted- ness incurred or assumed by Board for the acquisition, ex- tension or improvement of the electric system . . . Board and 22 These data are derived from a composite financial statement for the nearly 100 municipalities. Id., p. 45. 23 148 Distributors of TVA Power, 1955 Annual Report, p. 12. Many TVA dis- tributors are rural electrical cooperatives. They are included in this report but the few Federal projects that buy huge amounts of TVA power are not covered. 24 Ibid, 198 Part 2. Suggestions for Corrective Action TVA shall agree upon, and Board shall put into effect promptly, such changes in rates and charges as will provide the increased revenues necessary to place the system upon a self-supporting and financially sound basis. If the rates and charges in effect at any time provide revenues that are more than sufficient for such purposes . . . Board and TVA shall agree upon a reduction in said rates and charges and Board shall promptly put such reduced rates and charges into effect. Section 6. Use of Revenues.... (b) All revenues remaining over and above the require- ments described in sub-section (a) . . . shall be considered surplus revenues and may be used for new electric system construction or the retirement of electric system debt prior to maturity. Provided, however, that resale rates shall be reduced from time to time to the lowest practicable levels considering such factors as future circumstances affecting the probable level of earnings, the need or desirability of financ- ing a reasonable share of such new construction from such surplus revenues, and fluctuations in debt service require- ments.25 [Italics supplied.] We believe that a realistic application of these TVA policies and standards by Latin American public power authorities, in the light of conditions prevailing in the countries we have studied, would require resort to an adjustable investment base and the allowance of a substantially higher rate of return than public power authorities have heretofore been permitted to seek. The rates charged by the municipal power systems com- prising the distributors subject to the TVA contracts from which we have just quoted are not regulated by state commis- sions.26 In states having no commission regulation of govern- ment-owned utilities, the contracts pursuant to which bonds 25 Extract from the TVA standard contract with distributors (in this instance, with the Guntersville, Ala. Electric Board). Tennessee Valley Authority, 1957 Annual Report, pp. A 101-A, 102. 26 As to state commission regulation of electric rates, see Chapter V, note 1, at p. 110. Measures for Meeting the Cost of Capital. II 199 are issued to the investing public for the construction of elec- tric facilities commonly obligate the issuing body, whether a municipality or an autonomous power authority, to maintain rates at levels sufficient to meet not only the debt service but also operating expenses and depreciation and to provide funds from which transfers for other than electric power purposes are restricted.27 Whether government-owned electric utilities in any Latin American country should be freed from commis- sion regulation is a problem of basic governmental structure for the country involved, and into its consideration we do not wish to enter. However, we submit that, if government-owned electric utilities were to be relieved of regulatory commission controls, their rate policies should be guided by much the same standards as those embodied in the TVA contract quoted above. 27 An example of such a provision is to be found in the "Department of WVater and Power of The City of Los Angeles Electric Plant Revenue Bonds, Second Issue of 1955" reading as follows: "By the terms of the Charter [of the City of Los Angeles, §229] and by cove- nant expressed in the Resolution, officers of the City are obligated to fix rates and collect charges for service from the electric works of the City such as to provide revenues sufficient to pay . . . the principal of and interest on the Bonds, . . . in addition to paying . . . the necessary expenses of operating andl maintaining such works, including provision for depreciation of such works, and all other obligations and indebtedness payable out of the Power Revenue Fund. The Charter and covenants . . . in the Resolution impose upon the De- partment . . . other obligations to holders of the Bonds, restricting creation of further indebtedness payable from the Power Revenue Fund, restricting trans- fers out of that Fund to the reserve fund of the City, prohibiting issuance of bonds having any priority with respect to payment from the Power Revenue Fund and imposing conditions with respect to any sale of said electric works." CHAPTER X Handling the Problems of Depreciation The importance of depreciation expense and the depreciation reserve in electric utility rate-making is vividly shown by two facts derived from the composite statistics for all privately- owned electric utilities (very small ones excepted) in the United States in 1955. In that year, nearly one dollar out of every ten dollars obtained in electric revenues was required in order to meet depreciation expense. In the same year nearly one dollar for every five dollars of the original cost of electric plant had been placed in reserves for depreciation and amorti- zation.' In the Latin American countries studied, few electric utili- ties have been permitted by the regulatory authorities to charge rates for their service high enough to provide for depreciation expense, or to build up depreciation reserves, on a scale even approximating the foregoing. However, as has been seen in Chapter III, some privately-owned utilities met this cost out of their corporate pockets by diverting moneys allowed as earn- ings to depreciation reserves. In small part, the lower level of 1 Federal Power Commission, Statistics of Electric Utilities in the United States, 1955. Privately-Owned Companies (Washington, D. C.) p. xiv, Table 5. More precisely, depredation expense was 9.8% of gross electric operating income and depredation reserve was 19% of utility plant. 200 Handling the Problems of Depreciation 201 depreciation charges and reserves that appears prevalent in Latin America may be attributable to the higher proportion of hydro there since hydro generating facilities have longer lives than thermal. But social lag seems a more likely explana- tion; the attitude of the Latin American regulatory authorities is much the same as that prevailing in the United States, say, twenty-five years ago. This reflects the fact that the role of de- preciation in public utility rate-making is complex. Moreover, it is a factor that can readily elude scrutiny since it deals with problems which, if they are neglected, exact no immediate pen- alties. Three Functions of Depreciation One reason for the complexity of depreciation's role in util- ity rate regulation is that it is now expected to discharge three functions: (i) To allocate the prepaid cost of the utility's plant over the estimated life of that plant. (ii) To measure the loss in value of the plant comprising the utility's rate base. (iii) To maintain the integrity of the proprietary invest- ment in the utility. At the time the Latin American countries studied were pat- terning their regulatory systems along the lines of those in the United States, these functions were not clearly recognized here as constituting tasks for depreciation expense and the reserve built up by its accrual. On the contrary, many experts main- tained that the cost burden of retiring worn-out or obsolete plant was properly imposed on the users of the service at the time of the retirement and that the only regular charge needed to be imposed for depreciation purposes was a sum sufficient to even out the irregularities in retirement costs from year to year. This approach was embodied in the accounting standards 202 Part 2. Suggestions for Corrective Action which had been adopted by the authoritative National Asso- ciation of Railway and Utilities Commissioners (NARUC) in 1922.2 This "retirement reserve" method (still embodied in Mexican law) led to small annual charges and the accrual of small reserves. Depreciation and "Fair Value" In this period, "fair value" rate bases were employed by regulatory bodies in almost all of the United States. Since these were required to reflect present fair value, some measurement of the depreciation of the utility property being valued was nec- essary. For this purpose, the accrued reserves were disregarded; indeed, they could have little relation to the guesstimate that was taken for "fair value." However, engineers were called on to study the plant of the utility being valued and to report the percentages by which the various categories of property com- prising it had depreciated. This "observed depreciation" was then deducted from the figure representing "fair value" newv. Depreciation and Original Cost A change in the prevailing conception of depreciation's role came as the battle to establish original cost as a permissible measure of the rate base was being won, and as, under pres- sures exerted by the Federal Power and Securities and Ex- change Commissions, "water" was being squeezed from utility accounts. The turning point was reached when in 1938 NARUC's Special Committee on Depreciation recommended that all commissions require depreciation accounting in lieu 2 National Association of Railway 8c Utilities Commissioners, Proceedings, 1922, pp. 166-172; Uniform Classification of Accounts for Electric Utilities (1922) Acc't No. 251, "Retirement Reserve." Handling the Problems of Depreciation 203 of retirement reserve accounting and indicated a preference for the straight-line method of depreciation accounting.3 Under this method, the original cost of each category of plant being depreciated is allocated over its estimated service life in equal annual installments and charged as depreciation expense. The result is to distribute that cost more equitably by placing it on the consumers who are having the benefit of the property being depreciated. And the treatment of the annual installments of depreciation as a deduction from each year's income gives a truer picture, although still approximate, of what that year's earnings actually were. The Depreciation Reserve and the Rate Base This treatment of depreciation, however, gave rise to another and more difficult problem. The annual depreciation charges in the early years of a plant's life are likely to be much greater than the retirements. The resulting reserve in an expanding utility is likely to reach 25% or even 30% of the plant's original cost before leveling off. Obviously, for a company in need of capital for plant expansion, it is preferable to invest these funds in new plant and improvements than to keep them in liquid form, and typically this has been done. Yet if the new plant acquired with funds received from consumers (in effect, as advances against future retirements) were to go into the rate base, should the utility be allowed to earn a return on that part of its base? This problem was solved by allowing the return to be earned on the entire "used and useful" plant but requiring that the full amount of the depreciation reserve be deducted from the plant account, taken at original cost. This operation produced S National Assodation of Railway & Utilities Commissioners, Proceedings, 1938, p. 438. This action had been foreshadowed by the NARUC's adoption in 1936 of the Uniform System of Accounts for Electric Utilities whicl did not provide for retirement reserve accounting. 204 Part 2. Suggestions for Corrective Action what has come to be termed the "net investment" rate base,4 a concept employed in states adhering to the "prudent invest- ment" or "original cost" theories. This solution has never been wholly satisfactory to industry. Its spokesmen have complained that the depreciation reserve built up by the straight-line method did not measure, in any true sense, the loss of service value in the plant, one of the grounds on which the deduction of the reserve has been de- fended. Moreover, both they and the commissions encountered serious difficulties in shifting from a retirement reserve method to the straight-line method when the question of deducting reserves arose. If, as was common, the actual reserve was smaller than the amount which would have been built up by requiring use of the straight-line method from the start (the "reserve requirement"), then deduction of the actual reserve would appear to allow an inflated rate base. The deduction of the full reserve requirement caused a drastic curtailment of some rate bases. The telephone industry in particular was fear- ful that actual reserves would have to be deducted since it had voluntarily accumulated reserves well above the straight-line reserve requirement long before the deduction of any reserves was contemplated. Over a period of years, through the use of various methods, including the judicious employment of compromise and not a little litigation, the companies and the commissions in the United States have brought actual reserves quite closely into line with the reserve requirement,5 but the experience is one which Latin American regulatory authorities would do well to strive to avoid. Moreover, since World War II, the newly established principles for the treatment of depreciation have been under attack on a different ground. 4 The Federal Power Commission used the term in 1942 in computing the rate base in the Hope case. See Cleveland and Akron v. Hope Natural Gas Co., 44 Public Utilities Rep. (New Ser.) 1, 24 (1942). 5 The problem is discussed in Lippitt, "Net Investment Rate Making: The Deduction for Depredation," 62 Harvard Law Review (1949) p. 1155. Handling the Problems of Depreciation 205 The utility industry, especially in states using the net in- vestment rate base, complains that, by basing depreciation on the original cost of the plant being depreciated, the accepted method is neither measuring and allocating depreciation cost accurately nor maintaining the integrity of the company's in- vestment. The critics point to the continuing fall in the pur- chasing power of the dollar (about 50% since 1940). They argue that the sums accrued by depreciation based on original cost cannot possibly permit the company to replace plant as it has to be retired and so only the nominal dollar investment is being protected. Moreover, they insist that, since the straight- line depreciation expense understates the actual loss that the utility is experiencing through depreciation, its income is be- ing overstated, and it may actually be impairing its capital when it pays dividends on this basis.6 One of the more fre- quently advocated alternatives is termed "economic deprecia- tion;" it is proposed that annual depreciation charged be trended for changes in an index of public utility construction costs.7 Any such upward adjustment of the depreciation charge leads to another problem. Since the result of trending is to enlarge the depreciation reserve, then its deduction from the original cost rate base would curtail that base unfairly. To meet this hazard, it is sometimes proposed that the base itself be trended, a view that excites the opposition of the still-powerful cham- pions of the net investment base. Another alternative is to keep the amount of the adjustment separate from the depreciation 6 The problem in utility and other fields is discussed in A. H. Dean, "Pro- vision for Capital Exhaustion under Changing Price Levels," 65 Harvard Law Review (1952) p. 1339. 7Among numerous articles advocating "economic depreciation," see C. Ross "Responsibility of Utility Management with Respect to Depreciation," 50 Public Utilities Fortnightly (1952) p. 602; G. Jackson, "Depreciation-and Inflatibn," id., p. 617. A vigorous contrary argument was advanced by Charles W. Smith, Chief of the Bureau of Accounts, Finance and Rates of the Federal Power Com- mission, in "Public Utility Depreciation," id., p. 625. Mr. Smith argued that basing depreciation on original cost "avoids conjecture as to future prices," and "properly allows for technological improvement." 206 Part 2. Suggestions for Corrective Action reserve and not deduct it at all. But this allows the utility to earn a return on plant purchased with consumer contributions. Fortunately, we believe that an avenue of escape from these entanglements opens up if the suggestion we advanced in the preceding chapter is adopted. As will be recalled, we suggested in that chapter the desirability, first, of ascertaining the utility's investment base (the capital accounts comprising the equity) and then of applying the approved rate of return to that base, adjusted for intervening changes in the cost of living. The util- ity's capital accounts, rather than its plant accounts, would be the accounts with which the rate-making process would be con- cerned. Such being the case, there would be no need to examine the extent to which the plant might be depreciated. However, a fuller explanation of our proposal for adapting the treatment of depreciation to this procedure is necessary and will be pre- sented in the succeeding section. A Suggestion for Separating Depreciation Accounting from the Calculation and Adjustment of the Investment Base The prevalence in the countries studied of inadequate an- nual charges for depreciation expense and seriously deficient depreciation reserves points to the desirability of adopting without delay the straight-line method of determining depreci- ation expense so that, pending the determination of the in- vestment base, the further under-depreciation of existing plant could be halted. In the discussion that follows, it will be as- sumed that the utility has been permitted (or required) to use the straight-line method or one of its numerous variants.8 We shall also assume that the investment base has been deter- 8 For the purposes of our proposal, other methods of allocating the cost of plant over its service life are suitable; the fact that straight-line charges remain constant is advantageous in projecting electric rates based upon them for a period (hopefully) of years. Handling the Problems of Depreciation 207 mined as the starting point from which all further adjustments of the equity capital accounts are to be made. In fixing that base, the regulatory body may or may not have provided for a certain amount to be set aside on the right-hand side of the new balance sheet as a depreciation reserve. In any event, a reserve thus created could seldom have any validity as a meas- ure of the plant's loss-of-service value or have any logical rela- tion to the investment base. Finally we shall assume that the suggestions advanced in Chapter VIII for annual adjustment of the equity capital accounts have been put into effect. Let us suppose that the utility's long-term debt amounts to Ps. 100 million, that the investment base (evidenced, say, by common stock) is Ps. 100 million, and that the depreciation reserve set up at the same time is Ps. 20 million. During the ensuing year, the utility, after covering interest charges on its bonds, earns Ps. 12 million and pays out Ps. 7 million in divi- dends on its common. A 5% increase in the cost-of-living index has also occurred. The result of these two happenings is to add Ps. 5 million of retained earnings to Surplus and to begin the Capital Surplus Adjustment Account with Ps. 5 million from the initial adjustment for the increase in the cost-of-living index. The retained earnings are invested in new plant, in- creasing the plant account kept (at original cost) by Ps. 5 million. The Capital Surplus Adjustment Account created by this action is balanced on the asset side by a Ps. 5 million ently to a Capital Assets Adjustment Account. In the same year, Ps. 6 million have been charged as depreci- ation expense (3% on total plant of Ps. 200 million). How- ever, since much of the plant is new, we shall assume retire- ments of only Ps. I million, leaving Ps. 5 million net for investment in new plant. This would be reflected on the left- hand side of the balance sheet as an addition to Plant and on the right-hand side as Depreciation Reserve. It is at this point that familiar practice would require the addition of the new plant to the rate base and the subtraction of the Depreciation Reserve from the Plant Account in order 208 Part 2. Suggestions for Corrective Action to arrive at a net investment rate base. We submit that these operations are unnecessary. Note that the capital accounts comprising the investment base (the Common Stock and Sur- plus and the Capital Surplus Adjustment Account) represent only the adjusted equity investment and do not include any additions contributed by consumers through the operation of the 3% depreciation charge. Neither the addition to plant re- sulting from the investment of accrued depreciation nor the Depreciation Reserve itself is to be included in the amount on which the utility is to earn a return for its common sharehold- ers. Hence there is no need from the standpoint of fairness to investors and consumers to add the new plant and deduct the reserve. The question may be raised whether the failure to deduct the Depreciation Reserve from the investment base does not result in exaggerating the latter's size. Not if the utility is not shrinking and is following the practice of investing depreciation reserves