Policy Analysis of Shadow Pricing, Foreign Borrowing, and Resource Extraction in Egypt Kemal Dervis Ricardo Martin SWP622 Sweder van Wijnbergen WORLD BANK STAFF WORKING PAPERS Number 622 WORLD BANK STAFF WORKING PAPERS Number 622 Policy Analysis of Shadow Pricing, Foreign Borrowing, and Resource Extraction in Egypt Kemal Dervis Ricardo Martin Sweder van Wijnbergen £ThRNATIOPiAL MONETARY FUND JOINT II3RARY MIAR ') 1984 INT IMNATIONAL BANiK FOR MYIZZOl.TRUC.I10N J?D D£'XELOPMEllT The World Bank Washington, D.C., U.S.A. Copyright O 1984 The International Bank for Reconstruction and Development / THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. First printing January 1984 All rights reserved Manufactured in the United States of America This is a working document published informally by the World Bank. To present the results of research with the least possible delay, the typescript has not been prepared in accordance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. The publication is supplied at a token charge to defray part of the cost of manufacture and distribution. The views and interpretations in this document are those of the author(s) and should not be attributed to the World Bank, to its affiliated organizations, or to any individual acting on their behalf. Any maps used have been prepared solely for the convenience of the readers; the denominations used and the boundaries shown do not imply, on the part of the World Bank and its affiliates, any judgment on the legal status of any territory or any endorsement or acceptance of such boundaries. The full range of World Bank publications is described in the Catalog of World Bank Publications; the continuing research program of the Bank is outlined in World Bank Research Program: Abstracts of Current Studies. Both booklets are updated annually; the most recent edition of each is available without charge from the Publications Sales Unit of the Bank in Washington or from the European Office of the Bank, 66, avenue d'Iena, 75116 Paris, France. Kemal Dervis is chief of the Industrial Strategy and Policy Division in the Industry Department of the World Bank; Ricardo Martin and Sweder van Wijnbergen are economists in the Bank's Development Research Department. Library of Congress Cataloging in Publication Data Dervis, Kemal. Policy analysis of shadow pricing, foreign borrowing, and resource extraction policies in Egypt. (World Bank staff working papers ; no. 622) 1. Egypt--Economic conditions--1952- --Mathematical models. 2. Egypt--Economic policy--Mathematical models. 3. Shadow prices--Egypt--Mathematical models. I. Martin, Ricardo, 1947- . II. Wijnbergen, Sweder van, 1951- . III. Title. IV. Series. HC830.D47 1983 338.962 83-21882 ISBN 0-8213-0277-9 Abstract This paper describes a long-term optimal growth model for Egypt focussing on issues of investment allocation, optimal saving and foreign borrowing, extraction of natural resources (oil and gas) and shadow pricing. All sectors of the economy other than oil and gas are aggregated into nontraded or traded goods; only the latter can substitute (imperfectly) for imported goods. Oil and tradable goods can be exported, although the latter faces a downward slopping demand curve. The model is run first with exogenous total investment and foreign borrowing; then investment is left free, subject only to an absorption capacity constraint limiting the speed with which it can increase from period to period. Finally, foreign borrowing is allowed to vary, with the interest rate charged to loans being an increasing function of the amount borrowed. In all variants of the model we test the sensitivity of the results to changes in some of the basic parameters (future oil prices, discoveries of new oil/gas reserves, pure rate of time preference). Besides characterizing the optimal growth paths in terms of the level and allocation of investment, the rate of oil and gas extraction, and the mix of domestic savings andi foreign borrowing, the model generates paths for the shadow prices of foreign exchange, natural gas and the social rate of discount, with application for project evaluation. Extracto En este documento se describe un modelo de crecimiento 6ptimo a largo plazo para Egipto, centrado en materias relativas a la asignacion de fondos para inversiones, el ahorro y los empr6stitos externos 6ptimos, la extracci6n de recursos naturales (petr6leo y gas) y los precios de cuenta. Todos los sectores de la economia distintos del petr6leo y el gas se agrupan como bienes no comer- ciados o comerciados; s6lo estos ultimos pueden sustituir (imperfectamente) a los bienes importados. El petr6leo y los bienes comerciables pueden exportarse, aunque estos ultimos se enfrentan a una curva de demanda descendente. El modelo se pone a prueba primero con inversiones totales y empresti- tos externos ex6genos; luego, se deja libre a las inversiones, con sujeci6n solamente a una limitaci6n de la capacidad de absorci6n que restringe la veloci- dad a la cual pueden aumentar de un periodo a otro. Finalmente, se permite que varien los empr6stitos externos, constituyendo el tipo de interes cobrado sobre los prestamos una funci6n creciente de la cantidad tomada en prestamo. En todas las variantes del modelo probamos la sensibilidad de los resultados a los cam- bios en algunos de los parAmetros basicos (precios futuros del petr6leo, descu- brimiento de nuevas reservas de petr6leo o gas, tasa pura de preferencia temporal). Ademas de caracterizar las trayectorias 6ptimas de crecimiento desde el punto de vista del nivel y la asignaci6n de fondos para inversiones, de la tasa de extraci6n de petr6leo y gas y de la combinaci6n de ahorro interno y empr6stitos externos, el modelo genera trayectorias para los precios de cuenta del tipo de cambio, del gas natural y de la tasa social de actualizaci6n, con aplicaci6n en la evaluaci6n de proyectos. Ce document decrit un modele de croissance optimale A long terme pour l'Egypte qui fait intervenir les decisions d'investissement, l'epargne optimale et l'emprunt exterieur, 1textraction des ressources naturelles (petrole et gaz) et l'utilisation des prix virtuels. Tous les secteurs de l'6conomie autres que celui du petrole et du gaz sont agr6ges en biens non echanges ou echanges avec l'ext6rieur; seuls les biens echang6s avec l'ext6rieur peuvent se substituer (imparfaitement) aux biens importes. Le petrole et les biens echangeables peuvent etre export6s bien que, pour ces derniers, la courbe de la demande soit descendante. On utilise d'abord le modele pour obtenir l'investissement total exogene et l'emprunt exterieur; puis on laisse l'investissement fluctuer, 6tant entendu qu'une capacite d'absorption limitee freine le rythme auquel il peut s'accroitre d'une periode A l'autre. Enfin, on laisse varier l'emprunt exterieur, l'int6ret percu sur les prets 6tant une fonction croissante du montant emprunte. Dans toutes les variantes du modele, nous avons teste la sensibilit6 des resultats aux variations de certains des parametres de base (prix du p6trole a terine, d6couvertes de nouveaux gisements de petrole/gaz, taux de preference pure pour le present). Non seulement le modele permet de caract6riser les sentiers de croissance optimale du point de vue du volume et du choix des investissements, du rythme d'extraction du petrole et du gaz, et du dosage epargne interieure- emprunt exterieur, mais il permet de determiner des trajectoires pour les prix de reference des devises, le gaz naturel et le taux d'actualisation social, qui pourront servir pour l1'valuation des projets. Table of Contents Chapter Description Page 1. Introduction: The Basic Questions 1 2. The Initial. Conditions: Egypt at the Beginning 6 of the 1980's 3. The Model 19 3.1 Production Functions 22 3.2 Material Balances 23 3.3 The Balance of Payments and foreign debt 25 3.4 Capital Accumulation 27 3.5 Resource Exhaustion Constraints 28 3.6 The Objective Function 29 4. Valuation on the Optimal Path 32 4.1 Conditions for Static Efficiency 32 4.2 The Social Discount Rate 39 4.3 Dynamic Optimality Conditions 42 4.4 Changes in Relative Prices 45 5. The Nature of Optimal Growth Paths: Structural Adjustment in a Long-Run Perspective 47 5.1 Experiments with fixed total investment 47 5.1.1 A Base Case Scenario 48 5.1.2 The Base Case 50 5.1.3 Shadow Prices in the Base Case 61 5.1.4 Sensitivity to Oil and Gas Reserves 68 5.1.5 Sensitivity to the Real Price of Oil 75 5.1.6 Sensitivity to the Growth of Domestic Demand for Energy 88 5.2 Experiments with Endogenous Investment 92 5.2.1 A Fundamental Optimality Condition 92 5.2.2 The Base Run 99 5.2.3 Variants on the Base Run 112 5.3 Experiments with Endogenous Investment and Optimal Foreign Borrowing 124 5.3.1 The Base Case with Foreign Borrowing and Endogenous Capital Accumulation 125 5.3.2 An Alternative Oil Price Scenario 129 6. Conclusions 135 6.1 Some Robust Policy Conclusions 135 6.2 Shadow Prices and Project Evaluation 137 Table of Contents (continued) 7. Footnotes 142 8. References 144 9. Appendices 145 9.1 Appendix A: A Summary of the Egyptian Model 145 9.2 Appendix B: Calibration of the Model 155 9.3 Appendix C: Absorptive Capacity Constraints 158 9.4 Appendix D: The Steady State Capital Stocks 160 Figures Figure 1: Share of Exogenous Resources in Total Resources 9 Figure 2: The Impact of an Exogenous Foreign 11 Resource Transfer Figure 3: Share of each Sector in Gross Output, Value 15 Added and Employment Figure 4a: Pricing of Natural Gas When Use is 35 Demand Constrained Figure 4b: Pricing of Natural Gas When Supply Constraints 38 are Operative Figure 5: The Ratio of Exogenous Resources to Non Oil 52 domestic Value Added Figure 6: Oil Production and Exports 54 Figure 7: Measures of Export Expansion and Import 56 Substitution Figure 8a: Share of Tradable in Domestic Non-oil Economy 58 Figure 8b: Ratio of Consumption of Tradable to Non 59 Tradable Goods Figure 9: The Composition of Exports 62 Figure 10: Real Exchange Rates: Price of Nontrade 63 & 65 Goods in Terms of Various Traded Goods Figure 11: Decomposition of the Price of Oil 67 Figure 12: Decomposition of the Price of Gas 69 Figure 13: Production of Oil under Alternative Reserve 72 Scenarios for the Oil and Gas Sector Figure 14: Capital in the Traded Sector under Alternative Reserve Scenarios 73 Figure 15: Price of Gas under Alternative Reserve Scenarios 74 Figure 16: Alternative Oil Price Scenarios After Year 2000 79 Figure 17: Tilted and Permanently Lower Price Scenarios 80 Figure 18: Oil Output under Different Price Scenarios 81 After Year 2000 Figure 19: Oil Output under Different Price Scenarios 82 Figure 20: Real Consumption under Different Price Scenarios 83 Figure 21: Capital Stock in the Traded Sector under 85 Different Price Scenarios Figure 22: Price of Gas under Different Price Scenarios 86 After Year 2000 Figure 23: Price of Gas Under Different Price Scenarios 87 Figure 24: Real Consumption under Different Scenarios for 93 the Efficiency in the Domestic Use of Energy Figure 25: Stock of Capital in the Traded Sector under 94 Different Scenarios for Energy Efficiency Figure 26: Composition of Wealth, Base Run 100 Figure 27: Oil Revenues and Exogenous Transfers 101 as a Share of Total Income Figure 28: Fraction of the Labor Force Employed in the 104 T-Sector Figure 29: Cost and Benefits of Two Hypothetical 108 Projects Figure 30: ARI, *ARI and CRI in the Base Run .111 Figure 31: Composition of Wealth, Low Reserve Case 114 Figure 32: Share of Oil, Oil Revenues and Exogenous 115 Transfers in Total Income, Low Reserve Case Figure 33: Oil Extraction Path in the Base Run and in the Favorable Price Scenario 119 Figure 34: Oil Extraction in the Base Run and Low Time Preference Case 120 Figure 35: ARI and *RI in Base Run and Low Time Preference Scenario 123 Figure 36: Various Discount Rates in the Base Run with Optimal Foreign Borrowing 128 Figure 37: Different Oil Price Scenarios 130 Figure 38: The Real Exchange Rate (NT/FT) under Different Oil Price Scenarios 133 Figure 39: Optimal Borrowing Levels (Commmercial Debt) Under Different Oil Price Scenarios 134 Tables Table 1: Tradables, Non-tradables and "Exogenous Resources 14 Table 2: Exogenous Variable Estimates for the Base Case Scenario 50 Table 3: Average Annual Sectoral Growth Rates: 57 The Base-Case Table 4: Growth of Macroeconomic Aggregates: The Base Case 60 Table 5: Alternative Scenarios for Oil and Gas Reserves 70 Table 6: The Impact of Alternative Oil Price Scenarios on the Growth Macroeconomic Aggregates 76 Table 7: Alternative Scenarios for the Real Price of Oil 78 Table 8: The Impact of Alternative Reserve Scenarios on the Growth of Macroeconomic Aggregates 89 Table 9: Alternative Scenarios for Domestic-Energy Efficiency 91 Table 10: The Impact of Different Energy Efficiency Scenarios on production structure, energy and use and GDP growth rates 91 Table 11: GDP Growth in the base run with optimal capital accumulation 105 Table 12: Growth Rate of GDP and Consumption in Base Run and in runs with alternative utility 122 function parameters Table 13: Consumption Growth Rate with and without Foreign Borrowing 126 Table 14: Optimal Current Account Deficits under Different Oil Price Scenarios 132 1. Introduction: The Blasic Questions This study presents a model of the Egyptian economy that focuses on some of the most import-ant aspects of the trade strategy and investment planning related choices that Egypt faces over the next two decades. The approach is that of optimal growth theory applied to the specific challenge of long-term planning in Egypt. But the issues addressed are of more general interest and the approach taken should be useful in the context of other efforts of long-term development planning. The questions raised in this study reappear in different contexts and are of importance whenever the need for substantive "structural change" is anticipated. To focus on the essentials one has to discard many details, aggregate and oversimplify. We hope that the optimizing model presented in this study can provide the economic policy maker with a framework which makes reality more transparent by bringing out the most essential features of the complex choices that must be made. Once the most basic choices are brought into focus, the aggregate model should be useful as a unifying framework for more complete and detailed studies of selected economy-wide and sectoral issues. In particular, the resolutely long run and optimizing nature of the model should yield results that can be incorporated into more disaggregated medium-term consistency models. Similarly, the aggregate shadow prices derived from the dual solution of an optimizing model can become a useful input into more detailed sectoral estimates of shadow prices, by providing the crucial link between long-run optimal growth considerations and the detailed, but by necessity more partial, procedures of sector studies and project appraisal. - 2 - The most fundamental dynamic choices faced by an economy can be summarized in the form of two basic questions: (i) how much to invest, and (ii) where to invest. These are the two basic issues addressed in this study. The answers are by no means obvious. Much depends on the returns that society can get on additional investment, on the consumption requirements that must be satisfied, on society's long-term willingness to save and on the terms on which domestic resources can be complemented by foreign capital. Conditions vary from country to country. However, the question of how far to push the investment effort remains as an important challenge to the policy maker. There will never be an ideal answer to this question. To a considerable extent it depends on value judgements relating to intertemporal income distribution on which reasonable people can simply disagree. Nevertheless, every society does make choices relating to aggregate investments and it is imporant to analyze the key determinants of these choices. Trying to determine an optimal composition of investment leads to an even more complex set of issues. Should the emphasis be on housing, schools, social infrastructure, or do the export producing and import competing commodity producing sectors deserve top priority? In Egypt, these optimal saving and investment questions are further complicated by two important characteristics of the Egyptian economy: (i) a substantial part of Egypt's wealth lies under the ground in the form of oil and gas reserves. These oil and gas reserves are currently generating a substantial flow of income (close to one fourth of GDP), but they represent a stock of finite, non-reproducible wealth that cannot be renewed and expanded like physical plant and infrastructure. Moreover, there is a substantial amount of uncertainty attached to the quantity of reserves and to the price at which oil can be sold - 3 - on the world market; (ii) in recent years, Egypt has benefitted from an unusually large flow of foreign savings in the form of workers remittances and concessional foreign aid. Sustained growth in these flows is doubtful. These are two key features of the resource base in the early 1980's, and it is against this background that an attempt must be made to evaluate the choices open for the future. The economic model that will support the analysis throughout this study is an aggregative, optimizing, long-term growth model that distinguishes three sectors: (i) a sector producing tradeable goods (exportable commodities and import substitutes); (ii) a sector producing non-tradeable goods; and (iii) the oil and gas sector. The basic objective of the policy-makers is assumed to be the maximization of a discounted stream of domestic consumption over time. There is the usual trade-off: restraining consumption in the short-run allows more saving and, provided there are positive real returns on saving, more consumption in the long-run. There is a balance of payments constraint: in any given period the value of imports cannot exceed the sum of tradable commodity exports, oil exports, foreign transfers and net borrowings from abroad. There are also physical and technological constraints: increases in domestic production are limited by the amount of physical capital that can be installed, the size of the productive labor force and the technological improvements that can be achieved. Finally, there is an exhaustible resource constraint: the cumulative production of oil and gas cannot exceed available reserves. Within this framework the key strategic policy choices facing the country over the next 20 to 30 years can be summarized as follows: (a) How much should Egypt save? Should the savings effort increase over time, decrease or remain constant? - 4 - (b) In what form should Egypt save and invest? The country's total wealth can be decomposed into: (i) real capital assets (factories, infrastructure, etc. 1'); (ii) wealth in the form of oil and gas underground; and (iii) financial claims on the outside world (reserves in the Central Bank, net claims on foreigners). The process of saving and investing can add or substract to these three types of wealth. In view of the alternatives, what constitutes an optimal distribution of saving? Note in this context that oil extraction "decumulates" the second and foreign borrowing the third type of wealth. (c) Given a certain overall rate of "physical" investment (alternative (i) in (b) above), how should capital accumulation be allocated between the tradable, the non-tradable and the oil and gas producing sectors? What factors determine an optimal allocation? While we shall discuss specific "optimal" growth paths, the primary objective is to analyze the interdependence between the various decisions that must be made and the sensitivity of optimal policy rules to key parameters, such as the degree to which Egypt is willing to take a "long-run view", the estimated magnitude of oil and gas reserves, and the return Egypt can hope to get on physical investment. A second important objective of the modelling work is to provide insights into the price and incentive structure that should accompany an optimal savings and investment strategy. With a three sector model not much can be said, of course, on the microeconomic structure of shadow prices. But some of the aggregate parameters that are usually given exogenously to project - 5 - planners can be analyzed in a dynamic general equilibrium framework with the kind of optimizing model presented below. In particular, the following questions arise: (d) What is a reasonable time path for the social rate of discount? The degree of time preference and risk aversion remain fundamental value judgements that cannot be avoided, but the sensitivity of the social discount rate to long-run macroeconomic trends can be explored systematically with an applied optimal growth model. (e) How should the real exchange rate behave to support an optimal growth path? This "relative price" is important for investment planning and project appraisal procedures for it determines, loosely speaking, how costs and benefits that can easily be expressed in terms of foreign exchange should be valued in comparison to costs and benefits that are essentially of a domestic nature. (f) The third important macroeconomic price is the real wage. It too may vary over time, and its time path is of course intimately linked to the time paths of the social discount rate and the real exchange rate. These are fundamental, strategic questions that can be analyzed quite naturally with a numerical optimal growth model. While finding correct answers is difficult and will always contain a large dose of "arbitrary" judgement, it is perhaps worth stressing that in one way or another decisions have to be, and are being, made on all these issues. When a project is rejected on the grounds that it does not yield an internal rate of return of, - 6 - say, 10 percent, this decision reflects a judgement on the appropriate discount rate. When any particular exchange rate is used to convert, say, projected domestic construction costs to a dollar numeraire, there is implicit in the calculation a judgement on the future path of the real exchange rate. What this study attempts to do is to analyze these issues in an explicit dynamic and optimizing framework to bring out the interdependence of the key variables that affect the process of investment planning and to clarify the strategic choices that we made in any economy. 