in new plant. Any diminution in the value of the plant that appears to be taking place through depreciation is offset by the acquisition of new property with the proceeds of the depreciation charge. The investment in the utility remains a continuing one, and the depreciation charge is not amortizing it. Although, under the method we suggest, it is not necessary to go through the motions of adding the new plant to the rate base and then subtracting the reserve, the effect of our pro- posal is much the same, absent the various complications noted above. If retirements were to absorb more than the annual depreciation expense in a single year, this would reduce the Plant account and the Depreciation Reserve but would leave the investor's contribution in the capital accounts unaffected. If adverse economic trends made the exhaustion of the De- preciation Reserve likely, a problem of amortizing the invest- ment in the utility would arise, but here we are dealing with a set of contingencies that seem highly unlikely. Still another question may be raised. If the proceeds of the depreciation charge, though invested in plant, are not to serve Handling the Problems of Depreciation 209 as any part of the base on which a return to the utility is to be calculated, what benefit do the owners of the utility derive from the process of building up a reserve? How does it pro- tect the integrity of their investment? The main benefit to the investor is the fact that, through the depreciation charge, the users of the utility's service are con- tributing either to the replacement of worn-out plant or to plant expansion. Thereby the physical reality which underlies the investment is preserved-or would be if the charge is set at an appropriate level. However, in any inflationary economy, the investor seeks protection for the real value of the money that he has invested and the payment of a fair return upon it. If, as we have suggested, the investors receive a return on the investment base adjusted for changes in purchasing power, that will maintain the economic integrity of their investment if the physical integrity of the plant can also be maintained. If, as we suggest in the next section, it is found desirable for this purpose to augment the charge for depreciation ex- pense so as to reflect changes in the cost of living or in the index of construction costs, that may readily be done. Under the plan we suggest the effect of such action would not be to enlarge the base on which the equity is entitled to earn its return. The resulting higher annual charge for depreciation naturally will mean greater additions to plant than would otherwise occur, but the increase in the Plant Account is bal- anced by increases in Depreciation Reserve, not in the Com- mon Stock or the Surplus Account. If it seems wise to adjust the Depreciation Reserve by either of these indexes in order to keep in line with changing costs (and with the adjusted capital accounts), this can be done without affecting the investment base. If Depreciation Reserve itself is adjusted, then a balanc- ing account must be created on the asset side of the balance sheet but that will be equally without effect on the investment base. One of the advantages resulting from the suggested separa- tion of the two categories of accounts is that it permits the 210 Part 2. Suggestions for Corrective Action depreciation charge and reserve to be adjusted by a different index from that used in adjusting the capital accounts. This may become important if changes in construction and equip- ment costs should differ materially from changes in living costs. The separation also avoids complications with respect to the effect of retirements of adjusted plant on the rate base. Adjusting Depreciation Expense and Reserve for Price Changes The separation of the investment base and depreciation ac- counts that we have suggested leaves open the question whether to reflect in these accounts such changes as may take place in the price level in view of their probable effects upon replace- ment cost. If the changes in the price level appear to be long lasting, then the case for adjustment of depreciation expense is a strong one. In countries where inflationary pressures are strong and per- sistent, it is obvious that depreciation expense calculated (by whatever method) on the original cost of the plant being de- preciated will not yield sufficient funds to enable the enter- prise to replace the depreciated plant with new plant capable of rendering equal service. The deficit may be offset in part by technological gains, and at no time can the depreciation charge be set with foreknowledge of the cost of replacement. These considerations (together with the convenience of usinig a cost base) have led the case for "economic depreciation" and kindred proposals to be rejected by the regulatory com- missions in the United States.9 But in this country, the loss in purchasing power of money has been much more gradual than 9 A recent decision running counter to this trend is Re Indiana Teleplhone Corp., Indiana Pub. Serv. Comm'n, 16 Public Utilities Reports (3d Ser.) 490 (1957). See also Iowa-Illinois Gas & Electric Co. v. City of Fort Dodge, 20 id., 159, 190, 85 N.W. 2d 28, 53 (Iowa Sup. Ct. 1957). Handling the Problems of Depreciation 211 in many Latin American countries, and technological gains have been more rapid. Moreover, the electric utility com- panies and their regulators in the United States are not faced by a lack of demand for electric securities and do not have to solve the problem of how to establish investor confidence in themn. Such a predicament might well lead them to be more receptive toward proposals for adjusting depreciation charges.10 Still another consideration argues for adjusting depreciation charges for price changes: the application of the cost-of-service principle, especially as between present and future consumers. If no adjustment is made, are not present consumers paying less than their share of the cost of service when they leave to future users the burden of replacing at inflated prices the plant which is now in service? This consideration seems to underlie the views of the 1956 Parliamentary Committee of Inquiry into the Electricity Sup- ply Industry in Great Britain. In discussing the provision for depreciation, the Committee stated:" In the case of a state-owned monopoly .. . prices must be determined by a policy of calculated assessment. It seems to us that a proper element in arriving at these prices is the cost of the resources used up in the process of manufacturing and delivering the product-in this case electric power. It also seems to us that this cost should reflect current price levels rather than those ruling at the time when the plant wvas originally installed.... We have therefore thought it our duty to ascertain whether the industry is charging its consumers prices which . . . ex- 10 The champion of the Federal Power Commission's policy of basing de- preciation on cost, Charles W. Smith, in the article cited supra note 7, said, at p. 627: "If it is necessary to allow depreciation on a value basis in order to pre- serve the soundness of the utility industry, then I would be for it." 11 Report of the Committee of Inquiry into the Electric Supply 7ndustry (Cmd. 9672, H. M. Stationery Office, London 1956) p. 87. In discussing factors making for higher electricity costs, the OEEC experts noted that "cost price will more and more be based on plant replacement cost." OEEC, The Trend of the Selling Price of Electricity in its Relation to the Financ- ing of New Plant (Paris, Nov. 1958) p. 15. 212 Part 2. Suggestions for Corrective Action ceed or fall below those which would cover the real value of the resources used up in serving them. The Committee found that the industry had been making an allocation towards the difference between historical cost and replacement cost depreciation as an appropriation of surplus. The Committee criticized this accounting practice as mislead- ing, since no surplus could properly exist, in its view, until depreciation had been taken at current cost levels. The position of a regulated, though privately-owned, electric monopoly is essentially the same as that of a state-owned monopoly for the purposes of the point which the Committee was developing. We believe that, under Latin American condi- tions, even more than in Britain, there is need to provide for the adjustment of annual depreciation expense.12 12 If our suggestion that capital accounts be adjusted for changes in thc price level were to be adopted but not our suggestion for separating depreciation accounting from the determination of the investment base, the need for adjusting depreciation accounts would be imperative. If the revaluation of plant on the basis of reproduction cost is used in determining the rate base under the rate- base x rate-of-return formula, then adjustment of depreciation expense and re- serve would be called for unless "observed depreciation" were used to determine depreciated reproduction cost. Even in that event, adjustment of depreciation ex- pense would be desirable. CHAPTER XI Suggestions for Obtaining New Capital for Expansion In dealing with the problem of how electric enterprises in Latin America may obtain new capital from domestic sources to fi- nance the expansion they so sorely need, we may appear to be extending our inquiry beyond the subject of the regulation of the electric industry. Ought we not to consider only those meas- ures that would give investors in the electric industry reason- able assurance that its earnings would suffice to provide stable returns in terms of purchasing power, regardless of the extent of price fluctuations in their national economies? We do not believe we should thus limit our suggestions. It is not enough that the industry be assured adequate earning power; securities must exist with such terms as will enable the Latin American investor to benefit directly from the earnings thus obtained. Such securities have not yet come into general use. In recent years very few security issues have been offered to the public by electric power enterprises, not only in the five Latin American countries we have studied, but in the entire region. Domestic funds required for expansion have come largely from retained earnings, depreciation accruals, and loans from the special development banks or from private banks. In the cases of municipal enterprises such as those in Medellin and 213 214 Part 2. Suggestions for Corrective Action BogotA, their municipal character has precluded sales of stock to the public. Mixed economy corporations, such as the state enterprises in Brazil, have had little public acceptance of their stock. State-owned companies, such as Endesa in Chile, have found that the public is unwilling to accept the limited return which they offer. Locally owned private enterprises have had some little success in selling securities, sometimes with the im- plied, at least, promise that this would result in power connec- tions for the buyers. The most notable achievement in the sale of securities locally has been the sale of over Cr. 500 million par value of common stock by the subsidiaries of the American & Foreign Power Company in Brazil. For example, the Paulista Company sold stock on the basis of an annual 10% cash dividend and a peri- odic 5% stock dividend. After the sale of the shares, the Brazil- ian public owned 48.5%. Since issuance, cash dividends have been paid regularly, and stock dividends have been declared about every other year. The par value of the stock of the Paul- ista Company is Cr. 200, and its market price has tended to settle at or just below par. This performance is disappointing, however, when compared with the records of any representative group of Brazilian common stocks of companies in unregulated industries. Given the long history of limited participation by domestic investors in the financing of their own electric utilities, we be- lieve that affirmative policies and measures are needed to break through this long-standing reluctance. This would require ini- tiative on the part of the regulatory agencies having jurisdiction over the electric industry. Regulatory agencies have authority in most countries over the form taken by the securities issued by electric enterprises, and this seems to us to offer hope of constructive action. Participation in the ownership of electric plants by other industries represents another avenue for ex- ploration which clearly would require regulatory approval. Moreover, electric rates themselves may be used as a means of accumulating capital under safeguards that the regulatory au- Suggestions for Obtaining New Capital 215 thorities could best impose. Certain of our suggestions, to be sure, involve tax policy and the policies of institutional invest- ors, both public and private, but a failure to take these aspects of the problem into account would leave our suggestions in- complete. We have not allowed a narrow conception of our function to debar us from their consideration. There are serious impediments other than restrictions on earnings that affect the sale of these securities in many of the countries. To utilize potential capital markets, we suggest that foreign-owned operating companies should be reorganized as domestic companies where this is necessary to remove restric- tions on the sale of their securities based on their foreign char- acter. Of course, any reorganization for this purpose must be freed of tax consequences by legislation or administrative rul- ings. If appropriately designed securities of electric power com- panies are made available, they should be qualified as invest- ments eligible for banks, insurance companies, pension funds, and investment companies. Before turning to this aspect of the problem, we shall suggest certain types of provisions that should protect their holders against the impact of inflation. In so doing, we assume that enough of the measures suggested in the preceding chapters have been adopted, at least in substance, to make the suggestions which follow practicable from an earn- ings standpoint. 1. Securities to Protect the Investor Against Inflation Everyone seems to agree that it would be highly desirable to achieve greater domestic ownership of the securities of the foreign-owned companies through the sale of securities having the right to participate in earnings. We believe that not only the foreign-owned companies but all Latin American electric enterprises, public and private, could obtain a substantial por- 216 Part 2. Suggestions for Corrective Action tion of the domestic funds needed for expansion through issu- ance of senior securities (i.e., bonds, debentures, preferred stock) with features for adjusting the monetary return in the ways described below. The basic interest and preferred divi- dend rates payable on the suggested securities would be lower than the rates obtainable for fixed interest securities-assum- ing, indeed, that the latter could be sold. In case of inflation, the burden of increased interest and dividend payments would be borne by the consumers who would be benefiting by the enterprise's ability to finance its expansion. The measure of the increase in the principal due on the debt or the share of the preferred on liquidation is roughly the measure of the higher value of the plant purchased with the funds obtained by such issues and, under the suggestions we have made, these increases would be reflected in substance by the adjustment of the enter- prise's capital accounts. A. COST-OF-LIVING AND COST-OF-ELECTRICITY SECURITIES One of the most direct methods of increasing the attractive- ness of senior securities that ordinarily carry fixed rates of re- turn is to provide for the escalation of those rates by reference to an index such as the cost-of-living index. As the cost of living increased, so would the interest or preferred dividend payable. The escalation could be confined to the return, but where the inflation menace is serious, probably the escalation of the prin- cipal would also be necessary in order to attract investors. Though the amounts payable would vary up or down with changes in the index used, they would not be allowed to fall below the face value of the securities and their specified mini- mum rates of return. Other indexes could be used although the advantages of the cost-of-living index for measuring adjustments in the capital accounts of electric enterprises have already been noted.' 1 See Chapter VIII, supra, p. 171. Suggestions for Obtaining New Capital 217 Against the fear that escalation of interest and preferred divi- dend rates would add to inflationary pressures in the economy, we submit that the minor relaxation of a regulated rate of re- turn should not cause alarm, especially in any economy where most prices and profit rates are allowed to move freely upwards. Moreover, the fact that the cost-of-living index is used for ad- justing the investment base for equity securities makes it peculiarly apt as a measure of adjustments in the return on senior securities. In France, resort to indexed securities was authorized in 1952 and advantage was promptly taken of the opportunity, especially by the nationalized industries. France did not permit the cost-of-living index to be used as a measure, an understand- able precaution in view of the fact that indexed securities were not to be confined to issuers in regulated industries. The result of this limitation on indexed bonds (described more fully in the Appendix, reporting briefly the French experience with them) was to induce the use of a variety of ingenious schemes of escalation. One of the most relevant of these to the problems of Latin American electric enterprises was the cost-of-electricity securities issued by Electricit6 de France, the nationally-owned electricity undertaking in France. Between 1952 and 1954 Electricite de France issued its "Pro- duction Securities," amounting to fr. 18,000 million for the year 1952, and fr. 25,000 million for the year 1953. This secu- rity, first offered in 1952, was offered on a pre-emptive basis in the following years. The essential terms are: (1) The annual income is to be equal to the average sale price of 200 kwh of electricity, but may not be less than fr. 1,400. (2) The securities are divided into ten series, each series to be repurchased annually by Electricite de France, the first purchase being four years after the issuance. (3) The purchase price of a series by Electricite de France shall be equal to twenty dividends but not less than fr. 32,000, the issuance price. 218 Part 2. Suggestions for C orrective Action What Electricite de France achieved through this cost-of- electricity security was an issuance at the rate of 4.5%, subject to the revaluation of that rate and the principal amount due in the event that increases in the price level brought about an increase in the price of electricity. Under these circumstances, the investor who desires the assurance of a fixed income secu- rity is not forced to speculate on the trend of prices. He is as- sured the fixed income, and at the same time has some assur- ance that during the life of the issue it will have substantially the same purchasing power as at the time of issuance. Of course, the investor gains this assurance only under a system of electric rate determination that advances the price of electricity concurrently with increases in electricity costs. The suggestions we have made would, if adopted and faith- fully adhered to, achieve that result satisfactorily. An advantage of the plan lies in the fact that the tie between the interest payable on the security and the rates charged by the electric supplier gives assurance of the latter's ability to meet the es- calated interest charge. At the same time, however, it makes the fidelity of the regulatory authority to the terms of the elec- tric rate adjustment plan essential to the success of the security issue. It should be added that the provisions for repurchases are not an essential element of the plan. B. SECURITIES PAYABLE IN DOLLARS OR WITH OBLIGATIONS MEASURED BY THE EXTERNAL VALUE OF THE DOMESTIC CURRENCY Another method of giving flexibility to the return on debt securities is exemplified by debt securities payable in dollars. These have been issued recently in Mexico and have been well received. Such an obligation, because of its greater protection against currency depreciation, bears a lower interest rate than those payable in pesos. Two relatively recent issues of bonds payable in dollars are Suggestions for Obtaining New Capital 219 the U. S. $4,100,000 issue of Celulosa de Chihuahua, S.A., and the U. S. $7,000,000 issue of Compajifa Constructora Nacional de Carros de Ferrocarril, S.A. The former issue carries a 2.5% semi-annual coupon; the latter, a 4.5% annual coupon. The $4,100,000 issue of Celulosa Chihuahua was put on sale on January 25, 1955, and was completely subscribed for at the close of business on January 27.2 The $7,000,000 issue of Carros de Ferrocarril was subscribed for by Nacional Financiera and three investment companies. Of the total issue, $2,000,000 was offered to the public. Issuance of dollar bonds locally may impose the requirement of obtaining the necessary foreign exchange to meet the inter- est payments as they become due. The same objective of a more stable security may be achieved (except where multiple ex- change rates prevail) through the issuance of securities pay- able in the domestic currency with interest and principal keyed to the external value of the domestic currency. Such a security vould assure prospective investors that adverse exchange fluc- tuations would not affect their investment. Further, since in- flation and devaluation generally arrive at about the same point in terms of decreased purchasing power, such securities wvould constitute a desirable hedge against inflation. To enable the issuing enterprise to meet the obligation in case of an impor- tant change in the exchange rate, it would, of course, be neces- sary to provide for the escalation of rates upon the increase in charges due to exchange rate depreciation. C. CONVERTIBLE DEBENTURES No more than brief mention need be made of this financing device which has only recently come to be experimented with in Latin America and indeed, in France, though it has long been familiar in the United States. The power to convert a debt security into common stock provides a means of protect- 2 El Mercado de Valores, January 31, 1955, p. 52. 