2. The Initial Conditions: Egypt at the Beginning of the 1980's The 1970's witnessed a radical transformation of the structure of Egypt's resource base. In terms of overall growth in resources available to the economy, the last decade has been a period of rapid expansion. Defining total resources as the sum of gross domestic product and net imports (domestic resources plus net resources from abroad), one gets an average annual growth rate of about 12 percent over the 1974-1981 period. This means that real resources available to the economy more than doubled in seven years. While total resources increased at a remarkable rate, the composition of these resources underwent a radical change. Four major developments led both to the rapid growth in Egypt's resources and to the change in their composition: (1) The petroleum (and gas) sector emerged as the financially dominant sector of the economy with production rising from less than 8 million tons in 1974 (worth about $800 million at world prices) to more than 32 million tons (worth almost $8 billion) in 1981. In - 7 - constant dollars, this represents an average annual growth rate of about 27 percent. (ii) Remittances from workers abroad increased from $190 million to $2,800 million in 1981, a dramatic increase averaging to about 32 percent a year in constant dollars. (iii) The Suez Canal was re-opened and enlarged and Canal earnings reached about $900 million in 1981. (iv) Finally, direct foreign investment and net MLT lending increased significantly, particularly if one distinguishes between the pure balance of payments supporting grants and loans of the mid 1970s and the more autonomous development project directed capital inflows of the 1978-1981 period. Let us define "exogenous resources", RE, as the sum of Egypt's share of petroleum output valued at world prices (deducting the foreign companies share), workers- remittances, Suez Canal earnings and net foreign capital inflows: Egypt's Share Workers' Suez Net Foreign RE = of Petroleum + Remittances + Canal + Capital & Gas Output Earnings Inflows By calling these resources "exogenous" we do not mean to imply that their magnitude is totally unresponsive to domestic economic policy. They are "exogenous" more in the sense of having very little to do with the product- ivity of Egypt's domestic labor force that is overwhelmingly employed in agri- culture, industry and services. It is also true that their magnitude is very much dependent on factors beyond the policy makers control (world oil prices, the political situation in the countries hosting Egyptian workers, etc.). - 8 - Figure 1 plots the ratio of RE to total resources available to the Egyptian economy for the 1974-1982 period.The picture tells a striking story. The ratio of exogenous to total resources went from about 6 percent to 45 percent in the early 1980's. The most dramatic part of the increase occurred between 1976 and 1980. After 1980 the ratio starts showing a decline. Crucial questions facing Egyptian policy makers and planners relate to the impact RE has on the domestic economy and its most likely behavior in the future. Will RE continue to grow. in absolute terms? Will it grow or decline as a percentage of total resources? What implications does an alternative growth path of RE have for overall economic growth? For investment allocation? For exchange rate policy? For project selection? These are some of the difficult questions that lie behind the policy debate in Egypt and that will be discuss below. An important challenge for the economist is to try to derive from the macroeconomic analysis operational implications for investment planning and project appraisal. Tradables, Non-Tradables and the Real Exchange Rate: While there are some significant differences between oil revenues, remittances, Suez Canal earnings and foreign aid flows, they all basically represent a transfer the magnitude of which bears very little relationship to productivity, wages and resource growth in the non-oil domestic economy. Moreover, the transfer accrues in the form of foreign exchange and, therefore, implies immediate command over resources that are tradable on the world market. An increase in exogenous foreign exchange resources implies an immediate increase in the capacity of a country to consume easily traded ,t 20CI lost 0Kt U61 GM61 U6t 9M6t .S61 VSI 54t 0 l.' C) a saoanos9le T3oI0 UT saalosa SnoueoxR Jo ao saiS -6- - 10 - commodities and services. There are some commodities and services, however, that are not easily traded. For such commodities, availability of foreign exchange does not signify an immediate increase in potential consumption. Housing is one example of such relatively non-tradable commodities. In contrast to tradables such as clothing or food, it is not possible to increase domestic consumption of housing by increasing imports or reducing exports: housing is non-tradable in the sense that transport costs are simply too high to make significant trade a viable alternative. Substantial portions of an economy s output have this characteristic of non-tradability (many services, construction, electricity, transport and communication networks are all relatively non-tradable). A rapid increase in exogenous foreign exchange flows does not allow an equally rapid increase in the consumption of such items, since their supply can only be increased by raising domestic production. In contrast, the supply of a tradable commodity can be increased quickly by importing more or exporting less. The impact of a sudden increase of exogenous foreign exchange flows on the structure of an economy has traditionally been described with the help of a simple two sector model (see Buiter and Purvis (1982), Corden and Neary (1982) and van Wijnbergen (1980, 1981)). Figure 2 below reproduces the by now well established stylized representation of the impact effect of a foreign transfer. The economy has two sectors (tradables and non-tradables). In the absence of any other resources, the production possiblity frontier AB is also the consumption possibility frontier and the economy will produce and consume at P=C. The slope of the produ lion possibility frontier at P=C defines the equilibrium relative price of non-tradables in terms of tradables, a price that is defined as the real exchange rate. - 11 - Figure 2 The Impact Effect of an Exogenous Foreign Resource Transfer sooa5 e , A a-0) X < yi(a1KIt + (1 - a1)(Ltg1 ) ) (1) - P2 -p2 1/2 (V >0) 2t < Y2(2 2K + (1 - a2)(L g) )(2) 2t2t t 2 2t 2 2tg2 (V 3t > ) X3 f3(K 3t, R3t' t) 3 (V4t >0) X4 f4(K4t, t) (4) Here, Xit, Kit and Lit refer to output, capital stock and employment levels in the four sectors, while R3t is the stock of oil reserves at the beginning of the period. In the tradable and non-tradable sectors, output is constrained by neo-classical production functions that allow some substitution between capital and labor, with the substitution elasticity given by a KL = 1/(1 + P There is labor augmenting technical oi progress reflected in the growth factors gi. Production constraints in the oil and gas sectors, f3 and f4, are of a different nature. For oil, the cost of production depends critically on the relationship between the stock of reserves and the flow of extraction. Given a certain number of fields and certain levels of reserves, there is an upper limit on how much oil or gas can be produced without endangering the long-term productiveness of the fields. Nevertheless, with the aid of secondary and tertiary recovery techniques, oil production can be maintained at high levels even in old fields with small reserve to production ratios. The model captures this relationship by linking cost recovery share of the - 23 - foreign oil companies to the production-reserve constellation. Capital expenditures in the oil sector are overwhelmingly financed by the oil companies. The companies then recover their costs "in kind" by getting a certain amount of cost recovery oil in addition to their contractual share in profit oil. Attempts to push production beyond "normal" levels will run into diminishing returns to the Egyptian economy reflected in a steeply increasing cost recovery share. 3/ In the gas sector, the same kind of relationship does in fact exist between reserve levels and production flows but there is somewhat greater flexibility. There are also important institutional differences. The foreign companies role is minor and domestic investment is required to expand productive capacity. These considerations as well as computational constraints have led us to drop the variable production costs specification for gas and use a specification based on a simple capital output ratio plus quantity constraints on annual output 4/ 3.2 Material Balances The second group of constraints reflect the intermediate input requirements of production and are arranged in the form of material balance equations setting total use of a commodity equal or below total available supply from production and trade. We have: (P lt > ) Dit + Eit + sI Yt <^X lt (5) (P6t 0) C + E a < Y6(a6D1t + (1 - a )M (6) 6t 6t i 6~~jXj t6i - 24 - (P 0) c + Ea X + s Y °) C2t + ; 2jXjt 2+ t < X2t (P 3t 0 ) E3t+ a3jXjt + E;a 4Xjt -x4t) <-3t (8) (P4t ' ) x4t ' E a4JXjt 9 j~~~~~ The first three constraints refer to the domestic non-oil and gas economy. The supply of domestically produced tradables, Xit must be greater than what is required for domestic consumption or intermediate use, Dit, for investment, sjYt (when s, is a fixed ratio) and what is exported, Elt. The amount Dit then combines with imports Mit to supply the composite tradable commodity that is either consumed, C6t, or used as an intermediate input E a6t Xjt, in equation (6). The material balance for non-tradables is simpler: total supply, X2t, must suffice to meet consumption demand, C2t' intermediate demand, E a2jXjtp and investment demand, s2Yt. Turning to the oil and gas sector, we distinguish between two kinds of (intermediate) demands: (i) a demand that can only be met by petroleum, reflected in the input coefficients a3j, and (ii) a demand that can be met by petroleum or gas, reflected in the input coefficients a4J. For example, electricity can be generated by gas or oil; on the other hand, we assume that cars and trucks will continue to require petroleum products. Material balance equation (8) states that the total supply of petroleum, X3t, must be greater than total demand composed of exports, E3t and domestic demand, where the latter decomposes into a portion that is petroleum specific, E a3jXjtp and a portion that could be met partially or entirely by gas, E a4JXjt' To - 25 - the extent that the latter demand is in fact covered by gas output, X4t, domestic demand for petroleum declines. For example, if gas has totally replaced oil wherever it can, we would have E a 4jXjt = X4t and the demand for oil would consist only of exports and non-substitutable uses. Of course E3t could turn negative if domestic use outruns domestic production, in which case Egypt would become a net importer of oil. In addition to the material balance constraint there is also, however, a special constraint on the supply of gas. The output of gas is constrained by domestic demand through equation (9): given the probable range of Egypt's reserves and the market conditions in the region we do not consider exports of LNG to be a viable alternative, so that domestic demand in the sense described above (all possible uses where gas can replace oil), acts as a constraint on the total production of gas. 3.3. The Balance of Payments The balance of payments reflects society's budget constraint: the value of imports cannot exceed the sum of export earnings, factor receipts, and net foreign capital inflows. (P5t > °) PW14 M + PW M