220 Part 2. Suggestions for Corrective Action ing the purchaser of the former from the effect of inflation on fixed obligations. If a debenture is to be convertible into a utility's common stock, the privilege will be attractive only if the regulatory system permits the common stock to earn a suf- ficiently high return so as to offset the effects of inflation upon it. A convertible issue could be rendered more attractive by providing for the indexing of the interest payments. D. PARTICIPATING PREFERRED STOCK Not all financing can or should be carried out by debt secu- rities, but one of the problems encountered by both govern- ment-owned electric utilities and those owned as subsidi- aries by holding companies is that of raising capital through the sale of stock without losing voting control. Government- owned enterprises which have taken corporate form have tried to meet this problem through the issuance of preferred stock, but the rates, being both fixed and low, have effectively discouraged public investment. For any holding company, sale of the common stock of the underlying companies is limited by the desire to retain effective control. Outside the electric utility field, a similar problem has been posed by closely-held industrial corporations that require large sums of new capital for expansion. They have found an answer, in part, in the is- suance of participating preferred stock. Participating preferred stocks have already received wide public acceptance in Brazil. Thus, in the year 1954, out of fifteen underwritten security issuances, aggregating Cr. 843 mil- lion, ten issues, aggregating Cr. 561 million, were participating preferred stocks with preferred dividend rates of 10% and 12%. A participating preferred stock gives the same priority on earnings and assets as a straight preferred stock. It may be cumulative with respect to the fixed amount of the dividend and non-cumulative as to the remainder, or it may be com- pletely non-cumulative. A preferred stock may be designed to Suggestions for Obtaining New Capital 221 have more of the characteristics of a common stock than the usual preferred issues. It may participate in all dividends after the common shares have received a dividend equal to the preference, including stock dividends. It may have a priority over the common on dissolution and participate with the com- mon in the distribution of the surplus. It may be given full voting powers or the power to vote as a class on matters affect- ing its rights. The attractiveness of participating preferreds rests in the assurance of a priority in dividends in a limited amount and the ability to share in the success of the corpora- tion if it prospers-or if it does no more than keep its profits abreast of a rising price level. The summary in the stock offering of a non-utility company in Brazil states: The first offering of participating preferred stock was made in January, 1953 at a price of 115% of the nominal value. In March, 1954 these shares were selling at 190% of their nominal value. Further, the cash and stock dividends paid amounted to 32% of the nominal value, or 28% of the in- itial offering prices. One would not expect such a return by a regulated electric utility except in a period of sharp inflation after provision for the adjustment of capital accounts had been made. However, it may be noted that this is the very kind of competition for investment funds that electric enterprises face without ade- quate provision for such adjustments. If, in an economy with a depreciating currency, electric utility investors are restricted to fixed preferred dividend rates of 8% or even somewhat higher on preferred shares, they will certainly seek higher re- turns afforded by non-regulated industry. What the electric utility industry can offer, however, is a high degree of assurance that the regular dividend will be paid, should conditions become adverse. Participation with the equity or an indexed dividend is essential if preferred stock is to have a market not only under inflationary conditions but 222 Part 2. Suggestions for Corrective Action in countries where the risk of inflation remains real, even in times of relative price stability. E. SECURITIES ISSUED BY MUNICIPAL ENTERPRISES The problem of financing the electric enterprise owned by a municipality or by any other political subdivision is made difficult by an all-too-common tendency for the owner to ex- ploit the enterprise to further its other functions. To avoid this, where, say, municipalities aye engaged in the service of electricity supply, the electric service should be completely separate from the city's municipal functions and its other pro- prietary functions. The city should not look upon its electric department as a source of funds to shore up sagging municipal revenues. Rather, it should look upon itself as owner of the equity, to whom dividends are to be paid, if earned and not required for the expansion of the existing electric plant. We have already noted the resemblance of a government- owned electric utility to an operating subsidiary in a holding company system. The holding company or parent governmental body is the owner of the equity, and the operating company obtains funds from the public through the sale of its senior securities. Conceivably the credit rating of a well-run munic- ipally-owvned electric system might be better than the credit of the city itself. If it is to be autonomous in fact as in theory, the electric enterprise should depend on its own credit in order to finance expansion. Accordingly, the electric utility assets and accounts should be segregated from other municipal ac- counts; the electric utility could then issue its own mortgage bonds, secured by its physical assets. Government-owned enterprises are prevented from raising equity capital since, of course, stock ownership by the public would change their status as from governmental to "mixed" corporations. Accordingly, apart from funds raised by bonds and debentures, these enterprises must rely for domestic funds on bank loans, advances from the parent governmental body, Suggestions for Obtaining New Capital 223 and retained earnings and depreciation accruals. Yet clearly these enterprises should obtain domestic funds through public participation. For this purpose, given public attitudes toward fixed-return securities, the 'indexing of bonds and debentures is called for. However, a further attraction could be provided the investor by issuing the securities as earnings participating debentures. Debentures of this type would carry a fixed rate of interest and would be of a fixed maturity. Additionally, they would be entitled to share in the earnings of the enterprise, after the earnings should achieve a stated percentage of the equity in- vestment. In effect, this becomes a combination of a fixed inter- est and an income debenture. Even if it were not possible to issue the debenture with an indexed return, the ability to par- ticipate in earnings would give' some degree of protection against inflation and so remove the chief undesirable feature of fixed income securities in a regulated enterprise. Of course, the municipality would have to make adequate provision for adjustment in its electric rates to give reasonable assurance that a rise in prices would be accompanied by a rise in rates and so in earnings. 2. Potentialities for Institutional Investment in Electric Utilities In channeling substantial amounts of private savings in Latin America into electric utilities-whether government- or privately-owned-the role of institutional investors, such as insurance companies, pension funds, and investment compa- nies, may be important. In the United States, life insurance companies constitute the greatest single market for bonds of electric utility companies. Thus, between 1949 and 1953, life insurance companies increased their holdings of utility bonds by U. S. $3,063 million,3 accounting for approximately 25% of the new financing by utilities in that period. 3 Institute of Life Insurance, Life Insurance Fact Book, 1955, p. 69. 224 Part 2. Suggestions for Corrective Action While life insurance funds in the Latin American countries are much less important than in the United States, yet they do constitute a significant potential source of funds. In the United States, the ratio of life insurance savings to gross national prod- uct ranged from 1.25% to 1.5% in the period 1948-1952.4 For the Latin American countries, such indicated ratios would be .25%-.3% for Brazil, Colombia, and Costa Rica, and .15% for Mexico. An objective for the electric utilities might be to obtain 10% of the annual funds of insurance companies available for investment. To take Mexico as an example, this 10% might amount to no more than Ps. 10 million or about 5% of the funds appropriated by the Federal Government for CFE. As the desirability of such investments increases, however, elec- tric utility securities might well achieve a place of real impor- tance in the portfolios of insurance companies. They should offer a higher return than government obligations and be sub- ject to less risk than industrials. This market, moreover, would grow with the growth of the insurance industry. A more important source of funds than life insurance in the Latin American countries are the pension funds accumulated through the operation of their social security systems. Such funds are more extensive in Latin America than in the United States, and, to some degree, are the counterparts of private in- surance and private savings in the United States. The annual net accrual for social security funds amounted to 1.9% of total national income for Brazil (1947) ; 2.3% for Chile (1952); and .5% for Mexico (1953). A large part of the social security funds is now devoted to financing social betterments such as public low-cost housing, hospitals, etc. It is not our purpose to question the propriety of these investments. However, the actual existence, or even the risk, of serious inflation suggests the desirability that the portfolios of the social security funds also include securities with income protection against serious monetary changes. Of 4 Computed from National Income (1954 ed.) pp. 165, 167. Suggestions for Obtaining New Capital 225 course, fixed income utility bonds in the economic context of a country such as Chile would not attain this objective. If, however, electric utility bonds can be issued which have some satisfactory protective device against inflation, investment in such bonds not only would make more secure the future pen- sions of employees in the country but also would aid in achiev- ing similar social objectives such as investment in public housing, etc. At present, the restrictions on investments by institutional investors in a number of Latin American countries would pre- vent them from investing in the securities of electric enter- prises. If these securities can be made more attractive as to income without endangering their dependability, the restric- tions on their eligibility for institutional investment should soon be relaxed. The possibilities of using social security funds to aid in financing electric power expansion has been recog- nized by International Monetary Fund officials who described them as follows:5 Satisfactory means can be found for investing at least part of the social security reserves in enterprises of social and economic significance.... Such trust funds, however, should not be invested haphazardly without adequate supervision. The proper safeguard would be to establish a list of enter- prises whose preference securities would be eligible for in- vestment because of their record of continuity and stability in earnings. The public utility enterprises-electric power, telephone communications, etc.-are likely to meet these tests to a high degree. With rates reasonably adjusted to the cost and value of the services they provide, public utility enterprises should have no difficulty in providing a good return on the social security reserves invested in them. An instance of action making utility securities eligible for investment by pension funds is afforded by the recent experi- ence of the Cuban Electric Company. That company obtained a loan of Ps. 8,000,000 from Financiera Nacional de Cuba, an 5 E. M. Bernstein and I. G. Patel, "Inflation in Relation to Economic Develop- ment," International Monetary Fund, 2 Staff Papers (1951-52) p. 393. 226 Part 2. Suggestions for Corrective Action autonomous government agency, secured by first mortgage bonds of the Company. Financiera sold a like principal amount of its own bonds, which are legal investments for banks, insur- ance companies and pension funds, to a number of domestic institutional investors. As the strength of electric securities grows more apparent, the need for the interposition of a gov- ernment institution like Financiera should dwindle. The Cuban experience shows that domestic private savings may be channeled into electric securities. However, in coun- tries without Cuba's exchange stability and exposed to more serious inflation, a fixed interest security would not have the same appeal. Hence there would be need for some types of se- curity that have been considered earlier in this chapter, prin- cipally, in the case of institutional investors, debt securities. 3. Greater Participation by Industry in Electric Utility Financing The high rate of industrial expansion has been a major fac- tor in electric utility expansion in the developing countries. Not only has this expansion caused an increased demand for electricity for industrial use, but it has increased the demands of residential and commercial users in established urban cen- ters. Existing public service supplies of electricity have fre- quently been insufficient in amount or reliability to satisfy the needs of new or expanding industry. This has led many industrial concerns to install their own diesel generating facili- ties in order to have an assured source of power. These relieve the electric utility of only a part of the burden created by in- dustrial expansion since the problem remains of supplying the increased residential and commercial load it has caused. In most of the countries studied, industrial profits are high, and industrial enterprises have had little difficulty in attracting funds for expansion. Conversely, electric utility profits are low, and capital is not available. The profit-starved electric utility industry helps to swell further the profits of handsomely re- Suggestions for Obtaining New Capital 227 warded manufacturing industry by reason of rates that do not cover the cost of service. At the same time, industries that can- not exist without electric power provide virtually none of the funds required for expansion of electric utilities but reinvest a large portion of their profits in their own enterprises, thus further burdening the public service supplier. While increased industrialization is of undoubted benefit to any country, it also imposes a very heavy capital burden for the electric expansion it makes necessary. For an industrial firm to expand, lacking electric power from a public service supply source, it must itself bear the capital costs of installing diesel equipment. Such equipment is not only costly to install; its operating costs are much higher than the rates at which pub- lic utility electric companies can supply industrial consumers, even though those rates are set at adequate, compensatory levels. However, a manufacturing company using a public source of electric power may absorb a substantial percentage of the available capacity of the electric plant and hasten the time when further generating capacity will be essential to meet community needs. If, then, industry is to profit by the avail- ability of a public service electric power supply, it seems no more than fair that it contribute toward the capital cost of the expansion in generating capacity which meeting its power needs will sooner or later require. Once it is recognized that industrial users should bear a greater part of the financial responsibility for electric power expansion, the question of method arises. A method is needed which would not penalize power-consuming industries yet at the same time would draw on industry resources to a degree commensurate with their responsibility for increased demand. The first step toward a solution, obviously, is to increase the rates of industrial users. These, as a rule, are substantially below the low level of residential and commercial rates re- ported in Chapter IV. There is, of course, a limit on the rates that may be charged industrial users; when public service rates reach the point where it is cheaper for an industry to supply 228 Part 2. Suggestions for Corrective Action its own power, it will install its own generators. However, rate adjustments to assure a greater industrial contribution to the expansion of electric capacity need not approach this limit; to cast a heavy burden on the industrial user would, in the long run, be self-defeating. As new electric capacity is brought in, sustained industrial demand may be vital to its economic oper- ation. Since, as was noted in Chapter IV, the cost of electricity in industry is no more than from .5% to 1.5% of the cost of pro- duction, with few exceptions, chiefly the extractive, chemical and metallurgical industries, an increase in electric rates of 10%, 20%, or even 50%, should not seriously affect the prices of most products. Indeed, most industries could readily absorb the increased cost as part of their operating expenses. Suppose, moreover, that the Latin American electric power industry could improve its own operating performance, now gravely deficient as a result of overloading in many areas, and could steadily build up its existing capacity and take full advantage of technological progress in electric generation and distribu- tion. The result would be to achieve real savings for industrial consumers. These could greatly outweigh the increased charges for electricity needed at the outset. Instead of an attempt to meet the problem entirely by rais- ing rates to industrial users, a more direct approach might prove practicable. The subscription to a specified amount of the securities of the electric enterprise might be made a condi- tion to the installation of new or increased service in excess of a specified minimum demand. The amount of the subscription would, of course, be graduated with the demand required. In- creases in demand above the minimum would carry with them more than proportionate increases in the subscription require- ment. Despite the much higher operating costs of a diesel plant, the total subscription would have to be kept well below the capital cost entailed in installing such a plant with a capacity equal to the required demand in situations in which the exist- ing service provided by the utility was unsatisfactory and its Suggestions for Obtaining New Capital 229 correction by the completion of new facilities was still some time in the future. A subscription should not carry with it any greater assurance of firm supply than could be given to the existing customers of the enterprise. The securities to be made available for this purpose might be a portion of an issue already outstanding. They might repre- sent debt, preferred stock, or common stock, although in the case of government-owned enterprises only debt securities would seem satisfactory, with whatever provision for indexing the obligation, that the issuer had adopted. The terms of the subscription would have to have the approval of the regulatory agency to assure the fairness of the price. However, particu- larly in the case of a privately-owned enterprise, a more satis- factory method of handling these required subscriptions might well be to provide for the issuance of a special security designed for this very purpose. This might be a class of common stock, having the right to parity of treatment in dividends and upon liquidation with the common stock of the enterprise. The shares thus issued probably should be subject to a limitation on transfer for a period of, say, three or even five years. Other- wise, since their acquisition would not have been wholly volun- tary, there might be a danger, especially at the start of the plan, that the shares would be dumped at prices that would be detrimental to the market for the issuing electric company's other securities. Of course, the shares would be transferable with the purchase of the business and could be made subject to the liens of any general mortgages placed upon its property. To justify the imposition of such a requirement of invest- ment, the electric enterprise would have to earmark the pro- ceeds of these subscriptions and use them only to defray the costs of its expansion. Some care would have to be taken to define "expansion" for this purpose. The subscribing com- panies would have to be assured of a genuine quid pro quo for their contributions while at the same time not having their funds frozen in the hands of the electric enterprise for long periods. In this connection, a problem would be presented by 230 Part 2. Suggestions for Corrective Action the electric company which, instead of expanding its generating capacity, was purchasing additional power from large new generating plants. These frequently are under government ownership and sell only at wholesale. In such a situation, the regulatory agency might order an equitable division of each industrial firm's subscription between the securities of the dis- tributing company and those of the generating enterprise. Perhaps such a division would be called for only where the large scale of the subscribing industrial firm's demand would entail a large subscription. If measures such as we have suggested are taken to give to electric utility securities the kind of stability in practice that they have in theory, a requirement that industry invest capital in electric enterprises would create no hardships. True, the industrial entrepreneur would not receive as handsome a profit on his electric investment as that received in his own business, but, given a suitable regulatory regime, such an investment could be a stable, income-producing item. Measured against the cost of installing his own equipment, the investment in electric power securities would be of actual benefit, since it would re- sult in lower power costs, less money tied up in electric facili- ties, and none of the burden of operating electric facilities. The status of the entrepreneur as a shareholder-or credi- tor-of the electric enterprise, even though involuntary, would give him an interest in the welfare of the electric enterprise. This would strengthen the latter's position in the community served and, together with all other such enterprises, in the national economy. To require an initial subscription to the securities of an electric utility by industrial firms making a substantial demand on the utility's capacity represents a relatively painless way of raising funds for electrical expansion. Might not the same idea be extended and placed on an annual basis? In this way, con- tributions might be obtained not only from industrial firms making new demands for electrical capacity but also from firms already established. The annual requirement would, of course, Suggestions for Obtaining New Capital 231 be in the nature of a rate increase and the fact that other in- creases in industrial electric rates had also been made would have to be taken into account in determining its size. However, the burden of an annual subscription would be much less than that of a rate increase of equal size since it would confer on the subscribing industrial firm an income-producing, mortgageable asset which would in time become saleable. The problems and possibilities of this plan are such, however, that they can best be considered in connection with a still further extension of the idea: a construction surcharge to be imposed on all cate- gories of substantial users. First, however, consideration will be given to another possible approach to industrial participation in the financing of electric expansion: the joint ownership of electric facilities by an industrial user and an electric enter- prise. JOINT OWNERSHIP OF ELECTRIC PLANTS Situations may arise in which the needs of one or a group of manufacturers are so large that they cannot readily be met by the normal expansion of the electric enterprise that would otherwise serve them, even though this were aided by subscrip- tions to its securities by the industrial firms along the lines suggested above. Yet, obviously, it would be to the interest of the electric utility to have the capacity used by the manufac- turing plants available to serve other loads in the hours when they did not require it. Further, it would benefit the country as a whole if the same plant capacity could supply a greater demand through diversity of use. We believe these public ob- jectives may be achieved and savings effected by the manufac- turers through joint ownership of installations. One reason for installation of private plants is the fact that an industrial user cannot be assured of firm power by the pub- lic service enterprise. An agreement for joint ownership can, however, provide for firm power. This may be accomplished by the industrial user and the electric enterprise actually owning 232 Part 2. Suggestions for Corrective Action undivided interests in the plant or in a corporation formed to own and operate it. An example of such an agreement which was used with notable success in the United States is afforded by the case of Deepwater Light & Power Company. During 1926 and 1927, the American Gas & Electric Co. which operated in the area of southern New Jersey through its then subsidiary Atlantic City Electric Company, sought to in- crease plant facilities in that area. At the same time a group of companies, later designated as Delaware Power & Light Co., also sought additional sources of power. The E. I. DuPont de Nemours Co., which had a power plant in southern New Jer- sey, was considering modernization of its plant. Because their needs for power could best be supplied by a large plant and because load diversity would result in lower power costs to all concerned, the two power companies agreed to build the power plant and devote a turbine and certain other equipment to DuPont. The fixed charges applicable to the portion of the plant devoted exclusively to DuPont were to be borne by that company. Title to the property was vested in the two power companies, the plant being arranged in such a way that there were in reality two separate power plants. Operation of the company was entrusted to a custodian company organized for the purpose, "The Deepwater Operating Company." This company owned no property but was responsible for operation and maintenance of the plant. DuPont contracted for the pur- chase of steam from the plant for a term of ten years and was committed to bear its proportionate share of the carrying charges covering return on investment, depreciation, property taxes and insurance. The plan assured DuPont of a firm supply based on the output of its turbine and entitled it to purchase additional power at the average station cost. The two utility companies had prior rights for the use of the installed electric capacity except for the DuPont turbine. The result of this agreement has been described by Professor Ruggles as follows:6 6 C. 0. Ruggles, Problems in Public Utility Economics and Management (New York 1933) p. 106. Suggestions for Obtaining New Capital 233 Both utility companies and the DuPont Company profited from this unique cooperative arrangement. All three parties had been able to attain in the Deepwater Plant a degree of reliability and flexibility which would have been impossible had each of them erected a plant to serve its own territory or needs exclusively. A much greater investment in standby equipment would probably have been necessary if such a course had been adopted. The cooperation of the three parties made possible a more efficient plant than could have been economically constructed and operated by any one of them. In certain of the industrial areas in Latin American coun- tries, a group of industrial companies might contract with the electric enterprise serving the area for the construction of a generating plant on terms entitling the industrial owners to firm capacity to the extent of their share in the plant's owner- ship. Such a plant could either replace, or obviate the building of, a number of small diesel plants at lower installation and carrying costs to the industrial users and at a lower cost of oper- ation. Its fuel costs would be lower, and the diversity factor would lower the average cost to each user. The plant's output would be available to the electric enterprise's network, to the extent of its own right to firm capacity and to the extent that other users were not fully utilizing their respective shares of the plant's capacity. At the same time it would relieve the par- ticipating industrial companies from the need to operate a private electric power plant. In terms of cost, manpower, avail- ability of power, and conservation of fuel, this type of arrange- ment would appear to offer many advantages. Title to a physically separate part of the plant might be vested in the industrial participants, the rest being owned by the electric enterprise, or both the industrial companies and the utility might own undivided interests. In either case, the fact that the plant had been brought into being through the plan of co-ownership with the industrial companies would pro- vide a reasonable basis for enabling them to share in the power 234 Part 2. Suggestions for Corrective Action generated there without creating an unfair discrimination against the utility's other customers. Although responding to a different problem and using a solution that could scarcely be applied to an ordinary power installation, plans developed by the United States Atomic Energy Commission for cooperation with both the privately- owned electric industry and with public power bodies testify to the possibilities that can be opened up by innovation in ar- rangements for sharing the capital costs of power development. After competitive proposals had been received, the Commission arranged with the Duquesne Power Co. for the construction at Shippingport, Pennsylvania, of an atomic reactor to produce steam to operate turbines and electric generators. The United States is to remain the owner of the reactor, which has since been completed. It sells the steam to the Duquesne Co. which uses the 60,000 kw capacity to meet its system's needs. Duquesne will hold title to the turbines and generators and operate the electric plant. The same plan is being employed by the Com- mission in arrangements with Consumers Public Power District of Nebraska for a 100,000 kw plant, with the Rural Cooperative Power Assn. for a 22,000 kw plant at Elk River, Minnesota, and with the City of Piqua, Ohio, for a 12,500 kw plant.7 LEASE-BACK OF GENERATING STATION Another method may be used to effect savings in power costs and make power more generally available while at the same time obtaining funds from industry for the purpose. A power plant may be constructed by a group of industrial concerns and then leased to the electric enterprise serving them, a term of the lease being the firm supply of power to the lessors. An op- tion to purchase the plant, subject to the same condition, might 7 For a description of these plans and various plans for cooperation by groups of electric utility and industrial firms, see Cavers, "Atomic Power: The Quest for a Program," 27 Geo. Washington Law Review (1959) pp. 427, 435-442. Suggestions for Obtaining New Capital 235 be given the lessee. This method would be advantageous where the latter could not contribute a significant share of the capital for construction. Here again, the cost of firm power to the industrial users would be lower than the cost of separate instal- lation of diesel equipment, and the public service network would have the benefit of integrating the power supply into its system. Needless to say, any of these arrangements would require the approval of the regulatory agency having jurisdiction and pos- sibly implementing legislation as well. There seems no reason, however, why participation in joint-ownership or lease-back arrangements should be confined to privately-owned electric suppliers. 4. Consumer Contributions to Capital Through Construction Surcharges with (or without) the Issuance of Securities in Return As has been pointed out, the case for requiring an annual contribution to the capital of an electric enterprise by the in- dustrial firms served by it is rendered especially strong by the disproportionately heavy drain on electric supply that indus- trial firms frequently impose and the profits that such firms derive from this access to a service provided for the entire com- munity. But the whole community has a stake in providing electricity in a volume sufficient to meet all industrial, commer- cial, and residential needs. This fact argues for the extension of the capital contribution requirement to all consumers whose demand for service is in excess of a prescribed minimum. To expand the suggestion we made above with respect to indus- trial users, we suggest that the commercial and residential consumers also receive securities of the supplying enterprise (or enterprises) in return for their contribution to its funds for construction. 236 Part 2. Suggestions for Corrective Action More specifically, this would look to empowering the regu- latory authorities to authorize the supplying electric enterprise to add what we shall term a "construction surcharge" to its regular bills to all industrial users and to all commercial and residential users consuming more than a prescribed monthly minimum (which might be set at a level high enough to elimi- nate one-half or possibly two-thirds of all residential users from the plan). This surcharge might take the form of a flat per- centage, say, 10%, or it might be graduated so that those users whose demand was largest would make the largest proportion- ate contribution to the expansion of their supplier's capacity. This suggestion is made only on the assumption that the regular Tates for electricity would have been increased to the extent necessary to cover the full cost of the service. The fact some consumers had to invest in the capital of the enter- prise serving them should not relieve them of the duty to meet the costs of its operation, including the cost of capital. More- over, they should be assured of the soundness of the securities they would be receiving in return for the surcharge before it should be imposed. This would require not only that the regu- lar rates were adequate but that adjustment procedures were in existence that would keep them so. To provide for construction surcharges on all industrial, commercial, and substantial residential consumers poses a question as to the character of the securities to be issued to the consumers. We believe the suggestions we advanced in the dis- cussion of the industrial users would be applicable to the ex- tension of the plan here suggested, with certain qualifications noted below. Since the number of persons having the securities of the enterprise would be vastly increased by the operation of the plan, almost certainly a specially designed issue of securi- ties would be desirable to facilitate the plan's administration and to avoid complicating the internal affairs of the supplying enterprise. To attempt to give voting rights to all the thousands of holders of minute fractions of this issue would scarcely seem practicable. However, provision could be made, effective after Suggestions for Obtaining New Capital 237 the plan's operation had created a substantial volume of the special security, for the election of a director by all the share- holders holding, say, five or more shares each of the special security. This representation could be increased as the total contribution to the enterprise's capital from this source grew to greater size. One of the practical problems that the extension of the plan to a large number of consumers would present would be the small size of the contributions that would be made by many residential and small commercial users if the surcharge were imposed on a monthly basis. How could these small payments be evidenced by corporate securities? The problem would be a real one even though the minimum consumption level pre- scribed for the plan's operation were set at a point which eliminated a half or even two-thirds of the residential users and the small commercial users. The size of the smaller monthly surclharges would still represent small fractions of the smallest practicable corporate share. However, this is a problem that should not defy solution long. One way of handling it might be drawn from the experience of large merchandisers in the United States and doubtless other countries who issue "trading" or "premium" stamps to their customers in amounts propor- tionate to their purchases. These are pasted in stamp books issued for the purpose and, when a book is full, it is presented for a piece of merchandise offered as a "premium" to partici- pants in the plan. Similarly, stamps in small denominations could be issued to residential users paying small electric bills. These could be collected by the users in form books issued for the purpose and, when a book had been filled, it could be pre- sented to the electric company and exchanged for a share-or, in the case of publicly-owned enterprises, a debenture. A some- what similar plan was first employed on a large scale by the United States Government during World War I to facilitate the sale of "War Savings Stamps." Stamps of very small de- nominations were pasted in books, and, when the book was full, it could be exchanged for "Liberty Bonds." Much the 238 Part 2. Suggestions for Corrective Action same plan is in force today. Post offices sell U. S. Savings Stamps in denominations as low as 10. A book containing $18.75 in stamps may be exchanged for a bond. Would the complications introduced by the extension of construction surcharge on this basis to relatively small users be worth the trouble it would cause? The need for capital on the part of the electric industry in Latin America promises to be a vast and continuing one, and a steady, if relatively small, flowv of funds for expansion from any suitable source is not to be scorned. More important, however, is the desirability of estab- lishing widespread consumer-investor relationships. Financing of the expansion program of the electric industry, if this is not to take inflationary forms, requires that capital be drawn from private savings, regardless of whether the enter- prises are government- or privately-owned. The consumer- investment program could develop a nucleus of private invest- ors for the electric enterprises and in time do much to establish an active capital market in all their securities. (The transfer- ability of the special securities should, as previously suggested, be restricted for a limited period of time, except to the issuing enterprise or to persons succeeding to the supplied premises.) After the period of restricted transfer expired, the consumer who held the securities would often find himself reluctant to sell, if, at least, the enterprise had paid interest or dividends regularly during the period. Consequently if he were obliged to sell, the price he would demand for the securities would then equal or approach their real value. The argument may be advanced that the proceeds of a con- struction surcharge should not be reflected in any increase of the investment base or other rate base of the enterprise. This argument is based upon the proposition that the effect of re- flecting the fruits of the surcharge in the base would be to increase rates. This objection may be answered as follows: (1) The increased investment should not result in higher rates since the investment itself would be income-producing. Presumably, it should earn at the same rate as existing property Suggestions for Obtaining New Capital 239 and, therefore, would "carry its own freight." In point of fact, extensions to the system, by producing a more efficient opera- tion, would have a tendency to lower rates. True, the enterprise would have to receive more revenues to service the larger amount of securities outstanding, but the new investment should, in effect, service itself. (2) The amount by which total net earnings of the enter- prise would be increased would be presumably about the same as the earnings to which the consumer-owners or consumer- creditors of the enterprise would be entitled by way of divi- dends or interest. Not only would there be no need for rate increases to meet these payments, but the payments would represent an offset to the contributions which the consumers obtaining securities had been obliged to make. (3) If the expansion fund were not included in the invest- ment on which the enterprise was entitled to earn a return, the payments on the securities issued to consumers could come only out of the earnings available to the holders of the enterprise's other securities. Either these earnings would have to be in- creased to compensate for this diversion or the other securities would fall in value and attractiveness and so destroy the enter- prise's access to capital from sources other than its consumers. These considerations indicate that the argument against in- cluding compulsory contributions to the capital funds of an electric enterprise in its investment base must proceed on the assumption that the consumers paying the surcharge would not receive income-producing securities in return therefor. This indeed is the more familiar form taken by proposals for analo- gous funds. These proposals merit consideration if the gains resulting from better consumer acceptance and wider security distribution under the plan we have suggested seem outweighed by the practical problems that that plan entails. To distinguish a construction surcharge for which no securities are issued from a rate increase of equal amount, the proceeds must be excluded from the investment base (or the rate base if traditional meth- ods are adhered to). The treatment of the surcharge proceeds 240 Part 2. Suggestions for Corrective Action would be similar to the treatment of the proceeds of deprecia- tion expense described in the previous chapter. Resort to a construction surcharge without compensating securities has been had in recent and proposed legislation in two Latin American countries but with a complication of questionable desirability in this context-a requirement that improvements and extensions built with the proceeds of the surcharge be public property.8 Thus, the law for the electric industry recently adopted by Peru provides for an "extension fund." The annual appropriation for this fund may not exceed 5% of the value of the investments made in primary feeder and secondary distribution systems, including customers' contribu- tions. This percentage will be calculated on the property as revalued in accordance with the law.9 Improvements made with extension fund moneys are to constitute public property.'0 The proposed new law for the electric industry in Chile recom- mended by the Director of Electric Services contains a similar provision, whereby a company could collect from its consum- ers annually an amount for an expansion fund, the property constructed therewith being public property. Consideration of a special construction surcharge must take account of existing special taxes designed to finance electrical development by government-owned entities. The 10% tax on electricity consumption in Mexico payable to the Comision Federal de Electricidad and the Brazilian 1954 law creating the Electrification Fund which will obtain its main revenues from a tax on electric power were discussed in Chapter III, as wvere the special taxes in the Brazilian States of Minas Gerais and Rio Grande do Sul designed to raise funds for their respec- tive state-owned enterprises. 