itltlt Ot Ot= °) Kit < (1 - di)K it_ + Yi)t-1 (11) where di is the depreciation rate. In practice, we shall use two-year periods when running the model, so that we in effect assume that once physical investment has started, the average gestation lag is two years. Total resources devoted to capital accumulation are denoted by Yt so that E Yit 4 Yt. We experiment with both exogenous and endogenous total i investment. Moreover, we also impose an absorptive capacity constraint that introduces diminishing returns to aggregate investment beyond what is already implicit in the production functions. There is a critical investment - 28 - level, Yt , itself a function of time and previous levels of investment, beyond which an increasing part of the "excess" investment is wasted and does not yield effective new capital. Formally: *4 Yt if yt - Yti it-l 4 J~~~~~~~~~=l (PIt 2 ) i=1 Y it 4 (12) i=l g(Ytp Y) if t > ii it- t t j.=l t- The specific functional form of the functions Yt and g is given in Appendix A. This specification is designed to capture the difficulties inherent in trying to raise investment very rapidly beyond customary levels. These difficulties have been observed in many countries and are due to a multitude of causes ranging from limited port and transport capacity to lack of specific complementary factors such as sufficient management or specially skilled labor. To try and capture all this with a single absorptive capacity constraint is, of course, an imperfect solution. But the formulation adopted reflects an important aspect of reality that is central to the problems of investment planning and that should not be neglected. 3.5 Resource Exhaustion Constraints The last two constraints reflect the fact that petroleum and gas production over the years is ultimately limited by total recoverable reserves. It is this limit on cumulative production that distinguishes oil and gas from other industrial and agricultural commodities. Oil and gas are exhaustible resources that cannot be reproduced ad infinitum. In this context it is important to distinguish between "proven" or "discovered" reserves and - 29 - reserves that will be discovered in the future. The distinction is important because production costs and therefore, indirectly, production flows, are determined not by total reserves that will eventually be discovered, but by reserves in existing, known oil and gas fields. We have: (S3,t >-°) R3,t 3,t-1 ^ 3,t X3,t (13) (S4,t 2°) R4,t 4,t-1 ^ 4,t X4,t (14) where R and D refer to reserves stocks and discovery flows. Initial reserves and discovery flows ares exogenous. The magnitude of future discoveries, is of course subject to great uncertainty. But it is precisely by exploring the impact of different reserve scenarios that we can hope to gain a better understanding of the role of exhaustible resources in the determination of optimal levels and allocations of investment, and in the determination of shadow prices that wou:Ld support optimal strategies. 3.6 The Objective Function Subject to the constraints described above, the model tries to maximize a social welfare function which is taken to reflect the long-term development objectives of Egyptian society. In defining the social welfare function, several factors are taken into account. First, when society is faced with the choice between an additional unit of consumption today or the same additional unit tomorrow, the first option is preferable: in technical terms the pure rate of time preference is positive. This is taken into account by discounting the utility of future consumption by a factor 6 for - 30 - each period in the future. Second, society cares about the level of consump- tion of the representative family of each generation, and as society gets richer the value of an additional unit of consumption declines. Utility U is therefore a declining function of per capita consumption. On the other hand, it does matter how many families are consuming at that level,5/ so that the index of per capita consumption is multiplied by total population. Third, the model specifies a constant rate of substitution between tradables and non- tradables, where tradables are in fact a .composite" of domestically produced and imported goods. Thus we have in effect a nested two-level CES utility function. At the first level imports and domestic tradables aggregate into the composite tradable consumption good. At the second level, this composite commodity and the non-tradable commodity yield a CES index of utility. Thus we have: T - W = z L t U (15) t=O t where ut= y (a C*_Pu + (1 - a )C* P)/Pu u u 6t u 2t 6= the discount or pure time preference factor, which equals the inverse of (1 + the social discount rate). I= the consumption elasticity of social welfare (1 + elasticity of the marginal utility of real income). C*it = Cit/Lt = consumption per capita of good i, i = 2,6. - 31 - Finally, while our model stops at t=T , time does of course continue and the income of future generations is also valuable. In fact it could be argued that we should consider an infinite horizon, since by taking a positive rate of time preference we are already given enough of an advantage to present and near future generations. But to obtain numerical results we must, inevitably, work with a finite number of periods T . That is not really a problem with exogenous investment if T is not too small, since in that case the only instrument for intertemporal arbitrage is the oil/gas extraction policy, and all reserves would be exhausted by the end of that period anyway. But with endogenous investment, if no constraint or valuation is attached to the terminal stocks, a substantial part, or maybe even all, of the capital stock will be "eaten up" in the final periods, which is clearly not a reasonable solut:Lon. The way this problem is handled in the endogenous investment case is by exploiting the Turnpike property of optimal growth models: when the planning horizon is long enough the economy will spend most of its time growing at a constant rate, with the capital stocks in each sector equal to its steady state values.6/ That suggests as a good approximation to the infinite horizon problem, to constraint the terminal capital stocks to be not smaller than those steady state values. That is, * Ki,T KiT i = 1,2 (16) with K = steady state capital stock in the i sector. iT Notice that the "steady state" capital stock is a function of time because of technological progress; if capital is redefined in "efficiency units" that problem goes away. - 32 - Of course this method only works if the planning horizon is long enough to allow the economy to get close to the turnpike values. We therefore work with a longer horizon (30 periods or 60 years) in the endogenous investment case than in the fixed investment case, where the horizon is 20 periods or 40 years. This extension of T was chosen after some experimentation with different values; when extending T further ceases to influence the solution in the first 30 years significantly, T is considered long enough. A final point one should notice is that changes in oil price scenarios, energy I/O coefficients, utility function parameters, etc. will change those steady state values Ki . Whenever such experiments were performed, the steady state values were adjusted accordingly. 4. Valuation on the Optimal Path One of the attractions of working with an optimizing model is that it generates as part of the optimal solution a set of dual variables that measure the contribution to the objective function of relaxing the constraints of the model. Most of these variables can readily be interpreted as shadow prices for the different goods or factors. 4.1 Conditions for Static Efficiency Associated with each of the "material balance" equations in the primal problem, there is a dual variable (Pit) which is the shadow price of the corresponding good. There are shadow prices for domestically produced traded goods (Plt), nontraded goods (P2t)' oil (P3t), foreign exchange or imported goods (P5t), and the "composite" good defined from produced and - 33 - imported traded goods (P6td* We normalize the price of imports to one, so that P5t measures the value of foreign exchange in the model. The first two optimality conditions set the domestic shadow price of oil and of traded goods equal to the marginal revenue at border (international prices) multiplied by the value of foreign exchange: P3t P p5t W3t ( 1) Plt = p5t(1-1/ n pTe (2) where n is the elasticiLty of demand for exports. 7/ If Natural Gas were an internationally traded commodity, its shadow price should be its international price (adjusted for all extra costs incurred to make it tradeable). When international trade is not a relevant alternative (like in Egypt, where the size of gas reserves do not justify a liquefaction plant) gas should be priced at its fuel price equivalent if on the margin it substitutes perfectly for fuel oil. If, however, gas use is restricted by substitution possibilities and the level of demand at fuel oil parity price, a wedge may open up between the price of gas and its fuel oil equivalent. In this exercise we made the admittedly extreme assumption that gas is a perfect substitute up to 50%. of total energy use, and cannot substitute at all above that cut-off rate. The cut-off rate increases to 60% over the first 10 years. The variable P4t (the shadow price associated with the equation defining maximum demand for gas in any period) measures the gap between both prices: - 34 - PGt = p3t p4t (3a) so that when the use of gas is demand constrained, P4t will be positive and the shadow price of gas will be below its fuel oil equivalent. Consider first the case where P4t is larger than zero but where there are no supply constraints on production and distribution of natural gas (Figure 4a). If there is a finite amount of gas that will be exhausted, say, at time T , the price has only two components, extraction costs MCt and rent 4t~~~~~~~~~~~~~~~~~ to reserves, 4 PGt =MCt + S4t (3b) = X Pjt aj4 + Q4t b4 +54t Efficient exploitation of the natural gas resource will lead to a rental component rising at the relevant rate of discount with the price of gas reaching its fuel oil equivalent at the moment of exhaustion (cf. Section 4.3 below). Which discount rate is relevant will depend on the numeraire in which the price of gas is expressed if there are gradual relative price changes in the rest of the economy. This price path will provide a floor for the price of gas when supply constraints are operative. In that case the shadow price of gas picks _ 35 _ Figure 4a: Pricing of naturaL gas when use is demand constrained Price fuel oil equivalent P eprice rent toI gas reserves v Unit extraction costs T time - 36 - up a third component, rent to the factors constraining supply. The fuel oil price equivalent will provide a ceiling. Figure 4b is an attempt to clarify matters. At QA the "supply price" of fuel oil intersects with the inverse demand curve for energy (exclusive of the "non-substitutable" part of gas). If supply or distribution constraints limit gas output below QA ' substitution possibilities on the demand side are not yet exhausted and the shadow price of gas equals its fuel price equivalent, P oil (P3t in our model). If supply constraints limit output to any value above QB where the inverse demand curve intersects with the price floor set by rent to reserves and unit extraction costs, these supply constraints are not operative. Accordingly, they earn zero rent in that case and we are back in the case of Figure 4a where the shadow price of gas equals unit extraction costs plus rent to reserves. If constraints on production and distribution of natural gas limit output to an intermediate value, say Qc 9 the price of gas will be in between its ceiling P3t and its floor MCt + S4t and the supply constraint accordingly picks up a rent equal to Q - MC - S4 4t =P gt t 4t (4) where Pgt = pE C) . PE( ) is the inverse demand curve for energy exclusive of the part for which gas is no substitute. - 37 - Notice that QC is an interesting number that could be used in Q4t shadow-pricing projects expanding say distribution facilities (like pipelines). As time goes on, both the demand for gas and the supply constraints will increase; accordirngly the ratio P gt/Poil may go through several phases before it eventually reaches unity (for an example see Figure 12). In the case of oil, instead of using domestic capital, the cost of extraction additional to intermediate inputs is borne by foreign companies, which are compensated through a share in the value of output (a . The corresponding equation is: 3t= jPia3 + Q3t+ Z3t + P5t 3t ( + X3taQ /3X3t) * (5) The production functions have associated dual variables (Vit) which give unit value added per sector. That is: Vit = Pit - E Pjtaj, i = 1,2 (6) These variables also measure the contribution of primary factors to output: it' Fi/Kit - Qit = rental value of Ki (7) Vit' aFi/31 it Wt =shadow wage rate where Fi(Kit, Lit) are the production functions defined in (1) - (2). - 38 - Figure 4b Pricing of natural gas when supply constraints are operative Inverse Demand Curve for Energy, P (Qt) Price E t oil ! c ~~~~~ rent to supply .B rent to gas reserves . ~~ ~~I I I eunit extraction costs - 39 - The last equation already incorporates a further optimality condition, namely that labor is allocated in such a way that its marginal product is equal in all sectors. The rent on capital goods may differ between sectors, as capital once installed cannot be shifted to other uses. There are two more static efficiency conditions: one defines the optimal way of produciag the composite X6 from imports and domestically produced traded goods: Dlt P6 a6 P5t (8) mit 1-6 lit The other shows the opitimal consumption of nontraded goods (C2t) and the traded composite (C6t), defined by the equalization of the marginal rate of substitution in consumption and production: 1+ Pt (C 2tI/C ) = (1-Y2)P 6t/y2p2t (9) In both equations Pi reflect the substitution possibilities (as Pi approach its lower bound, -1, substitution becomes easier), and a6 and 2 are "share" parameters that are obtained from base year data. 4.2 The Social Discount Rate The price variables discussed above are useful for project evaluation as they give the social value of the corresponding goods and/or factors. At any point in time only relative prices matter, so that any numeraire can be used to define prices. It is important to remember, however, that when evaluating projects with costs - 40 - and benefits spread over several periods, the numeraire chosen determines the appropriate discount factor. Our model provides a natural generalization of the one-consumption good model (where that consumption good is the obvious numeraire): total utility is a homothetic index of consumption of both goods, so that we can define an exact price index for real consumption to use as numeraire. Let us consider briefly how the appropriate discount rate is obtained in models with only one consumption good-8. Suppose a project uses one unit of that good today and produces b units tomorrow. By definition, social objectives are defined by the function: W = EL C ^- (10) t t to be maximized among all feasible paths of per capita consumption (C t) The change in W when the project is undertaken is (when the project is small) AW E6-tC*c lb t t t By definition bo = -1 and b1 - b. The project should be approved, if and only if it results in an increased value for the objective function, i.e. when: _ C*~l + c6 lC"_1b > 0 (11) That is, future benefits must be discounted at the Consumption Rate of Interest (CRI) defined by: - 41 - CRI -g I (12) C where g = C /C is the growth factor of real consumption. The social discount rate is larger the larger is 6 , as the future is considered relatively less valuable. Also, with declining marginal value of per capita consumption (i.e. f < 1), the more consumption per capita is expected to grow the more the future should be discounted: if consumption is expected to grow, the marginal utility of consumption decreases and consumption in the future will be less valuable for that reason. The rate of growth of the population drops out of the expression for the consumption rate of interest due to the specification of social objectives in (10). An alternative specification used in the literature does not weight per capita consumption by the size of the population. Practically this has an impact identical to having a smaller time preference factor, so that a specif- ication that maximizes the sum of unweighted per capita consumption with a high pure time preference factor, is equivalent to a specification maximizing the sum of population weighted per capita consumption with a low pure time preference factor. Whichever specification one choses, it is necessary to test the sensitivity of the growth path to variations in the time preference factor. As noted above, the assumption of homothetic preferences allows us to interpret the model as having only one consumption good. The CRI can be obtained as in (12), diefining C* as total real consumption (i.e. utility, Ut, which is equal to the value of consumption divided by the exact price index defined by the function U(C2t, C6t)). In Chapter 5 we will also compare the - 42 - CRS with discount factors in terms of other numeraires. 4.3 Dynamic Optimality Conditions The optimal solution also generates values of all stocks in the model: capital in each sector and reserve of oil and gas. The prices associated with the capital updating equations, PKit, must satisfy two optimality conditions: (i) they must equal the discounted value of the future contribution of Kit to real consumption over the planning period, plus their terminal values. (ii) they must also equal the cost of reproduction of capital, with allowances for lags in the maturation of investment and constraints on the quantity of investment that can be usefully undertaken during a given period (i.e. absorptive capacity). If we define all variables in terms of real consumption and assume, to simplify notation, a constant social discount factor, CRI, we must have: PK = T-Et Qist+sCRI -s +dQi t T (13) it ,t+s +-i$TCR where Qi,T is the terminal value of Ki, and Qi,t is the marginal productivity of capital in the ith sector, defined in (5) and (7). Capital goods are assumed to be composed of nontraded goods, domestically produced traded goods and imports, in fixed proportions. The cost of new capital goods is therefore: PKt =sPit + s2P2t + s5P5t (14) - 43 - There are two factors mediating the connection between the value and the cost of reproduction of capLtal. First, investment takes time in becoming fully productive. In our case the maturation lag is just one period, so that: PIt = PKit/CRIt i = 1, 2, 3 (15) The fact that (15) has to hold for all three capital goods reflects our assumption that alL capital goods are homogeneous before being committed to any particular sector. A second reason why PIt is not necessarily equal to the cost of reproduction of capita:L goods is the presence of absorptive capacity constraints: resources devoted to investment are less efficient in generating production capacity beyond a certain level (due to lack of qualified workers, bottlenecks in infrastructure, etc.). In general we must have PIt > PKt (16) with strict inequality when absorption capacity is a binding constraint. The actual connection between PIt and PKt in the model is also affected by the impact of current investment on relaxation of future capacity constraints (see Appendix A). The value of the remaining two stocks in our model, oil and gas reserves, are more easLly described, as they can not can be reproduced nor have any use besides aLlowing future extraction. An efficient allocation of social wealth between all available assets will lead to equal rates of return on all of them. For natural gas the - 44 - only component of the rate of return is capital gains on reserves, so the rental value of gas follows (13a). s4t s4t+l / CRIt+l (13a) The value of gas reserves increases at the social discount rate: when they equal the fuel oil equivalent minus marginal extraction costs, reserves will be exhausted. For oil the marginal return on keeping one more unit in the ground has two components because of the assumption that extraction costs depends on the level of reserves. If Q3t+l is the reduction in marginal extraction costs during period t+l (incurred at the beginning of the period) by having one more unit in the ground at the beginning of period t+l 9/, the rental value of oil must follow (13b): s3t= Q3,t+l + S3t+1/CRIt+1 (13b) The rental value of oil reserves may decline, contrary to the case of gas, as their increasing scarcity pushes up marginal extraction costs. It is possible that the rent becomes zero at a positive reserve level, indicating that extraction costs have become too high to make extraction worthwhile: economic exhaustion may come before physical exhaustion if extraction costs depend on remaining reserves. - 45 - 4.4 Changes in relative prices In section 1 we discussed how changes in the level of Exogenous Resources were expected to affect relative prices in a simple traded-nontraded model: increasing Exogenous Resources results in an appreciating real exchange rate (i.e. an increase in the relative price of nontraded goods), and a real depreciation should be associated with declining inflows, to facilitate substitution, away from traded goods in consumption and toward them in production. As we noted there, the magnitude of the adjustment required depends on "how different" domestic tradeables are from domestic non- tradeables, in the way they combine factors of production if production characteristics were identical between the