8 Since presumably these public acquisitions will be added, piece by piece, year by year, in a system that is expanding by the investment of its own funds, the requirement seems likely to produce an oddly fragmented pattern of owner- ship and to achieve little, if anything, that exclusion from the investment or rate base would not do-except to exert some deterrent effect on private investment. 9 Peru, Law No. 12,378, D. 0. July 14, 1955, Article 131. 10 Id, Article 132. Suggestions for Obtaining New Capital 241 Are such taxes consistent with our suggestion that electric enterprises be permitted to impose a construction surcharge and give their securities in return as a method of raising some of the domestic funds they need for expansion? Obviously, there are limitations on the total charges that may be imposed for the use of electricity. Certainly to some consumers, forced investment would at first appear no different from a tax or a rate increase: it would merely constitute additional money being paid for electricity. It may appear that to abolish special taxes earmarked for construction and to substitute a construction surcharge for which securities would be issued would prejudice the govern- ment-owned enterprises and favor the privately-owned com- panies. Where the government enterprise sold directly to consumers, no problem should arise since it should be quite as able as the private company to administer a security-purchase plan, issuing debentures in return for the surcharge payments, perhaps including an adjustable-return feature in their terms. Moreover, even where the government-owned enterprise is a wholesaler, it can cooperate with the privately-owned dis- tributing company in administering a securities-purchase plan. The amount of the surcharge could be turned over to the enterprise making the expenditures for expansion, and its se- curities would be issued to the consumers, or a process of divid- ing the proceeds regularly in agreed-upon proportions could be worked out. This may be illustrated by a specific example: In Mexico, a large part, in fact, the major part, of CFE's expansion has been concentrated on increasing generating capacity in the area served by Mexlight. For the year 1953 generation of the combined CFE-Mexlight system amounted to about 50% of the total generation of electricity for the entire country. If CFE continues to expand generating facili- ties in the Mexlight area, a part of a construction surcharge imposed on Mexlight consumers could be turned over to CFE and a part to Mexlight, based on their respective ex- pansion programs in the area. Users of electricity would re- ceive, in turn, the securities of both enterprises. To avoid 242 Part 2. Suggestions for Corrective Action needless fractionalization of securities, relatively small con- sumers might receive only the securities of the distributing company, the other's securities being held in the distributing company's treasury. Or perhaps no securities would be issued against that part of the surcharge going to the public body. The same situation prevails in other areas where govern- mental enterprises operate primarily as wholesalers of elec- tricity. Allocation of the construction surcharge would provide funds for both the generating and distributing enterprises. Another objection to a construction surcharge in lieu of special taxes is that the more backward areas of the country would not receive as much in the way of funds as the more de- veloped, since the former contribute less. While this is true, it should be borne in mind that the major expenditures that have been made in the post-war expansion have been to supply the areas already supplied with electricity, and expenditures in the more isolated areas have been quite meager. Further, the re- quirements of these areas are so great that the amounts derived from special taxation of electricity consumption could repre- sent only a small part of their overall needs. Funds for this purpose must be derived largely from other sources. Moreover, do the users of electricity in developed areas represent a suit- able source of funds for subsidizing electric developments in other areas? Experts in public finance have long deprecated the tendency of governments everywhere, but especially in the less developed nations, to meet many pressing demands on their develop- mental budgets by the imposition of special taxes, the proceeds of which are to be earmarked. The possible sources of govern- mental revenue become pre-empted by institutions which be- come dependent on these taxes. In consequence, the taxing governments, faced by new needs, find themselves deprived of the financial flexibility to meet them. The construction surcharge suggested here, even when em- ployed by governmental enterprises, is less open to this objec- tion than the tax for construction purposes, at least where the Suggestions for Obtaining New Capital 243 consumer receives securities in return for his payment. In time these will be paying income equal to the surcharge, and he therefore retains a tax-paying capacity to a much greater extent than the consumer paying a special tax. And the government has not been required to subsidize the expansion of the enter- prise. Rather, the enterprise has been aided by government to pay its own way. CHAPTER XII A Summary of Suggestions for Corrective Action In the preceding five chapters, we have advanced a series of suggestions for the regulation and financing of the electric in- dustry in Latin America. Translated into law and practice, we believe these would enable that industry not only to become self-sustaining but also to accumulate and attract the capital it needs for expansion. If the Latin American electric power in- dustry were free to assess the real costs of its service against the industries and the people whom it serves, it could cease from competing for limited government funds with other pub- lic services that have no income-producing potentialities. The electric industry then could hopefully face the challenge of the, dawning atomic age with its vast demands for new capital. The industry would not have to fear that the only way it could get domestic funds would be through the inflation- feeding round that has grown familiar: subsidies from the government to the industry, loans from the central bank to the government for the subsidies, and the creation of new money by the bank to provide the wherewithal for the loans. As we have repeatedly declared above, the cost-of-service principle is not confined to the fixing of the rates of privately- owned electric suppliers and such differences as may be neces- sary in its application to government-owned suppliers are not fundamental. However, if government is to play an im- 244 Summary of Suggestions for Corrective Action 245 portant proprietary role in the development of the electric industry, it can best do this through autonomous, financially independent authorities; otherwise it will be hard to tell whether a given producer has in fact resisted either subsidiza- tion or exploitation for other governmental purposes, a com- mon practice among municipalities. The suggestions we have advanced have not been tailored to the needs of any one country, though they are of special rele- vance to countries that have suffered from, or are threatened by, serious price inflation. However, we are convinced that any country with capital shortages which is seeking to strengthen its electric industry to meet the capital demands of expansion will find that providing protection in its regulatory system against inflation will prove worthwhile as a means of attracting investors to its securities. This should be true even though its price level and exchange rates have long been stable. If prices rise dangerously or the exchange value of the domestic money begins to depreciate, the built-in protection these measures give would become invaluable. One may ask why protective devices of this sort have not been resorted to more extensively in the United States (some states do use methods of valuing the rate base which are de- signed to protect against change in the price level) . There has been argument for the adoption of more such measures, but conditions in the economy of the United States have prevented the need for them from growing acute. The industry has bene- fited from the combined effects of steady technological im- provement, a plenitude of institutional capital seeking sta- bility of monetary income rather than stability of purchasing power, and a rate of price increase which has been low rela- tive to the rate in most of Latin America. These factors have enabled the United States to preserve regulatory standards for electric rate adjustment which are neither fully nor promptly responsive to changes in price levels. At the same time, the industry has been able to attract the capital it has needed for continuous and rapid expansion. 246 Part 2. Suggestions for Corrective Action In Chapter VI, we stated three major lines of regulatory policy that wve believed should be pursued by Latin American nations seeking to place electric rates on a cost-of-service basis. We shall reiterate these policies below and note under each the suggestions we have advanced in the preceding chapters for its realization. We wish to repeat here an earlier recognition of the fact that there are other ways in which the cost-of-service principle can be realized. Our primary concern is for the reali- zation of the principle, rather than for the particular methods we have suggested, helpful though we believe them to be. First, the industry should be able to adjust its rates rapidly enough so that operating cost increases will not prevent it from realizing an adequate level of earnings. Any such plan of ad- justment should apply equally to assure rate decreases in re- sponse to cost decreases. Since changes in rates which Tequire official approval before they can take effect are typically delayed long after the cost changes they are designed to meet, we have suggested in Chap- ter VII the authorization of automatic compensatory adjust- ments for cost increases and decreases of the kinds listed below: (1) Increases or decreases in the price of fuel, the price of purchased power, and in wage and salary rates. (2) Or, as an alternative to the foregoing, increases or de- creases in an index constructed to reflect the principal costs of operation and maintenance experienced by an electric utility of the type for wvhich the index is to be used. (3) Increases or decreases in the rates of taxes applicable to the utility. (4) Increases or decreases in the exchange rate for the dol- lar or other foreign currencies in which any of the en- terprise's securities are to be serviced, or to which the amount of payment in national currency is tied. Compensatory adjustments for these purposes would be con- Summary of Suggestions for Corrective Action 247 trolled by commission-approved formulas designed to assure that the changes in rates would be proportionate to the changes in total costs caused by the particular cost changes for which adjustment was sought. The changes in rates could be made either at stipulated periods, say, quarterly, or whenever the net change in the costs exceeded a specified percentage. Second, as a means of enabling the industry to compete with unregulated investment opportunities for national savings andl foreign investments, provision should be made for regular ad- justments in the capital accounts of electric enterprises, public as well as private, to reflect significant changes in the domestic price level and, when relevant to outstanding securities, in the exchange value of the national curr.ency. The measures we have suggested to this end are developed in Chapters VIII, IX, and X. They are the following: Revaluation of Capital Accounts We suggest that the regulatory body pursuing this course should be freed from basing its rate determinations either on the original cost of the utility property or on unrealistic, time- consuming property revaluations. We suggest the abandon- ment of the property rate-base x rate-of-return formulas to which, largely for historical reasons, United States regulatory bodies had become legally tied and which are still being used in the countries studied. We suggest instead that the deter- mination of the earnings to be allowed the utility in calculat- ing its revenue requirement be based wholly on the capital accounts of the enterprise and that these be revalued every year, more frequently in case of rapid price rises. The rate of return allowed to the equity in the enterprise would be applied to its common stock and surplus accounts (or to the govern- ment's proprietary interest in a public authority not taking the form of a stock corporation). The adjustments in these ac- 248 Part 2. Suggestions for Corrective Action counts would in this manner be reflected in the earnings of the enterprise. The revaluation of capital accounts would have as its start- ing point the initial determination of "an investment base" for each enterprise. This should be an equitable judgment figure which, after provision had been made for the enterprise's out- standing debt and non-participating preferred shares, would measure the base to which the rate of return to the equity in- terest would be applied. At its inception, the investment base could not be determined by a formula of general applicability but should reflect both the past history of the enterprise and its current capital structure and economic situation. In the initial determination of the investment base, care would have to be taken to prevent overvaluation of the equity interest since ordinarily a part of the capital would have been furnished the enterprise by security issues which were fixed as to yield and principal amount. After the investment base had been determined, it would appear in the capital accounts of the enterprise as its common stock and surplus accounts or as the government's proprietary interest in an electric authority not organized as a stock cor- poration. The annual adjustment would call only for the application of changes in the selected index to these accounts. However, probably the amount of the adjustment could best be indicated by a special Capital Surplus Adjustment Account, balanced by a Capital Asset Adjustment Account. At intervals, if increases shown in the account appeared likely to be perma- nent, they could be capitalized. A participating preferred stock presents special questions. Since, in a regulated industry, opportunities for participation wvould be restricted unless the preferred as well as the common capital was adjustable, it seems likely that, for purposes of re- valuing capital accounts, participating preferred capital stock should be included along with the common. Whether the ef- fect of changes in exchange rates on debt payable in foreign moneys or of changes in indexes on indexed issues of debt and Summary of Suggestions for Corrective Action 249 preferred shares should be shown by changes in the capital accounts for those securities or by proper notations is a prob- lem of accounting practice, rather than of substance. The ac- counts representing such securities would not be in the invest- ment base. The cost-of-living index is suggested as the basis of revaluing the capital accounts since that index measures roughly the current purchasing value of the investment. The application of this index to electric utility securities also provides an ap- proximate equivalent to the revaluation, or inflation, of the capital investment in unregulated industrial corporations and so helps the electric industry to compete with them for domes- tic capital. Without provision for revaluation, utility shares could thus compete only if permitted to earn (on an original cost or investment base) a markedly higher rate of return than that enjoyed by industrials. Since this would scarcely be de- sirable in a regulated industry, it seems clear that adjustments in a utility's rate of return do not constitute a practicable sub- stitute for revaluations of its capital. Rate of Return The rate of return, even if expressed as a single figure, should nevertheless represent a weighted average of the rates of return on the respective elements in the capital structure of the electric enterprise. For those securities on which the return was contractually determined in amount (whether or not adjustable), the allowed rate should be that determined by the contract. The main problem would be that posed 'by the equity shares-or by the proprietary interest in a government power authority. In the case of privately-owned utilities, once provision had been made for the periodic revaluation of capital accounts comprising the investment base and for adjusting the return to the equity interest reflected therein, the rate of return could 250 Part 2. Suggestions for Corrective Action be set at whatever figure appeared likely to be necessary to at- tract equity capital after the new system of regulation had had a chance to demonstrate its merits in operation. This figure should allow for the retention of a part of the earnings for reinvestment in the enterprises; not all should be paid out in dividends. The rate set should be lower than the return which probably would have to be paid to attract equity capital at the very start of the new system. It would take time for domestic investors to come to appreciate the protection afforded them by the combination of automatic compensatory adjustments and periodic revaluations of capital. However, the rate of re- turn needed to attract capital should drop as investor confi- dence grew and domestic demand for utility investments en- larged. New equity financing might well be deferred until the earnings performance of utilities had been tested in experience. The equity interest in a government-owned electric en- terprise must be contributed initially by the governmental body creating the power authority, although the plan adopted may look to the latter's eventual emancipation by its repay- ment out of earnings of the capital thus advanced. If this is not done, the public body creating the power authority will bear a relation to it like that of a holding company to its sub- sidiary, irrespective of whether the corporate form is used and the subsidiary's shares issued to the parent body. In any event, if subsidization is to be avoided, the latter should receive a rate of return on its continuing investment at least equal to the real cost to it of the capital it has supplied. If, however, the electric enterprise is not to remain dependent on its gov- ernmental parent for capital, the return should be set at a higher level. To enable the power authority to continue expanding its plant in response to the needs of the public it serves, the earn- ings allowed to it should enable it to finance a substantial amount of the required construction directly and to support such additional borrowing as may be needed. The level of re- turn necessary for this purpose must be reckoned in the light Summary of Suggestions for Corrective Action 251 of the circumstances of each particular enterprise. In general this policy has been pursued by the municipally-owned electric authorities of the United States which, by maintaining returns on their equities at a high level, have built up their proprie- tary interest, in the aggregate, to more than 50% of their elec- tric plant. (This has an original cost, less substantial deprecia- tion reserves, of well over $2 billion.) This policy has also been written into the standard contract that the Tennessee Valley Authority has entered into with the municipal electric distrib- utors it supplies, requiring them to increase their electric rates whenever these fail to yield returns assuring a sound financial position, taking into account their need to raise funds for ex- pansion. In pursuing this policy, the municipalities supplied by TVA have been able to meet a substantial part of their needs for financing construction from internally-generated capital funds, while maintaining rates well below the national aver- age. Since Latin American power authorities confront infla- tionary conditions much more serious than those in the United States and also lack a market for fixed-return electric securi- ties, they will require higher returns than the United States municipal authorities have secured if they are to achieve com- parable results. Depreciation Belated efforts to revise inadequate systems of depreciation to enable depreciation accounting to discharge its multiple functions in electric utility control run into serious difficulties. In the United States, the problems grew most acute in rate cases when it became necessary for rate base purposes to decide what deductions for depreciation should be made from the utility's plant accounts. Happily, under the suggestion we ad- vanced in Chapter VIII that the rate of return should be ap- plied to the capital, rather than the property, accounts of the enterprise, it would not be necessary to reduce the investment 252 Part 2. Suggestions for Corrective Action base by the amount of depreciation accruing annually. This accrual would be placed in a reserve and the proceeds used for reinvestment in the enterprise. However, since the funds thus accumulated and invested would have been contributed by the consumers through the allowance for depreciation expense in the electric rates, the enterprise should not be allowed to earn a return on the assets thus acquired. In this respect, the de- preciation reserve would differ from surplus built up by the retention of earnings. The latter would enlarge the investment base since it would represent a contribution to the capital in- vested in the enterprise by the proprietary interest. Typically depreciation expense allowances in the Latin Amer- ican countries studied have been insufficient. We suggest that they be required to be increased at least to the straight-line level promptly and provision made to reflect this increase in the electric rates. Since in inflationary economies the original cost of most assets being depreciated will have fallen far below replacement cost, a fair distribution of the burden of replace- ment between present and future consumers suggests that the current level of depreciation expense be adjusted periodically to reflect the changing level of utility construction and equip- ment costs. It should be noted, incidentally, that this does not have to be done by the same adjustment factors as are used to adjust the capital accounts if, as we have suggested, the deprecia- tion reserve has been kept independent of the investment base. Third, the regulatory authority (and any financial authori- ties having jurisdiction) should be prepared to approve forms of securities that are adapted to the industry's needs in an in- flationary economy. They should adopt measures, where neces- sary, to open up markets for the distribution of those securities among institutional investors and perhaps should require in- vestment by such of the electric industry's business and even residential customers as make especially heavy demands upon its capacity. Summary of Suggestions for Corrective Action 253 Policies designed to adjust electric rates to changes in op- erating costs and to meet changes in the cost of capital by periodic revaluations of the industry's capital would not alone suffice to bring about a substantial flow of domestic investment funds into the industry. Equally necessary would be the issu- ance of securities designed to pass the benefits of such rate and capital adjustments through to investors. Moreover, given the long alienation of investors from the industry in Latin Amer- ica and the high returns paid on other forms of investment there, special measures, not excluding compulsion, may be re- quired to establish the practice of placing investment funds in the industry. The affirmative measures that were suggested in Chapter XI are outlined here. (We repeat that they can suc- ceed only if measures giving effect to the first and second poli- cies are adopted.) 1. The protection of investors against persistent price infla- tion by (a) providing for the adjustment of interest rates on industry obligations (and possibly dividends or preferred shares) in accordance with the cost-of-living index or, as in France, with the cost of electricity;* (b) issuing obligations payable in the issuing country in dollars or in the local cur- rency adjusted by its rate of exchange with the dollar; (c) is- suing obligations, possibly indexed as to return, convertible into common stock; (d) issuing preferred shares with provision for participation with the common in all earnings after divi- dends on the common had equalled those paid on the preferred; and (e) in the case of municipal or other publicly-owned authorities, issuing income debentures enabling the holders to share in earnings above a fixed rate. If any such securities were to be issued, the issuer of course would have to be able to in- crease its electric rates along the lines previously suggested so that it could meet the increased claims of its security holders in case the price level continued to rise. 2. The encouragement of institutional investors to acquire electric industry securities might open up a substantial market for these securities, as has long been the case in the United 254 Part 2. Suggestions for Corrective Action States. Although in Latin America life insurance has not been a major source of savings for investment, the social security and pension funds are such. If legal barriers to such investment were removed in a country, the investment of substantial amounts of these funds in the electric industry (once its re- turn and securities had been placed on an inflation-proof basis) would be beneficial not only to the industry and the national economy but also to the social security funds and their bene- ficiaries. 3. The benefits accruing to industry from the availability of abundant electric power argues strongly that Latin America's power-using industries might be required to make a more sub- stantial contribution to the electric industry's capital needs than they have. As a consequence of making such contribu- tions, these industries would as a rule experience lower costs than they now incur from power bottlenecks and irregularities or from efforts to escape these by the private purchase of diesel generating plants. The following means of enlarging the in- dustrial contribution to the electric industry's capital needs have been canvassed: (a) By substantially increasing electric rates charged indus- trial consumers, the rates being held to levels that would not exert serious pressure on product prices or on in- dustrial profits; (b) By requiring that an industrial concern wishing new or added electric supply purchase an amount (propor- tioned to the increase in supply) of the electricity sup- plier's stock or other securities, preferably from a secu- rity issue specially designed for this purpose, with restricted transferability; (c) By requiring instead that every industrial customer above a certain minimum demand purchase every year a reasonable amount of electric supplier's securities, presumably smaller than would be required under (b) above and probably graduated in relation to demand; Summary of Suggestions for Corrective Action 255 (d) By permitting one or more industrial companies to join with an electric enterprise for the joint construction and ownership of generating facilities, the output to be shared by the industrial concerns and the electric sup- plier; and (e) By permitting one or more industrial companies to con- struct a generating plant and to lease it back to the elec- tric supplier on a power-sharing basis. 4. The conception of compulsory annual subscription to an attractively designed electric security might be extended be- yond industrial users to the commercial and larger residential consumers. This might take the form of an annual construction surcharge for which a special category of stock or debentures would be issued in return. If the issuance of many fractional shares were necessitated, a stamp book system might be de- signed to enable the individual user to cumulate his fractional shares into a full share. This device (employed by the United States in the sale of its obligations to small purchasers) would be used in the hope that it would also increase consumer in- terest in the well-being of the electric supplier and enhance the likelihood of voluntary subscription to more stock in the future. The device is one that seems preferable to a special electricity tax or construction surcharge which provides funds for governmental construction while giving no specific return to the electricity consumer who must pay it, not even the as- surance that the funds so collected will be expended on the system serving him. * ~~* Clearly, the effect of placing the electric industry of Latin America on a self-sustaining basis by setting electric rates on the cost-of-service principle would involve increases in existing rates and would require further increases if-but only if- present inflationary tendencies persisted. But the cost to Latin American economies of the chronic underpricing of electric 256 Part 2. Suggestions for Corrective Action service is far greater than the amounts that would have to be paid in increased electric rates. That cost is now being paid in inadequate and irregular electric supply, in the acquisition by industry of expensive alternative sources of supply, and in the accentuation of inflationary pressures by the means that now have to be used to finance the electric industry's expan- sion. In the meantime, this potentially self-sustaining industry drains away government funds and credit needed for public works that have no like earnings potential. Experience elsewhere is reassuring. As the recent study of the European electric industry by OEEC makes clear, those coun- tries which took steps to bring electricity prices into line with costs did not suffer as a consequence. The resulting increases were found to have "represented only a very small fraction of the increase in overall expenditure on private and public con- sumption" in the four-year period covered by the study. It seems, moreover, that the "changes in rates have not had any appreciable effect on economic activity or the standard of living." Indeed, "far from having had unfavorable repercus- sions, the measures taken in several countries have somewhat improved the electricity supply industry's position on the capi- tal market." After pointing to the danger of keeping the selling price of electricity "artificially below cost," the report warns that a "tariff policy which paid no heed to future problems and prospects would not only be proved wrong in the light of ex- perience in recent years, it would also hamper economic de- velopment in the Member countries."' At the end of Chapter I, we called attention to four circum- stances that had quieted such concern as we might otherwise have felt in suggesting electric rate increases sufficient to meet the industry's cost of service: the startlingly low level of exist- ing electric rates in the countries studied; the economic feasi- bility of not extending proportionate rate increases to small residential users; the high indirect costs of poor, inadequate 1 OEEC, The Trend in the Selling Price of Electricity in its Relation to the Financing of New Plant (Panis, Nov. 1958) pp. 23-24. Summary of Suggestions for Corrective Action 257 electric service; and the low percentage of electric costs to the value added by manufacture. These circumstances reveal how challenging is the oppor- tunity now open to the Latin American governments in the countries we have studied to remodel their systems for the regulation and financing of their electric power industries. We believe a like opportunity may also exist in other Latin Amer- ican nations and in many other parts of the world. Everywhere countries have been discovering that their electric supply, often in surplus in the 1930's, is now in short supply. No longer is there a need to encourage greater consumption of electricity. Policies, whether of regulation or of pricing, must now be adapted to the fact to which we pointed earlier: that what a developing nation needs more than cheap electricity is plenti- ful electricity. And the best way to get it is to pay what the service costs. APPENDIX Post-War Use of Indexed Bonds for Financing Public Utility Expansion in France In the period since World War II, France has had to contend with problems of financing expansion of electric power and other public utilities similar to those encountered in the Latin American countries studied. Demand has grown Tapidly; new facilities have been provided mainly by public bodies (Elec- tricite de France for electricity, Gaz de France for gas, etc.); the government budget has been under almost continuous pressure, in part because of investment demands; private in- vestment has proceeded vigorously and has offered strenuous competition to public utilities for a share in the supply of new savings; inflation has been a fact or a threat. Nevertheless, the French problem has not been as difficult as that faced by most Latin American countries: France possesses a well-organized capital market, backed by a historical tradition of thrift and of willingness to transfer savings to enterprises not under the con- trol of the saver. Thus the chore of selling fixed-interest securi- ties in exchange for new savings should have been less onerous in France than in Latin America. Even so, inflation had badly impaired this market in the post-war years, and special new securities have been required to restore the flow of savings 259 260 Appendix available to public-utility and other borrowers who can at best promise only a limited return. Developments in the critical period 1950-54 are summarized in Table A-1. Table A-1. New Issues through the French Capital Market, 1950-1954 (in billion francs) Type of issue, by type of issuer 1950 1951 1952 1953 1954 Fixed income: Nationalized industries ....... 16 24 - 14 3 Other public (including banks) 67 19 44 93 192 Private issues ......... ...... 11 11 7 8 9 Total fixed income ......... 94 54 51 115 204 Indexed bonds: Nationalized industries ....... - - 31 48 67 Government ................. - - 194 - - Private issues ................ - - - 16 35 Total indexed bonds ....... - - 225 64 102 Variable Income (shares of stock) 31 44 67 54 80 Source: OEEC, The Supply of Capital Funds for Industrial Development in Europe (Paris, Jan. 1957) p. 80. The OEEC points out in its discussion of these statistics: The consistently weak demand for straight debentures issued by private companies, and the absence of a revival since 1952, are mainly due to the risk of erosion of capital against which this type of security offers no hedge. Since the introduction of indexed bonds and debentures, the investors' response has been most significant: the annual total issued by the private sector more than doubled between 1953 and 1954. Until 1952 index clauses in bonds and debentures were prohibited by law. Under the present rules loans of any kind may include an index clause, provided that its effect is not exclusively to compensate the lender for depreciation of Appendix 261 the currency. Indexes are therefore related in some way to the borrower's production, productivity, the price of the products manufactured by him or of the services he performs. Participating debentures have also been much in favor in the last two or three years. Some of the issues of this type of security (e.g., the debentures issued in 1954 by the Groupe- ment des industries de la construction electrique) carry clauses providing for variations not only in interest but in principal also, based on the borrower's turnover and hedged by a guaranteed minimum of progression. In other cases the variations are geared to the increase in ordinary dividends and/or reserves. Since February, 1953, debentures carrying options of conversion into ordinary shares have also been permitted, but this innovation has not led to significant re- sults so far.' One of the most noteworthy features of Table A-1 is the great reliance of the nationalized industries in France (mainly public utilities, or with similar characteristics with respect to capital needs) on indexed bonds since their issuance was first permitted in 1952. These bonds surpassed fixed income issues by these industries in a ratio of more than 9:1 in the triennium 1952-54, and were issued in steadily greater amounts in each succeeding year. In 1953 and 1954, indexed bonds accounted for at least one-third of all bonds issued by all borrowers. The index feature blends into the participating feature of some debentures. Therefore the main forms of both will be summarized together: 1. Index Based on Gold Value of the Franc This index cannot legally be used in France except for direct government borrowing authorized by special legislation. The 1 OEEC. The Supply of Capital Funds for Industrial Development in Europe (Paris, Jan. 1957) pp. 80-81. 262 Appendix 3V2% Rentes of 1952, indexed as to principal but not as to interest, account for the "Government" issue of indexed bonds shown under 1952 in Table A-1. These bonds are to be retired over the years 1960-2012 at a price equal to 25 times the average price in a recent period of a twenty-franc gold piece, but not less than 100,000 francs. They are also receivable at this price in payment of certain taxes. 