two sectors, the output of tradeables could expand without the need for a rise in their relative price (this of course assumes; perfect mobility, we in fact postulate that capital goods are mobile ex ante, but sector specific export, but given the high rates of investment and forward looking nature of optimizing models, the constraints on mobility is almost never binding). However, our description of the Egyptian economy is considerably more complicated than the simple two goods model, making the dynamic behaviour and indeed even the definition of "the" real exchange rate also more complex. In the first place, we specify the need for changes in the terms of trade as exports expandL. That makes the real exchange rate relevant for consumption decisions dLifferent from the one determining the relative productivity of domestic factors in the traded and home goods sectors (since importables and exportaLbles enter in different proportions in the consumption and production baskets; in fact we assume that only imperfect substitutes of imported goods are produced domestically). As the economy adjusts to lower - 46 - exogenous inflows by producing more exports, the price of imports (the foreign good) rises relative to the prices of both domestic goods (tradeables and non- tradeables). This terms of trade effect forces a "devaluation" not in the sense of a rise in the price of domestically produced tradeables relative to the non-tradeables, but in the sense of a rise on the price of foreign goods compared to domestic goods. A second factor that complicates the dynamics of relative prices is that, in addition to a reduction of exogenous inflows, which other things equal, leads to a rise in the relative price of tradeables compared to non- tradeables, there is a sustained increase over the planning period in the overall capital/labor ratio of the economy and in the price of energy. That tends to reduce, for a given composition of output, the price of the sector which is relatively more capital intensive and less energy intensive, respectively. Stylized facts about the Egyptian economy suggest that the nontraded sector is more energy intensive (since it includes such energy users as transportation, electricity) and less capital intensive (because of the inclusion of services), than the sector producing tradable goods. As a result, both capital accumulation and the increasing cost of energy tend to counteract the tendency toward an increase in the relative price of tradeables, associated with the reallocation of resources required to reduce the dependence on Exogenous Resources. - 47 - 5. The Nature of OptimaL Growth Paths: Structural Adjustment in a Long-Run Perspective 5.1 Experiments with fixed total investment Once an economic model has been constructed and can be operated, it becomes a kind of laboratory with the help of which alternative scenarios can be explored, different views of the future can be contrasted and different strategies can be evaluated. The experiments reported below illustrate some of the analytical features described in the preceeding sections. They were designed to contribute ito a better understanding of some of the basic characteristics of Egypt-s future growth and are designed to help evaluate the most important policy options. For the first set of experiments the growth of physical investment is kept approximately constant and the trade balance is fixed exogenously. This assumes that total investment escapes the control of the policy maker, that Egypt cannot or will not use net foreign borrowing as a significant policy variable and that: workers remittances and Suez Canal earnings grow at some exogenous predetermined rate. Given these constraints, there are two major issues that the optimizing model can explore: the allocation of investment between tradables and non-tradables and the optimal pace of oil and gas extraction. Note that in so far as oil and gas extraction require investment in the oil and gas sector, investment allocation also determines the pace of extraction, so that experiment set A can be defined as focusing on resource allocation given exogenously fixed total streams of investment and foreign capital and an exogenously given profile of oil and gas reserves. - 48 - 5.1.1 A Base-Case Scenario To give us a basis against which to test the sensitivity of the model to variations in exogenous variables and parameter estimates, it is useful to start by exploring a base case scenario which can then be used as a kind of reference path. A key factor determining the characteristics of any optimal growth path derived from the model is, of course, the view one takes about oil and gas reserves. Clearly, there is a tremendous amount of uncertainty relating to the magnitude of hydrocarbon reserves. Proven reserves in Egypt are low compared to annual production levels and projected growth of domestic use. For oil, a comprehensive recent study estimates proven reserves to be between 2.6 and 3.6 billion barrels in 1980 (360 and 500 million tons). This amounts to no more than 10-14 years worth of production at current rates. Prospects for new discoveries are good, however, and there is a great deal of investment in exploration. Over the past few years this search for oil has produced a steady stream of small but significant discoveries, adding between 20 and 40 million tons on average to proven reserves every year. For natural gas, the situation is similar. The point estimate in 1980 for total proven reserves from established fields and new well finds was 8.3 trillion cubic feet, equivalent to about 200 million tons of oil equivalent. On the other hand, most analysts are quite optimistic about reassessments of existing fields and new discoveries. The same puts probable and possible "guesstimated" undiscovered gas at 23 Tcf or 550 million tons of oil equivalent. Note that this is a kind of upper bound for the next twenty years. The study in fact works with a "proven reserves + half probable + one quarter possible" formula that yields 15.5 Tcf as the most likely reserve scenario for natural gas. - 49 - There are, however, more optimistic estimates, ranging up to 35 Tcf or 840 million tons of oil equLvalent and it is also worth stressing that the study is not concerned with the post-2000 period so that small discoveries that might still be made in the outer years are discounted. In light of the above, we start by focusing on a base-case scenario with the characteristics presented in Table 2. As described in Section 2 above, oil extraction costs are endogenous and depend on the relationship between the rate of extraction and the stock of "already discovered" reserves. While a similar relationship exists for gas, it is weaker and given the relatively more plentiful supply, we are ignoring it in the range of variation explored with these experiments. Production costs in the gas sector are assumed constant and do not depend on the reserve to extraction ratio. Based on these assumptions about hydrocarbon reserves and other parameter values described more fully in Appendix A, the model has been solved numerically to generate a base case optimal growth path and associated shadow prices. We shall first describe in some detail the characteristics of the base case optimal growth path and then test the sensitivity of the most important variables to variations in reserve scenarios, oil prices and foreign exchange inflows. - 50 - Table 2 Exogenous Variable Estimates for the Base Case Scenario Proven Reserves Projected New Total 1981-/1982 Discoveries 1981-2011 Reserves Petroleum Reserves 500 m.t 400 m.t. 900 m.t. Gas Reserves 250 m.toe. 300 m.toe. 550 m.toe. World Price of Crude Oil 2.7 percent real annual growth 1983/84-1989-90 1.8 " " 1989/9 0-2000/ 01 0.5 " " 2001/ 04-20! 0/11 Suez Canal, Remittances 3.5 " " " 1981/82-2010/11 and Net Foreign Capital Economy-wide Investment 5.5 Economy-wide Labor Force 2.7 5.1.2 The Base Case The most important underlying factor explaining the characteristics of the optimal growth path is the behavior of what we have called "exogenous resources", RE, composed of Egypt's share of petroleum output, workers' remittances, Suez Canal earnings and net foreign capital inflows. Figure 5 below describes the ratio of exogenous resources to total resources for the base case scenario. It is clear from Figure 5 that under base case assumptions Egypt will have to adjust to a very substantial transformation in its resource - 51 - base. This R ratio t:o domestic value added goes up to 82% while oil and gas production is expanding and declines gradually there after to about 40% around 2010. Developments in the oil and gas sector are the most important determinant of the behavior of the RE ratio. Figure 6 summarizes output and export paths for oil and gas, given the base case discovery and reserve assumptions. In the base case scenario total oil production continues to grow throughout the 1980's, peaks at the end of the decade at about 52 million tons a year (a little more than one million barrels a day) and then declines quite rapidly. By the year 2010, oil reserves are nearly exhausted. Egypt's share of oil production also broadly follows this pattern with the decline accentu- ated by the decrease of the ratio of Egypt's share to total production, reflecting higher production costs and therefore higher cost recovery transfers to foreign oiL companies as the ratio of output to reserves falls over time. While oil production follows this pattern, natural gas production increases steadily between 1981/1982 and the year 2000. It then levels off at a rate of 20 million tonis of oil equivalent. The dotted line in Figure 6 describes the behavior of Egypt's oil exports derived as the difference between oil and gas production (net of the foreign companies' share) and domestic consumption. Egyptian energy exports, caught between a levelling off in production of oil and gas, increased extraction costs and steady growth in domestic demand, start declining after 1988/89 and, by 2003/2004, the base case projects net energy imports, growing rapidly thereafter. Two points are worth stressing in this context: - 52 - Figure 5 Ratio of Exogenous Resources to Domestic Value Added 0.9- - 0.9 0.8o- - 0.8 0.7 - 0.7 0.6- 0.6 0.5- .- 5.S 0.4- \ 0.4 0.3- -1 , 0.3 1980.0 1990.0 2000.0 2010.0 2020.0 - 53 - (1) The behavior of net exports (or imports) of energy often appears to be the crucial variable. In fact, what matters is the magnitude and rate of change of total oil and gas production. For example, consider the adjustment problems faced by two hypothetical countries. Country A initially produces 80 million tons of oil and gas, exports 40 million and consumes 40 million domestically. Country B produces 40 million, consumes the same 40 million and has no net crude. Five years later, Country A still produces 80 million tons, consumes 60 million and exports 20 million tons. On the other hand, Country B produces 50 million, consumes 60 million tons and has, therefore, become a net importer. Nevertheless, other things equal it is Country A that will have suffered from a greater adjustment problem. (ii) The model projects net oil imports after 2004. These imports can also be interpreted as imports of "alternative energy" in the form, for example, of foreign exchange expenditures on nuclear power plants, nuclear fuels and nuclear technology. Egypt is, in fact, planning to generate nuclear power starting around 1990 and this interpretation of the projected "oil" imports is not inconsistent with the view of the future of Egyptian energy planners, although the base case scenario seems to suggest that these expenditures should not occur before the end of the century. Of course nuclear power requires bulky investmenits so that the path of foreign exchange expenditures would be much more irregular than that depicted in Figure 6. - 54 - Figure 6 Oil Production and Exports ,_________ OIL AND GAS OUTPUT …____ SHARE OIL COMPANIES . C : OIL EXPORTS O c) __.