2. Participating Bonds or Debentures, with Index Based on Gross Revenues a. Compensation bonds have been issued by the Caisse Na- tionale de l'Energie to previous security-owners in privately- owned enterprises acquired by Electricite de France and Gaz de France as a result of nationalization. These are 3% sinking- fund bonds (to 1996), which in addition participate each year in a sum equal to 1%, of the gross revenues of Electricite de France and Gaz de France. This 1% is divided each year into an "interest share" and an "amortization premium," with the former fixed at 3%, of the latter. Thus the securities have up- ward leverage from two sources: the assumed increase through time of the gross revenues on which the 1% bonus is calculated, and the steady decline in the number of bonds outstanding to share in this larger bonus as the sinking fund does its work. b. Participating bonds with similar characteristics have been issued for reasons other than nationalization. For example, a Steel Industry issue of November, 1953 (Emprunt du Groupe- ment de l'Industrie Siderurgique) provided for basic interest of 6%/, and for amortization over a twenty-year period. Both the interest and the minimum premium of 5% on bonds called for the sinking-fund are subject to upward adjustment of 10% of the percentage increase in steel production above a base of 10.5 million tons per annum, plus 25% of the percentage in- Appendix 263 crease of steel prices calculated on a base of 30,000 francs per ton. Thus a 10% increase in production would add 1% to both interest and the principal payable on retirement, and a 10% price increase would add 2.5% to interest and principal. These bonds are only partially indexed, in two respects: the index coefficients are only 0.10 and 0.25, so the bondholder shares in only one-tenth of any production increase and one- quarter of any price increase; moreover, price and quantity increases are added and not multiplied, so that a doubling of both prices and production would increase the rate of interest on the original value of the bond by 12% (200% of 6%) to a total rate of 18%, and not by 18% (300% of 6%) to a total of 24%. The electrical construction industry has issued a bond, also 6% and with a twenty-year maximum maturity (Emprunt du Groupement de l'Industrie Electrique), which is based on the multiplication of price and quantity increases via an in- crease in both interest and sinking-fund values with each in- crease in gross revenues. Moreover, the relationship is 1:1; each 1 % increase in gross revenues is accompanied by a 1% increase in interest and capital value. Finally, the purchaser is guaranteed a minimum 17% increase in interest and siinking- fund price for each year he holds the bond. Thus his interest return will advance from 6% in the first year to at least 6% plus (19 x 1% x 6%), or 7.14% in the twentieth and last. 3. Bonds Price-Indexed as to Interest and Principal Issues of the electricity supply industry are most relevant for this discussion. They include the Parts de Production de l'Electricite de France. The annual coupon for these bonds is 200 times the average price of one kilowatt hour of electricity in the preceding year, with a minimum of 1,440 francs. They are amortized in ten equal drawings at 4,000 times the average 264 Appendix kilowatt-hour price, with a minimum of 32,000 francs. Begin- ning in 1958, the bearer gains an option to demand redemption at 32,000 francs. Gaz de France has issued similar securities, to be amortized in thirteen equal drawings. The same method has been employed, in various forms, by the National Railways (SNCF). For example, a 1953 issue re- lates both interest and principal to the price per kilometer of third-class railway tickets. The annual interest payment is 450 francs, as a minimum but the bondholder is entitled at his option to receive a sum equal to the average price of 90 kilo- meters of third-class transportation; as a further option, he may take "transport coupons" which provide 100 kilometers of third-class transportation or any other passenger or freight service of equal value. Amortization gives the bondholder a choice of 10,000 francs, or of the value of 2,000 kilometers of third-class transportation, or of "transport coupons" which may be used for 2,100 kilometers of third-class transportation or the equivalent. Thus the SNCF bonds not only index both interest and principal; they also give thie holder an option of payment in kind at bargain rates.2 4. Bonds Price-Indexed as to Principal Only This method has been used by the nationalized coal industry (Charbonnages de France, Bons a Capital Indexe). These bonds were issued in 1953, with maturities of three, six and nine years. The holder of a 10,000 franc bond is entitled to 2 This option would provide an interesting precedent for Latin America if the idea could be effectively transplanted in the form of "electricity bonds" to be sold to industrial users. Although the French Railway issues are offered for voluntary subscription, the arrangement might prove particularly desirable as a means of payment for capital contributions required as a condition for starting or expanding public-utility supply of electricity to industry, as is suggested in Chapter XI. Appendix 265 575 francs interest for each of the first three years and 600 francs thereafter, with no index feature. They are repayable on demand in 1956 and 1959 at 10,000 francs, or in 1962 at 10,000 francs plus a premium based on the French coal price index. These various examples are a, tribute to the ingenuity of management of French industry, private as well as nationalized, in appealing to non-speculative savers who fear inflation. The statistics of securities sold indicate that this ingenuity has re- ceived at least some reward. It is quite possible that such issues have even reduced inflationary forces in France by providing a new incentive for saving. Moreover, indexed bonds are cheaper to the borrower at the time of issue, in terms of cost of issue as well as rate of guaranteed interest: ... indexed bonds cost about 8% to the borrower, although they carry 6% interest; for non-indexed bonds the cost is 8.5%o to 9% for an interest rate of 6.5% to 7% per annum.3 Even this contrast probably understates the difference: it was made after a period of relative price stability in France; it does not take into account the degree to which each type of security appeals to voluntary savers as opposed to buyers who create money to make their purchases or are subject to legal limita- tions which force them into the fixed-interest market. It cannot allow for the probability that the very growth of the indexed bond has improved the marketability of fixed interest obliga- tions by reducing the supply. The general principle is at least as applicable in Latin America as in France. Certain features of the French system reflect conditions more or less peculiar to the country, however, and therefore should not be copied di- rectly in Latin America. a. The compensation bonds issued in connection with the nationalization of electricity and gas supply, with their return adjusted to total revenues, would be much too generous as a continuous financing method for a growing industry. Each past 3 OEEC, op. cit. supra note 1, p. 161. 266 Appendix issue would then secure a claim on the revenues derived from all future plant expansion. The result would be a cumulative financial burden. The cyclical stability of the electricity supply industry is great enough to permit adequate financing without recourse to a gross-revenue index. b. Conversely, the indexed issues of the electricity and gas supply industries might not be attractive enough for voluntary savers in Latin America. The value of an index based on the average price of electricity depends entirely on the methods used for pricing electricity. In France, late in 1952: The former specifications still in force include a clause for varying prices in relation to an economic index for electric- ity based on fluctuations in the price of coal, the cost of labor and the index of retail prices.4 Although the French Government had intervened to modify this index by direct price-fixing, as of November 1952 the high-tension index was 7,300 as compared with 8,300 in accord- ance with the formula, the low-tension index was 11,800 and the domestic lighting index 10,850 as compared with 13,700 based on the formula. Moreover, the principle of indexing had been adopted as long ago as 1919. So French kilowatt-hour bonds were offered against a background of indexed adjust- ment of the price of electricity which does not exist in Latin America. The suggestions for Latin America contained in the main body of our report are based on relating electricity prices to costs, including the cost of capital. This seems preferable, when an entirely new approach must be devised, to the French system of relating part of the costs of capital to prices deter- mined on the basis of independent criteria. c. The French prohibition against the use of general cost-of- living indexes may be only nominal if indexed bonds are based on the price of one product which is in turn influenced by the cost-of-living index. To the extent that the prohibition is mean- ingful, it must reflect a basic fiscal policy decision to attempt 4 OEEC, The Price of Electricity (Paris, Nov. 1954) p. 57. Appendix 267 to check inflation by forbidding sliding-scale arrangements throughout the economy. These arrangements are already very common in many Latin American countries; the absence of cost-of-living bonds has discriminated against would-be savers at the expense of saving generally and therefore at an added cost in inflationary pressure. The most important practical issue concerns the relative extent of latent desire for bonds in France and in Latin America. France has had a potential mar- ket for bonds which had to be restored. In Latin America, the market must be created. Therefore, in Latin American coun- tries, reliance should be placed on the cost-of-living index, the simplest and most familiar, as well as the best-adapted to the task of protection of savers against future price movements. Index* Agricultural users Baurti, Brazil, 74, 75, 78. of elec.: low nunibers of, 22; high Belo Horizonte, Brazil, 13. consumption of individual, 22. Bogotd, Colombia, 17, 23, 73, 74, 76, American & Foreign Power Co. 78, 90, 98, 214. nature of systems owned by, 13; Bonds subsidiaries of, 13; original acquisi- see Securities. tions by, 13-14; acquisition of Cali Brandeis, Louis D. system from, 17; sale of securities Justice, U. S. Supreme Ct.) quoted. by, to doniestic investors, 37-38, 214, 123; assumptions underlying "orig- to U. S. investors, 39. inal cost" theory of, 166. American Gas & Electric Co., 232. Brazil Argentina, 13, 14. gov't ownership of elec. plants in, Automatic compensatory adjustments 16, 18; per capita consumption of in clec. rates: need for, in inflation- elec. in, 22; urban growth rates in, ary economies, 147, 150; determina- 23-24; increases in industrial use of tion of amount of, 151; control of, clec. in, 25-26; rates of return in, by regulatory agencies, 152; inade- of elec. andl other industries, 29-30; quacy of surcharges to compensate capital issues for elec. industry in, for past cost increases, 152; lack of compared to U. S., 31-32; revalua- cost decreases to offset need for, tion of assets permitted in, 30, 37, 152-153; effect of, on efficiency of 117; policies of, in financing gov't management, 153-154; not inflation- owned elec. and other utilities, 39- ary in effect, 154; use of, for operat- 45; division of responsibility for ing and maintenance expense, 156- clec. power development between 158, for taxes, 158-159, for changes Federal and State gov'ts, 42, 44-45; in foreign exchange rates, 159-160; use of "mixed" corporations in, 58- provision for, in Brazil, 125. 59, 214; financing elec. expansion Bahia, Brazil in, by development bank, 60-61, by see Salvador. special taxes, 64-66; trend of elec. Bank of Mexico rates in, 1929-54, 72-75; 1954 loans of, to Mexican Government, monthly elec. bills in, 79, 81; ratio 51-52. of elec. cost to value added by Barran quilla, Colombia, 13, 17, 78, manufacture in, 94-95, to family 81, 90, 98, 99. budgets, 98; effect of clec. supply *In this index, "electric," "electrical" and "electricity" are abbreviated as "elec."; "government," as "gov't"; "Latin America," as "L. A."; "United States of America," as "U. S." 269 270 Index shortages in, 96-97; households and to Endesa's (Chile), 50; Mexican communities in, with elec. but lack- gov't investment in, 52-53; gross op- ing water and sewer services, 99-100; erating profits of, 53-54; special rate-base x rate-ol-return formula taxes as source of capital for, 64-66; prescribed by law of, 116; provisions 1954 monthly elec. bills of, 80; pos- for use of escalator clauses in, 125- sible sharing of construction stir- 126; tabular view of regulatory sys- charge with Mexlight, 241-242. tern of, 131; public acceptance of, Ciapala participating pfd. stocks in, 220, (Nueva Compafiia El6trsca de Cha- 221; rate of life insurance savings pala, Mexico) , 18, 80. in, 224; net accrual of social secur- CHESF ity fund in, 224. (Companhia Hidro-Eletrica do Sao Brazilian National Developnment Bank Francisco, Brazil), 16; failure of, to (Banco Nacional do Desenvolvi- attract private investment, 58-59. mento Econ6mico), sources of capi- Chihuahua, Alexico, 74, 76, 80. tal available to, 60-61; elec. devel- Chile opment loans by, 61, 125. gov-t ownership of elec. plants in, Brazilian National Economic Council 16-17, 18; per capita consumption quioted, 118. of clec. in, 22; increases in indus- Brazilian Traction, Light & Power Co. trial use of elec. in, 25; financilng of origin of, 12; size of, 13; moves to gov't owned. utility (Endesa) in, permit sale of securities to domestic 47-50; use of "mixed" corporations investors, 38; rate of consumption by in, 59; trend of elec. rates in, 1929- residential customtiers of, 90; results 54, 72-75; 1954 monthly clec. bills of power shortages of, 96-97. in, 79, 81; ratio of elec. cost, to value Cali, Colons biea, 17, 90. added by manufacture in, 95, to Canada family budgets in, 98; rate-base x holdings in, of L. A. elec. power rate-of-return formula prescribed companies, 14; early technical as- by law of, 116; "regulatory lag" in, sistance given by, 14. 126; depredation terms of Chilena Capital Assets Adjustment Account contract, 128; tabular view of regu- use of, to balance Capital Surplus latory system of, 131; rate of life Adjustment Account, 173, 207, 248. insurance savings in, 224; net ac- Capital Surplus Adjustment Account crual of social security funds in, use of, to reflect adjustments capital 224; proposed law in, provides for accounts, 173, 248; to be included expansion fund, 240. in "investment base," 173; capitali- Colombia zation of, 173, 188, not to be taxed, gov't ownership of elec. plants in, 188. 17, 18, 19; per capita consunmption CEEE of clec. in, 22; urban growth rates (Comissao Estadual de Energia in, 23-24; low returns allowed mu- Eletrica, Rio Grande do Sul, Bra- nicipal utilities in, 46; use of zil), 42. "mixed" corporations in, 59-60; CEMIG trend of elec. rates in, 1929-54, 72- (Companhia Eletrica do Minas 75; 1954 monthly elec. bills in, 78- Gerais, Brazil), limited private in- 79, 81; ratio of elec. cost, to value vestment in, 59; special taxes as addted by manufacture in, 95, to source of capital for, 65-66. family budgets in, 98-99; cotumuni- CFE tics in, served by elec. and water (Comisi6n Federal de Electricidad, and sewer systems, 100-101; rate- Mfexico) , 18; policies of, compared base x rate-of-return formula pre- Index 271 scribed by law of, 116; tabular view to, in countries studied, U. S., and of regulatory system of, 131; rate of OEEC countries, 98-99; advantage life insurance savings in, 224. of, index for adjustments in capital Combined Mexican Working Party, accounts, 171-172, 249; usc of index 51. of, on French indexed securities, Compainia Chilena de Elect ricidad 260, 266-267. origin of 14; expansion expected "Cost-of-service" principle of, 17; trend in rates charged by, in rate regulation: as objective of 74, 75, 104-105; low rates of, to suggestions made in study, xiii, 141; small consumers, 108-109; "regula- major lines of policies for imple- tory lag" affecting, 126. menting, 138-139, 246-247, 252; ap- Compailia Colombiana de Electrici- plicable to both gov't and privately dad, 17, 74, 76, 119. owned elec. suppliers, 141, 244; dif- Companlia Nacional de Fuerza y Luz, ficulty of defining, 142; similarity 128. of, to existing rate-making theory, Construction surcharge 142; conditions impeding realiza- suggested as source of capital for tion of, 143-144. elec. expansion, 236-237; issuance Costa Rica of securities in return for, 237-238; gov't ownership of elec. plants in, arguments for allowing return on 17-18; per capita consumption of proceeds of, 238-239; compared with elec. in, 22; stability of price level special elec. taxes, 240-243, with in, 55; public ownership of gov't "extension fund" in Peru law ancd debt in, 55-56; financing of new proposed Chilean law, 240. elec. plant in, 56-57; trend of elec. Consunzers rates in, 1929-54, 72-75: .1954 of elec.: see Agricultural users; In- monthly elec. bills in, 79, 81; yield d ustrial users; Residential consum- on gov't debt in, 84; ratio of elec. ers; Consumption of electricity. cost, to value added in, 95, to family Consumption of electricity budgets in, 98; rate-base x rate-of- increases in, in L. A. relative to rest return formula prescribed by law of world, 20; increases in per capita of, 116; provision for depreciation rate of, 20-21, as between urban in, 128; tabular view of regulatory and industrial uses, 22-23. system of, 131; rate of life insuLrance Corporacidn de Fomento savings in, 224. (Chile) . as source of capital for En- Cuba dlesa, 48, 59. 13, 39, financing of Cuban Electric Cost of capital Co. by social security funds, 225- of elec. power industry: high level 226. of, in relation to revenues, 83; in Deepwater Light & Power Co. U. S. compared to, in L. A., 84-85, creation of, by elec. utilities and 186; persistent rise in, 144-145; DuPont Co., 232-233. automatic adjustments for foreign Depreciation allowances exchange effects on, 159-160; diffi- importance of, in financing expan- cult to ascertain, for equity and sion of elec. power industry, 6, in proprietary interests, 182, 184. See protection of investor, 127, 201, 205; Rate of return. effect of inadequate, on rate of re- Cost of living ttirn of Mexlight, 35-36, 129, of elec. rates adjusted for changes in, CFE (Mexico), 54; extraordinary, indexes in countries studied, 74-76; provided by Endesa (Chile) rates, exchange rates adjusted to reflect 48, used for self-financing, 48-49; changes in, 77-78; ratio of elec. costs change to straight-line method of 272 Index calculating, in U. S., 127-128, 201- Electric generation 202, partially in L. A., 128, need for, Canadian technical aid in develop- in L. A., 206, 252; of gov't owned ing, 14; efficiency of U. S. and L. A. elec. systems in U. S., 195; of pri- hydro plants compared, 85, of vately owned elec. systems in U. S., thermal plants, 86; technological 200; functions of, in utility rate reg- progress in, 86, capital cost of new ulation, 201; based on original cost, capacity for, in L. A. compared to criticized as inadequate, 205, 210- U. S., 87-88; economies of scale in, 212; proposals for basing, on re- 88, 143; load factors in, in U. S. and placement cost, in U. S., 210-216, in L. A. compared, 89; investments Great Britain, 211-212, in L. A., for, and distribution compared, 92; 212, 252. difficulty in replacing high-cost Depreciation reserves plant, 144; joint ownership or lease- importance of, in financing expan- backs of plants for, 231-235. sion of elec. power industry, 3, 6, Electric operating costs 200; failure to provide adequate, in labor factor in, low for hydro Colombia, 46; of gov't ownled elec. plants, 86; customer costs in U. S. systems in U. S., 195, 197; of pri- and L. A. compared, 90-91; effect vately owned elec. systems in U. S., of uncompensated increases in, 124- 200; deduction of observed depre- 125; review of managerial discre- ciation rather than, in finding "fair tion as to, 126-127; rate increases value," 202; deduction of, from needed to cover increases in, 138, original cost, 202-204; difficulty of by automatic adjustments, 156-158. deductinlg adjusted, 205-206; sepa- Electric rates ration of accounting for, from in- effect of, on demand for electricity, vestment base suggested, 207-210, 69, 136, on contribution to national 252. income, 70; residential, in 1929 and Development banks 1939, in countries studied, 72, in as means of tapping new sources of 1939 in selected areas, 73, compared capital for elec. financing in Brazil, to U. S., 7, 72-73; "real," in coun- 60-61, in Mexico, 62-63. tries studied and in U. S., for 1929, Diesel generators 1939 and 1954, 74-75; monthly clec. high cost of, compared to public bills in 1954 in countries studied, utility charges, 7; resort to privately 78-80, compared by weighted aver- owned, in Brazil, 97; cost of install- ages to U. S. bills, 81; structures of, ing, as limiting rate increases to 1in- in L. A., for industrial power, 93; dustrial users, 228; joint ownership relation of elec. cost, to value added of elec. plant as alternative to by manufacture, in L. A., 94, 228; private, 233. in selected industries in Brazil and Dividends Mexico, 95; to family budgets in "pay-out ratio" of dividends to countr-ies studied, 98; "promotion- earnings, 122-123, 187-188, 250; use al" rate strcyc structurs e of, 102- of stock, to capitalize retained earn- 103; change to anti-promotionial ings, 188. stru;cture of, in Chile, 104-105, in DuPont de Nemours Co., 233. Medellin, Colombia, 105-107; level ECLA and structure of, reviewed by regula- see United Nations Economic Com- tory agency, 113-114; use of escalator mission for Latin America. clatises in, 125-126, defenided, 152- Ecuador, 13, 39. 155; frequent small increases in, Electric Bond & Share Co., 13. preferable, 154-155; changes in, un- Index 273 der suggested rate-making plan, 40; "inadvertent" and deliberate 178; in form of required subscrip- subsidization in, 40-41; Federal and tion to elec. supplier's capital by State policies for, in Brazil, 41-45; industrial user, 228-230, by com- by subsidization in Colombia, 45- inercial and residential users, 235- 46; use of self-financing by Endesa 238; in form of construction stir- (Chile) 47-49; by gov't investment in charge, 239-240. CFE (Mexico), 52; subsidization in, Electric service in Costa Rica, 56-57; from proceeds deterioration of, 29, 96-97. of special taxes, 64-66, 240; by pro- Electric traction industry vision for self-financing in rate of as factor in development of elec. return, 191-192, 250; in the U. S., power industry, 12-13. 