___ _ 0: Zc, DO. _ __ __ __ ___ E-ec.,, O '., _°~~~~~~~~~~~~~~~~~~9, C) . 0 a) I 192 0 1980.0 1990.0 2000.0 2010.0 2020.0 - 55 - Given exogenous overall investment and an exogenous trade balance, there are only two possible adjustments to the decline in exogenous foreign exchange receipts described above: export expansion or import substitution. Figure 7 below depicts both processes at work on the base case optimal growth path. Export expansion occurs in the form of a greater share of tradables having to be exported. From Figure 7 it is apparent that the late 1980's and early 1990's would be a period of dramatic export expansion under base case assumptions. Non-oil merchandize real annual export growth would average 5.3 percent from 1981/82-1986, 12.7 percent from 1987 to 1991, 13.7 percent from 1992 to 1995 and then decline gradually to 10.2 percent from 1996 to 2001 before stabilizing around 8.3 percent thereafter. The share of exports in domestic production of tradable commodities would rise from about 13 percent in the first half of tlhe 1980's, to 32 percent at the end of the planning horizon. While this is a very large increase over a relatively short period of time, it would bring Egypt's share of exports in tradable output to a level only a little bit above the average 25-30 percent characterizing most semi- industrial economies. Nonetheless, given the severe land limitations constraining agriculture, such an increase requires sustained and very rapid growth in manufactured exports over a period of 10 to 15 years. Parallel to the process of export expansion, the share of imports in domestic intermediate and final consumption would have to decline from 37 percent in the early 1980's to 27 percent at the end of the planning horizon. - 56 - Figure 7 Export Expansion and Import Substitution SHARE OF EXPORTS IN PRODUCTION OF TRADABLES ......... SHARE OF IMPORTS IN DOMESTIC USE OF TRADABLE GOODS . 0.40- ...... - 0.40 ....... ............ ...... 0.35/ 0.35 0.30 0 - 0.30 0.25 - - 0.25 0.20 - IP. 20 0.l5 - 0.15 0 0.10 I I I 0.10 1980.0 1990.0 2000.0 2010.0 2020.0 - 57 - All this amounts to a substantial increase in the share of the tradables producing sector in the economy. Figure 8a describes the base case results in terms of the share the tradable sector has in the capital stock, employment and output in the domestic non-oil and gas economy. All these shares increase nearly 15 points during the planning period. The structural adjustment depicted in Figure 8a is necessary to allow the required degree of export expansion and import substitution. But that only describes the supply side of the economy. The demand side also adjusts, although given the limited substitutability that is possible between the broad groups of tradable and non-tradable commodities, the adjustment is not that dramatic, as shown in Figure 8b. The structural adjustment characterizing the optimal growth path can also be described in terms of growth rates. Table 3 below summarizes the growth pattern for the four sectors distinguished by the model. Table 3 Average Annual Sectoral Growth Rates : The Base-Case (Percentages) Tradables Non-Tradables Oil Gas 1982-1985 7.0 6.4 6.0 30.0 1986-1991 6.7 5.6 2.2 8.9 1992-1995 7.2 4.4 4.3 1996-2001 6.6 4.6 -9.4 4.0 2002-2011 6.2 4.8 -16.5 0.0 - 58 - Figure 8a Share of Tradable in Domestic Non-oil Economy |___________CAPITAL . , EMPLOYMENT --------- VALUE ADDED 0.75- 0.75 0.70- 0.70 0.65- 0.60 - 0.60 0,55 ./ ,,w' . - 0.55 0.50 , 0.50 0.45 .. . ...-.0.45 0.40- - 0.40 I980.0 1990.0 2000.0 2010.0 2020.0 - 59 - Figure 8b Ratio of Consumption of Tradable over Non tradable Goods 1.27- - 1.27 1.26 - 41.26 1.25 - 1.25 1.24- - 1.24 1.23- 1.23 1.22- - 1.22 1.21- - 1.21 19.013 1990.0 2000.0 2010.0 2020.0 - 60 - Over the entire planning horizon, domestic production of tradables has to grow more rapidly than domestic production of non-tradables. The deviation from balanced growth is most pronounced in the 1992-1995 period during which the structural adjustment process is particularly dramatic. After the end of the century, the difference between the growth rates again narrows as the economy gradually adjusts to the near-exhaustion of oil reserves and the levelling off in the gas sector. At this stage, it may be useful to remember that throughout, by assumption, investment grows at an annual rate of 5.5 percent. Table 4 below summarizes the behavior of the main macroeconomic aggregates. Table 4 Growth of Macroeconomic Aggregates: The Base Case (Percentages) GDP Imports Consumption Investment Exports 1982-1985 8.7 6.6 7.6 5.5 10.2 1986-1991 5.1 5.2 4.7 5.5 6.3 1992-1995 2.7 3.8 2.7 5.5 2.8 1996-2001 2.7 3.9 2.6 5.5 2.9 2002-2011 3.0 7.3 2.8 5.5 8.2 GDP plus imports always equals the sum of consumption, investment and exports. With investment growth fixed and constant and imports-exports also fixed and growing at an approximately constant rate (between 3.2 and 4.1 percent) it is clear that the growth path of consumption must closely follow the growth path of GDP. As GDP growth declines to less than 2 percent in the - 61 - mid-1990's due to declining production in the oil sector (see Table above), consumption growth also must decline, since neither a reduction in investment, nor an increase in the resource gap can take place. The behavior of total exports (and therefore imports) constitutes an interesting feature of the growth path. Total exports first grow very rapidly (10.5 percent), then the growth rate becomes very small during the period of peak oil exports, before rising again. Figure 9 below separates the growth path of exports into its two components: oil exports and non-oil, tradable goods exports. The combination of very slow GDP growth and slow export growth during the 1992-1995 period, leads to low import, and even lower consumption, growth rates. Given that population growth is likely to remain above 2.9 percent per annum in tlhe early 1990's, 1.9 percent growth in total consumption implies a small decline in per capita consumption during that period, when the structural adjustment process reaches its most difficult stage. 5.1.3. Shadow Prices Ln the Base Case (a) Price of traded and nontraded goods The same process of transition to a post-oil economy can be seen from the valuation side of the model. The initial period of rapid expansion in oil revenues is reflected in an increase in the price of nontraded goods measured either in terms of imports, exports, or the domestic price of traded goods produced at home (Figures 10). Soon enough, however, the relative decline in Exogenous Resources as proportion of the domestic economy begins to make tradable goods scarcer and - 62 - Figure 9 The Composition of Exports 0TOTAL EXPORTS ........... ......NET OIL EXPORTS ----------- EXPORTS OF TRADABLE GOODS BO.O 80.0 60.0 2 -0.0 40.0 - 40.0 20.0 - 20.0 0.-"',.,,,- 0.0 -20.0 . -.20.0 -40.0- --40.0 1980.0 1990.0 2000.0 2010.0 2020.0 - 63 - Figure 10a Real Exchange Rates: Price of Nontraded Goods in Terms of Various Traded Goods PRICE IN TERMS OF IMPORTS .......... PRICE IN TERMS OF THE COMPOSITE TRADED GOOD 0.95- _ - 0.95 0.9go - ' .' ... . 0.90 0.85- - 0.85 0.80- - 0.80 0.75 - -0.75 1980.0 1990.0 2000.0 2010.0 2020.0 - 64 - therefore, more expensive to consumers and more valuable to producers. As Figure IOA shows, there is a substained fall of the price of nontraded goods in terms of imports and the composite traded good consumed domestically. This drop in the consumption real exchange rate is just the price counterpart of the declining share of nontraded goods in consumption seen in Figure 8. In the production side the story is slightly more complicated. On the first place, there is a widening gap between the domestic price of exportables and the price of exports, reflecting the fact that when the volume of exports cannot be expanded without reducing the unit value obtained from them, it is optimal to divide the output of tradable goods between the domestic and foreign markets in such a way that a wedge (an "optimal tariff") exists between international and domestic prices: the latter should be equal to the marginal revenue from exports, which is lower than the price paid by foreign consumers. Figure 10B shows that the average values of traded goods (i.e. the average of domestic and foreign prices, with weights proportional to the share of output going to each market) does in fact increase with respect to the price of nontraded goods (dashed line). There is a second reasons why the ratio between domestic prices of nontraded and traded goods is not a perfect indicator of the incentives for allocating domestic resources among sectors: part of the increase in PN/PT is needed just to compensate for increasing cost of energy, without really implying that domestic factors could earn more there. As Figure 10b shows, in fact the ratio of value added in both sectors hardly changes from the late 80-s on: most of the increase in PN/PT is due to increasing price of oil. The domestic cost of energy is also the reason for the fall in real exchange rates in the late 1980's and its rise in the late 1990's: these two periods mark the - 65 - Figure lOb Real Exchange Rates: Price of Nontraded Goods in Terms of Various Traded Goods Price of Nontraded Over Price of Traded Goods .............. Ratio of Unit Value Added -------------- Price of Nontraded Over Average Value of Traded Goods Price of Nontraded Over Price of Exports 1.00 - 1.00 0.95- - 0.95 090 - 0 #t ~~~~--- ,------ 0.985- . ~ . -0.90 0.85 - 0.85 0. 