193-196, 251; by sale of securities to Electricite de France institutional investors, 223-226; by issuance by, of bonds indexed to required sale of securities to indus- price of elec., 217-218, 263-264, de- trial users, 228-230; to all types of pendent on methods of pricing users, 236-237; by joint ownership elec., 266. and leaseback arrangements, 231- Endesa 235. (Empresa Nacional de Electricidad, Financing of privately-owned electri- S.A., Chile), origin and growth of, cal suppliers 16-17, 47; pressure of inflation on, as alfected by relative rates of re- 47-49; rate and depreciation poli- turn for elec. and other industries, cies of, 47-48, 58; new financing re- in Brazil, 29-30, in Mexico, 32-33; quirements of, 48-50; policies of, volume of capital issues for, in Bra- compared to CFE's (Mexico), 50; zil and U. S., 31-32; effect of exist- private investment in, 59, 214. ence of capital markets on, 32; re- Escalator clauses cent Mexican experience in, 32-36; permitting increases in elec. rates by sale of shares to domestic in- to reflect cost increases, 124-126; vestors, 37-38, 214; by sale of securi- Brazilian procedure for, 125-126; as ties to institutional investors, 223- device for automatic compensatory 226; to industrial users, as condition adjustments, 151; statutory author- to expansion, 228-230; to all types ity for, not needed, 156; use of, for of users, 236-237; by joint ownership fuel and other operating and main- and leaseback arrangements, 231- tenance expenses, 156-158, in Flor- 235. ida, 157. Fink, Marvin S. Export-Import Bank conduct of study by, ix-xi. see U. S. Export-Import Bank. Florida, 157. Federal Power Commission. Foreign exchange see U. S. Federal Power Commnis- effect of depreciation of local cur- sion. rency on elec. costs, 7, 68, 129, 144, Federal Power Comnmission v. Hope on foreign shareholders, 129-130; Natural Gas Co., 115. effect of devaluation of Mexican Financiera Nacional de Cuba peso on Mexlight valuation, 36, on arranged for financing of Cuban Impulsora valuation, 130; revalua- Electric Co. from social security tions of Endesa (Chile) debt to re- ftunds, 225-226. flect movement of, 47-48; compon- Financing of government-owned elec- ent in total investment needs of trical suppliers L. A. elec. industry, 67-68; rates of, recourse to central banks in, 28; for converting elec. rates, 72, 77, 78; effects of, by creation of new money, "purchasing power parity," rates, 274 Index 77-78; use of automatic adjust- Harvard Law School, viii. ments for rate changes in, 159-160; ICE suLggested balance sheet treatment (Instituto Costarricense de Electri- of debts payable in, 172; issuance cidad), 17; financing of new elec. of bonds payable in dollars or plant of, 56. other, 218-219. Industrial users France of elec.: resort to diesel generators indexed bonds authorized by, 217, by, 7, 226; increasing intensity of 260-261; bonds indexed to price of tise by, and factors causing, 24-25; clec., 217-218; factors affecting se- increase in growth of Brazilian, curities market in, compared with with heavy demand for elec., 25-26; L. A., 259, 265-267; new issues in ratio of elec. cost to value added by capital market of, 260, prevalence manufacture, in L. A., 8, 94, in se- of indexed securities in, 261; meth- lected industries in Brazil and Mex- o(ds of indexing securities in: on ico, 95; costs to, of poor elec. service, gold value of franc, 261-262, on 96-97; investment by, in elec. sectiri- gross revenues, 262-263, as to both ties, 130; need for, to take greater interest and principal, 263-264, as to part in financing clec. industry, 227, principal only, 264-265, with option by paying higher elec. rates, 227- for payment in kind, 264; use of 228, by required subscription to se- cost-of-living index prohibited by, curities of elec. industry 228-230; 260, 266-267. sharing by, in joint ownership of Fuel clauses clec. plant, 231-234, in leaseback see Auttomnatic compensatory adjust- arrangement, 234-235. ments, Escalator clauses. Industrialization Government-ownied electric systems rapid, as complicating problems of in Latin America: growth of, after elec. power indtistry, 6, 226-227. World War II, 15, in Brazil, 16, in Inflation Chile, 16, in Colombia, 17, in Costa as complicating problems of elec. Rica, 17-18, in Mexico, 18; source power regulation, xii, 4, 6, 37, 175; of fuLnds for, 18, 138, 189-191; em- effect of financing of elec. indtistry phasis in, on hydroelectric power, on, 28, 40, 44-45, 47-49, 52-53, 244; 18, on generation, 19; multi-purpose rate increases not catise of, 139; ati projects in, uncommon, 19; gov't totnatic adjustments not cause of, regulation of, 111; adjtistment of 154; risk of, as affecting pay-out capital accounts of, 170-173; at- ratio, 189; as affecting adequacy of traction of capital by, 189; deter- depredation allowances, 205, 210- mining rate of return on "equity" 212, 252; elec. industry securities interest in, 190-193; provision for designed to protect against, 215-223, self-financing by, 191-192, 250; re- in France, 259-267. turn earned by, in U. S., 194-197; Insutrance companies by TVA distributees, 196-197; rate- must contribute to Brazilian devel- making standards in TVA contracts opment bank, 61; holdings by, of with, quoted, 197-198; advantage of U. S. utility bonds, 223; as possible autonomy in, 245. market for L. A. elec. securities, Great Britain 224. 14, Parliamentary Committee of In- Interest rate quiry into the Electric Supply In see Rate of return. dustry in, quoted, 211-212. International Bank for Reconstruction Cuanajuato, Mexico, 13, 80. and Development Guatemala, 13, 39. (WVorld Bank) , viii, as source of Index 275 capital for elec. power industry, 2, example of successful, 232-233; anal 4, in L. A., 66-67; limitations on lend- ogous arrangements for atomic re- ing by, 2; loans by, to Colombian actors, 234. .,mixed" corporations, 59; loans by, Leaseback of generating plant compared to total capital needs of as means of financing expansion of L. A. elec. industry, 67-68. elec. system, 234-235. International Monetary Fund Los Angeles, Calif. officials of, quoted, 225. excerpt from Electric Plant Revenue International Power Co., 14. Bonds of, 199. Investment Medellin, Colombia in L. A. elec. power industry: need 17, 23, 72, 78, 90, 98, 99, 213; struc- for expanding, 1, 3, total require- ture of elec. rates in, 105-107; low ments estimated, 67; other public rates to small consumers in, 108-109. services competing for, 2-3; early Metropolitan Life Ins. Co., 39. foreign, 12-14; relation of, to na- Mexican Light & Power Co. tional product, 21; difficulties in at- ("Mexlight"), 13; control of, 14; tracting, for elec. industry, in Eu- generating capacity of, 18; Nacional rope, 3, in L. A., in 1930's, 15, after Financiera loan to, 33; rate increases WVorld War II, 15, 27, 137, 213-214; granted to, 34, 119; rates of return capital issues for, in Brazil and U. S. received by, 35-36; 1954 elec. rates cornpared, 31-32; by sale of shares charged by, 80; possible sharing o1 to domestic investors, 37-38, 214, construction surcharge with CFE, despite adverse regulatory policies, 241-242. 38; by sale of securities to U. S. in- Mllexico vestors, 38-39; failure of "mixed' gov't ownership of elec. plants in, corporations to attract, 58-60; types 18; per capita elec. consumption, 22, of securities used to attract, by Na- increase in industrial use of elec. in, cional Financiera (Mexico) , 62-63; 25; financing private elec. suppliers advantage of diversified securities in in, 32-36; rates of return on capital attracting, 182-183, of properly de- in elec. and other industries, 32-33, signed securities, 185; securities pro- 84; recent elec. rate increases in, tecting, against inflation: indexedl 34, 119; heavy gov't investment pro. bonds, 216-218, dollar bonds, 218- gram in, 50-51, its effect on public 219, convertible debentures, 219-220, cleficits, 51-52; financing elec. ex- participating pfd. stock, 220-222, pansion in, by Nacional Financiera, participating municipal debenttures, 62-64, by special taxes, 64-66, 240- 222-223; potentialities of, by insti- 241: trend of elec. rates in, 1929-54, tutional investors, 223-226, 253-254; 72-75; 1954 monthly elec. bills in, suggested required, by industr ial 80, 81; ratio of elec. cost, to value users, 228-230, 255. added by manufacture in, 94-95, to Investment base family budgets in, 98; rate-base x concept of, 171; need for determin- rate-of-return formula prescribed by ing initial, 175, 248, by case-by-case law of, 116; rate of return calculated method, 176-177; restricted to equity before income taxes, 35, 123; esaila- or gov't proprietary interest, 177- tion of rates not granted in, 126; 178, 248; adjustment of, separated tabular view of regulatory system from depreciation accounting, 206- of, 131; issuance of dollar securities 210. in, 218-219; rate of life insurance Jalapa, Mexico, 80. savings in, 224; net accrual of social Joint ozvnership of electric plants security funds in, 224. advantages resulting from, 231, 233; AMexico City, 12, 73, 75, 76, 80, 98, 102. 276 Index Mining by laws of countries studied, 116- use of elec. in, 13. 117; objectives of advocates of "pres- "Mixed" corporations ent value" and "original cost," as use of, to finance elec. development measures of, 164; how "fair value," in Brazil, 58-59, in Chile, 59, in aided common shareholders, 165; Colombia, 59-60; failure of, to at- "investment base" suggested in place tract piivatc investment, 58-60, 214. of, 170, 173, 246; "reproduction Monterrey, Mexico, 80. cost," as alternative to adjustable Morelia, Mexico, 80. investment base, 179; deduction of Nacional Finsanciera depreciation reserves in determin (Mexico), financing by, of Mexlight ing, 202-204, difficult when reserves construction, 33; methods used by, are adjustable, 205. to obtain investment in clec. ancd Rate of return other industries, 62-64, 219. relation of, to risk, 6; in U. S. an(d National Assn. of Railway & Utilities L. A. compared, 6; rates of, in elec. Commissioners, 202. andl other industries compared, in Organization for European Economic Brazil, 29-30, in Mexico, 32-33; Cooperation ("OEEC") earned by Mexlight, 35-36; inade- quoted, on financial problems of quacy of, in Brazilian utilities, 41- European elec. power industry, 3-4, 45; increases in, obtained by Endesa 70, on effect of elec. rates on demand (Chile), 47; low level of, obtained for power, 69-70; on return on gov't by CFE (Mexico), 53-54; does not equity, 191, economic effects of rate include interest on construction, 54; increases, 256, on acceptance of in- need for adequate, by "mixed" cor- dexed bonds in France, 260-261; porations, 58-60; on various securi- ratio of elec. cost to gross national ties issued by Nacional Financiera product of members of, 94. (Mexico), 62-64; on debt and equity Panama, 13, 39. capital in U. S. and L. A. compared, Paulo Afonso Project, Brazil, 16. 84-85, 184-186; low level of, on new Pemex plant in Medellin, Colombia, 106: (Petr6leos Mexicanos), 50; issue of determining fair, as step in rate- indexed bonds by, 60, 64; low price making process, 113; high, not sub- policy of, 102. stitute for low rate base, 118, 167, Peru 198; statutory maximums prescribed as leader in modernizing elec. regu- in, Chile, 16, Colombia and Costa lation, 130; provision for "extension Rica, 119; statutory minimum pre- fund" by, 240. scribed in Mexico, 118-119; as af- Porto Alegre, Brazil, 13. fected by capital structures, 119, by Price of electricity "leverage," 120-121; usually calcu- see Electric rates. lated after income taxes, 123-124; Public ownership effect of "regulatory lag" on, 124; see Government ownership. reduced when market recognizes Puebla-Veracruz system, Mexico, 13, protection against inflation, 171, 73, 74, 75, 76, 80, 103. 250; calculated separately for eacti Rate base class of security, 181-183; difficulty determining, as step in rate-making in determining, for equity, 184, 250: process, 113; "present value" andI in gov't owned system, 190-193, 250- "original cost" as measures of, 114; 251; provision for self-financing in, "present fair valtie" as U. S. Con- 191-192, 250; earned by gov't owned stitutional standard, 114, 163, re- elec. systems in U. S., 194-197, by jected, 115, adhered to, 116; required TVA distributees, 196-197, 251; Index 277 standards to govern, of non-reguilated ing water and sewver services, 99-101; elec. system, 196-199; arguments for "Mexican cycle" rates for, 102-103; allowing, on securities obtained by changes in structure of rates for, in construction surcharge, 238-240. Chile, 103-105, in Medellin, Colom- Rate regulation bia, 105-107; relation of use by, to process used in passing upon elec. assessedl value of residences, 107; rates in U. S. and L. A., 112-113; construction surcharge on, to raise burdensome nature of, in U. S., 115; capital, 235-239, and issuance of se- controversy over rate base, 114-117: curities to, 236-238. problems in determining rate of re- Retained earnings turn, 117-124; combating regulatory as source of capital for expansion of lag in, 124-125; use of escalator alec. power industry, 3, 122-123, 187, clauses in aid of, 125-126; tabular 252: in Brazilian manufacturing as view of systems of, in countries stud- compared to elec. industry earnings, ied, 130; assumes long intervals be- 30; use of, in financing Endesa tween rate cases, 145, and stable (Chile) , 48; importance of, in at- economy, 146; criticism of traditional tracting capital, 122-123, 189; risk of approaches to cost of capital, 168- inflation affecting, 189. 169; new method of determining Return cost of capital in, 169-174; cdetermin- see Rate of return. iig investmenit base as step in, 175- Revalutalion of assets 177; by standards in TVA contracts, permitted by law in Brazil, 30, 37, 197-198. 117, 167: to reflect revaluation of Rates Endesa's (Chile) foreign debt, 47- see Electric rates, Rate of return. 48; necessary in determining invest- Recife, Brazil, 13, 74, 75, 78, 100. ment base, 175-176. Regulatory agencies Revaltuation of capital accouints position of, in governments in U. S. of L. A. elec. suppliers: need to pro- and t. A. compared, 110-111; re- vide for, 138, 148; method for, sug- view by, of managerial discretion as gested, 170-175, 248, illustrated, 207: to operating expenses, 126-127; not applicable to contractually de- power of, to control automatic com- termined accounts, 172, 174, partici- pensatory adjustments, 151, 152, pating pfd. stock excepted, 248; usc 155; need for initiative by, in de- of Capital Surplus and Capital As- veloping elec. industry securities, sets Adjustment Accounts, 173, 207, 214. 248; separation of, from handling Regulatory laig depreciation suggested, 206-208, 252. in adjusting elec. rates to meet cost Rio de Janeiro, 12, 73, 74, 75, 76, 78. increases, 6, 124-125, 143; use of Rio Grande do Stul escalator clauses to combat, 125-126f, (State of, Brazil) , electrification tax 151. imposed by, 65-66, 240. Residential consumers Ruggles, Prof. C. 0. of elec.: small proportion of total quoted, 233. revenues obtainecl from small, 7, in Salrvardor, Brazil, 13, 74, 75, 78. Medellin, Colombia. and Chile, 108- San josd, Costa Rica, 13, 74. 75, 76, 109; increase in number of, in L. A. 79, 90, 98. as compared to U. S., 90; volume of Santiago, Chile, 12, 13, 74, 75, 76, 78, use by, in L. A. as compared to 90, 98,103. U. S., 90, in Mexico City, 102; ratio Silo Paulo, Brazil, 12, 23, 43, 73, 78. of elec. expenditures by, to faniilv 98; inflationary financing of elec. budgets, 98-99; proportion of, ha'- utilities by State of, 44-45. 278 Index Sargent, H. B. Supply shortages quoted, 39. in elec.: effect of, on purchase of Schantz, Radford L., ix. diesel generators, 7-8, in Brazil, 97; Securities heavy economic burden imposed by, for L. A. elec. suppliers: sale of, to 8, 96-97; as check on demand for domestic investors, 37-38, 214; need clec., 21. to develop marketable forms of, Tacoma, Wash. 138, 214; opportunity of well-de- 1954 elec. rates in, 79. signed, to attract investors, 185, 215, Taxes 252; with returns indexed to cost of use of, for elec. development in living, 216, on price of elec, 217- Brazil, 60-61, 64-65, in Mexico, 64- 218, 263-264; payable in, or indexed 65, compared and appraised, 65-66; to, dollars, 218-219; convertible de- payable on revaluation of assets in bentures, 219-220; participating pfd. Brazil, 117, but not under suggestedl stock, 220-222; to be made eligible plan, 188; rate of return usually cal- for sale to insurance and social se- culated after income, 123; use of curity funds, 224-225, 253-254; spe- automatic adjustments to compen- cial terms for issue of, to industrial sate for changes in, 158-159; pay- users, 229, to commercial and small ment of, by gov't ownled elec. sys- residential consumers, 237-238; for rtems in U. S., 195-196; comparison methods of indexing, used in France, of special elec., with construction see France. surcharge, 240-243. Shareholders Torredn-Chihuahua System, AMexico, in elec. utility corporations: wind- 13, 80. fall to common, from use of "fair TVA value" rate base, 165; protection (Tennessee Valley Authority, U.S.A.) against dilution of interest of, 184; 19; 1954 electric rates in supply area return to common, in U. S., 186; of, 79; earnings of distributees of, protection of, by suggested treat- 196-197; rate-making standards in ment of depreciation, 209; repre- contracts of, 197-198, 251. sentation oE user, on board of di- United Nations Economic Commission rectors, 236-237. for Latin America Smyth v. Ames, 163, 176. report by, on "Energy in Latin Social security America," xi; estimates by, of in- contributions from Brazilian, to de- vestment needed by L. A. elec. in- velopment bank, 61; funds as source dustry, 67; quoted on effect of elec. of capital for elec. industry, 224- rates on elec. demand, 69, cottribu- 226; 253-254, approved by Intcrna- tion to national income, 70; use by, tional Monetary Fund officials, 225. of purchasing power parity ex- Sofina, 14. change rates, 77; quoted, as to ef- Subsidization fective hydro capacity in L. A., 85, of elec. power industry: "inadvert- as to relative efficiency of L. A. in ent" contraste(d with deliberate, 40- thermal generation, 86; capital cost 41; resort to, in Brazil, 44-45, of of generating capacity in L. A. as Colomhian municipal uitilities, 46, reported by, 87. in Mexico, 52, 54; of residential United States of America users in Medellin, Colombia, 107; similarity of regulatory systems in, of small consumers, 108-109; by to those in countries studied, xiii, granting favorable exchange rates, 4, 5; differences in economic situa- 129; of elec. service to developing tion of, from L. A.'s, xiv, 5, 6; elec. region, 142. rates in, compared to those in coun- Index 279 tries studied, 7; increase in, incdus- needs of industry, 67-68. trial use of elec. in, 25; capital issues U. S. Federal Power Com77nission, 128, for elec. industry in, compared to 202. Brazil, 31-32; trend of elec. rates in, U. S. Supreme Court compared to countries studied, 1929- Constitutional decisions of, govern- 1954, 72-76; 1954 monthly elec. bills ing rate regulation, 115, 163. in, 79, 80, 81; cost of capital in, com- Urbanization pared to cost in L. A., 84-85; ratio rapid, as complicating problems of of elec. cost to family budgets in, elec. power industry, 6, 22-23, urban 98-99; method of rate regulation growth rates in Brazil, Colombia, used in, 111-114; rate base standards and Mexico, 23-24. in, 115-116; determination of rate Valparaiso, Chile of return in, 118, 119-120, 122, 123; see Santiago, Chile. use of escalator clauses in, 125; earn- Value ings of gov't owned elec. suppliers see Rate base. in, 193-197, including TVA distrib- Venezuela, 13, 39. utees, 196-197; use of stamps for Water and sewer systens sale of bonds by, 237, 238; reasons households using elec. and having for limited use in, of suggested rate access to public, in Brazil, 99-100, plans, 245. communities with elec. but lacking, U. S. Export-Import Bank in Brazil and Colombia, 100-101. ("Eximbank"), as source of capital World Bank for elec. power industry, 2, 4, 39, in see International Bank for Recon- L. A., 67, compared to total capital struction and Development. (continued fro?n front flap) Professors Cavers and Nelson con- cIude that the evidence accuLmlulated by the study points to the necd to acljust charges for electricity services, wlhether uljlic or private, so that costs can be fully mlet ancd suLrpluses accumulatccl frol0m ealllilngs to finance future expansions. This conclusion is strongly enClorseCd both by the AVorlcl B1anlk andl by ECLA. Thlese agencies believe that many of the prevailing systemiis regulatilng electtic power UilitiCS in Lati[n Amiier-ica arc obsolete and are hindrl-anices to the rapid ac- cuLMulation of capital for expansion which is required by greatly increased power demanjd in Latin America. David F. Cavers is Fessencden Plro- fessor of Law, Harvarcl University, and mames R. Nelson is Charles E. Merrill Plrofessor of Economiiics, Am- herst College. THE JOHNS HOPKINS PRESS Ballim.or-e 18, Mlla yloa d Other books from THE INTERNATIONAL BANK FOR RECONSTRUCTION AND DEVELOPMENT QEBT SERVICING CAPACITY AND POSTWAR GROWTH IN INTERNATIONAL INDEBTEDNESS BY DRACOSLA V A VRAMO VI C 241 pages $5.00 "Carefully presented, the report olfers a Lreasure of information onl the clevelopmlelnt of external clebt, debt service, interniationlal trade, productioll, governmiiient revenues, andl goldl andl foreignl exclhange reserves of the various countries dul-illg tlle first post World War .11 decadle." The A nnais THE DESIGN OF DEVELOPMENT BY JAN TINBERGEN 128 pages $2.50 paperbound "TThis volume sets forth a nlumbCer of pr-inciples underlyillg the indLustrializa- tion of the less developed COUlntries of the wvorkl. The book starts by outlining those general conditionis essential to sustain ecomomiiic develop- ment, sketchinlg various policies which might be applied tlnder different circumstances as well as the gener-al measures employecl to stimtulate private and governmncit investilients." Foreign Commerce W4eekly DEVELOPMENT BANKS BY T1VILLIAM DIAMOND 141 pages $3.00 paperbound "The first publicationi of the Economic Development Institute, this book sets forth the objectives, funlctions, problemiis of formation and administra- tion of developmenit banks, drawing on the record of experience of such institutions in advanced as well as underdeveloped countries." Economnics Librar-y Selections THE JOHNS HOPKINS PRESS Baltimore 18, Maryland