80- 0.80 0.75 -0.75 1980.0 1990.0 2000.0 2010.0 2020.0 - 66 - beginning and end, respectively, of the decade of cheap natural gas (see b), below), as its production/distribution capacity expands (temporarily) beyond domestic demand. Nontraded goods are more energy intensive, so they benefit more from low cost energy. (b) Price of Gas; Shadow Price of reserves of oil and gas The price of oil is exogenous to the model. In the base case scenario it is assumed to be constant in real terms until 1984 and grow from then on at an annual rate that declines from 3% to around 2% in 2010. We discuss alternative price scenarios in section 5.1.5. The model divides the price of oil in three components: rent to the stock of reserves, cost of extraction and rent to capacity limits. As Figure 11 shows, the latter component is significant at the begining of the planing period, while production capacity is being built up. The rent component rises up to about 60% of the price of oil, but later declines as lower reserves lead to higher extraction costs. The price of gas and its decomposition in rent to reserves, extraction cost and rent to capacity limits, is presented in Figure 12. As discussed in section 3.2, (see also section 4.1) we approximate the substitution possibilities between oil and gas by differentiating sharply between uses in which they are perfect substitutes (55% of total demand for energy), and uses in which gas cannot replace oil products at all. In addition, there are output constraints (including limitations in the capacity to transport the gas to users). The value (shadow price) of gas is therefore eqbal to its fuel oil equivalent (i) whenever the production constraints do not allow us to exhaust domestic substitution possibilities, or (ii) when - 67 - Figure 11 Decomposition of the Price of Oil (US$/barrel) VALUE OF OIL RESERVES .......-- PLUS RENT TO CAPACITY CONSTRAINTS - - EGYPTIAN SHARE ! -.-.--.-PRICE OF OIL 70.0- - 70.0 60.0 7 60.0 so. 0 - // -50.0 40.0 ./," ' 40.0 30.0 -~. - - - 30.0 - 30.0 20.0 / 20.0 10.0- I- 1980.O 1990.0 2000.0 2010.0 2020.0 - 68 - reserves are so scarce that it is optimal not to produce so much as to substitute fuel oil wherever it is possible, in order to allow some production in future periods.42' As shown in the figure, the result of these factors is that only during about 10 years from the late 1980's the price of gas is below (by 25-30%) to its fuel-oil equivalent. At exhaustion time, production capacity constraints are not binding, so the price of gas equals extraction costs plus rent to reserves. As the cost of extraction is independent from the stock of reserves, the value of reserves is constant in discounted value (i.e. it grows at the social rate of discount). The difference between the price of gas, net of extraction costs, and the value of reserves is the imputed rent to the production (and transportation) capacity constraints of gas. They show the considerable payoff that would accrue to projects increasing the quantity of gas that can be produced and delivered to users (Figure 12). 5.1.4. Sensitivity to Oil and Gas Reserves It is clear that oil and gas reserve estimates have an influence on the nature on the optimal growth path. In a first set of experiments we have tested the sensitivity of the projected "structural adjustment path" to variations in oil and gas reserves, keeping all other exogenous variables constant. Table 5 describes three of the alternative reserve scenarios that have been explored. These various reserve scenarios are based on a fairly non- controversial estimate of initial proven reserves (500 m.t. for oil and 250 m.t.e for gas) and projected new discoveries the magnitude of which is, of course, much more uncertain. It would be a fundamental mistake, however, to - 69 - Figure 12 Decomposition of the Price of Gas -_____ VALUE OF GAS RESERVES .......... PLUS RENT TO PRODUCTION/DISTRIBUTION - - … PRICE OF GAS PRICE OF OIL 70.0- - - 70.0 60.a- //- 60.0 so. a 7/ 50.0 1 0.0- , ,' /- 40.0 30.0 - 30.0 20.0 / 7 - 20.0 10.0 10.0 1980.0 1990.0 2000.0 2010.0 2020.0 - 70 - attach zero probability to new discoveries given: (i) the large investment effort in exploration that is underway, and (ii) past performance which has produced a steady stream of small discoveries of both oil and gas. There is little doubt that new oil and gas will be discovered in Egypt. The question only relates to the magnitude and time pattern of these discoveries. Table 5 Alternative Scenarios for Oil and Gas Reserves (Million tonnes Oil equivalent) Base High Oil Low Oil Case and Gas & Gas Reserves Reserves Proven Oil Reserves 500 500 500 New Oil Discoveries over the Planning horizon 400 655 100 Proven Gas Reserves 250 250 250 New Gas Discoveries over the Planning 300 590 150 horizon Regarding oil, the base case assumes annual discovery flows starting at 25 million tons in the mid-1980's and then declining gradually to 10 million tons in the mid-1990's before ending in the year 2000. The optimistic scenario assumes higher initial discovery rates, again declining but with - 71 - discoveries continuing throughout the planning horizon. The pessimistic scenario, on the contrary, projects only very small discoveries ending after the early 1990's. For gas, the situation is somewhat different. Exploration in the very promising Nile Delta region has only recently got underway and the next 10 to 15 years are likely to be characterized by steady rather than declining discovery rates. Figure 13 below describes the production schedules for the three scenarios explored: the Base Case compared to the high reserves (oil and gas) and low reserves (oil and gas) scenarios. Larger reserves postpone the period of self sufficiency in oil by about 5 years, while with low reserves imports are necessary about 8 years earlier than in the base run. The production of oil/gas and traded goods are (non perfect) substitutes, so that the stock of capital in the traded sector becomes large with low reserves and smaller with large reserves (cf. Figure 14). The most dramatic effect of changing the reserve estimates is on the pricing of gas. As we discussed above (Section 4.1) it is only when all domestic opportunities for substituting fuel oil by gas have been exhausted that the price of gas can fall below that of its oil- equivalent. In the base reserve scenario that happens only for a 10-years- period in the late 1980's and 1990's. As Figure 15 shows, with low reserves the period of cheap gas entirely disappears, while large gas reserves extend it to most of the plannlng horizong (after the initial period of capacity constrained gas output). - 72 - - Figure 13 Production of Oil Under Alternative Reserve Scenarios [o- - - - - LRWC OIL ND CflS RESERVES CA Q C;_ _ci 0 {D LF) C) C) I~~~~~~~ 2zC>_ * \ci ci - ~~~~~~~~~~~~~~~C3 *~~~~~~ t 980. .1990. 0 2000.0 2010. 0 2020 .0 - 73 - Figure 14 Capital in the Traded Sector Under Alternative Reserve Scenarios [71 BRaSE mcn [L - LRRE OIL AND GAS RESERVES G) o LOW OIL FN GAS RESERVES a:o~~~~~~~~~~~~~~~~~t O :^-*-. O~ .I--. ' Cr) G: . . 4 c H Cy * 0~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~- G ~ ~~~~~~~~ . . *. G O v--' ~~~~~* 4 C) 's CD c \ t___ o aS) , _%. Lc'R . v~~~~~~ v Cr) 9BS* 93.0 2000 21.0 2:' 0 C3 CC CD~~~~~ 'I 02~~~~~~~~~~~~~~~~~~~r LLJO v ci C 9B. 199.I2000 2100 22. - 74 - Figure 15 Price of Gas Under Alternative Reserve Scenarios § R~~~WSE CRISE - - - - - LNRrE OIL RND CS R£SERES O .............. LOW OIL RND PRS RESERVES C .~o _ : o , _ ~~~~~~~~~~~~~~~~C) -f 2 C) M~~~~~~~ I~~ C~L) cmII Q cm~~~~~~~~ 0-~~~~~~~~~ 0~ to L/ 0 cr-~~~~~~~~~~~~~~~~~~~ LU') / CD ~ ~ ~ ~ ~ ~ 10 (90 I~~~~~~~~~~~~~~~~~C CS ci~~~~~~~~~~~~~ 19800 190. 200.0 0100 220. - 75 - Table 6 describes the sensitivity of GDP, consumption and export growth to the alternative reserve and production scenarios. High reserves allow a larger rate of growth of GDP and consumption through the planning period, although the difference in average annual rates is no more than 1%. The early 1990's is still the most difficult period, while there is an improving trend after the year 2000, as the economy successfully completes the transition toward a much reduced role of the oil and gas sector. As the last part of the table shows, with pessimistic scenarios for new oil discoveries, non oil exports must grow at very large rates during the late L980's and early 1990's (near 15% per year) just to keep total export revenue from falling behind import requirements. The bright side of low reserves is that the economy reaches its non-oil based steady growth path sooner. And indeed, toward the end of the planning period the three scenarios hardly differ from each other. 5.1.5 Sensitivity to the Real Price of Oil So far the growth path of the real price of oil was kept constant throughout the experiments. The price scenario adopted is the World Bank's "official" scenario of 3 percent real growth until the early 1990's and a gradually declining price growth thereafter. According to this scenario, the price of a barrel of crude in constant 1981 prices would be $38 in 1990, $49 in the year 2000 and $59 in the year 2010. Any public or private economic decision maker must project the real price of oil: unfortunaLtely there is no escape from this, despite the obvious- ly massive uncertainty attached to future oil prices. The path of world oil prices will depend on a multitude of factors such as worldwide discovery - 76 - Table 6 The Impact of Alternative Reserve Scenarios on the Growth of Macroeconomic Aggregates (Percentages) Base High Low Case Reserves Reserves GDP 1982-1985 8.7 8.9 8.7 1986-1991 5.1 5.8 3.0 1992-1995 2.7 2.9 1.8 1996-2001 2.7 3.0 2.6 2002-2011 3.0 4.1 3.9 Consumption 1982-1985 7.6 7.8 7.6 1986-1991 4.7 5.2 3.0 1992-1995 2.7 2.8 1.8 1996-2001 2.6 3.0 2.4 2002-2011 2.8 3.9 3.6 Exports 1982-1985 10.3 10.4 10.4 1986-1991 6.3 7.2 3.3 1992-1995 2.8 3.1 2.1 1996-2001 2.9 3.2 8.6 2002-2011 8.2 5.4 7.5 Non Oil Exports 1982-1985 7.7 6.6 8.1 1986-1991 11.6 11.0 14.9 1992-1995 11.8 11.9 12.4 1996-2001 10.5 10.1 9.7 2002-2011 8.6 7.8 7.5 - 77 - rates, worldwide economic growth, technological developments affecting other forms of energy (solar, nuclear, biomass, etc.), conservation and domestic price policies and, last but not least, political developments of various sorts affecting the production and worldwide trade of hydrocarbons. All these factors could combine Ito drive the barrel of oil steadily up in value. On the other hand, the combination of a series of technological breakthroughs, major new discoveries and slow economic growth in the major consuming nations could lead to constant or even declining oil price. To explore the sensitivity of the optimal growth path to variations in the assumptions about future oil prices, we conducted two sets of experiments with alternative oil price scenarios. In the first group we only change prices in the far future (after the year 2000). In the second group oil prices are higher (lower) when the economy is a net importer and lower (higher) when it is a net exporter,i.e. the price schedules are tilted. For comparison we also include a scenario with oil prices uniformy below the base run values. We discuss; a scenario with even lower prices throughout (falling initially to $25 per barrel) in Section 5.3, where optimal foreign borrowing is allowed. Table 7 and figures 16 and 17 describe the scenarios discussed here. If the expected price of oil in the latest part of the planning period is larger than in the Base Case, there is an incentive to save oil initially, to be able to extract it later, when it will be more valuable. That is exactly what Figure 18 shows. There are changes in the allocation of capital, which are just the mirror image of the changes in oil output. - 78 - Table 7 Alternative Scenarios for Real Price of Oil Price per barrel of Oil (constant 1981/87 US$) (1981/2 US$ per barrel) High Low Prices Prices Low Low First High First Year Base After After Price High Later Low Later 2000 2000 1985 31.8 31.8 31.8 26.7 26.7 32.4 1990 37.9 37.7 37.7 27.3 27.3 40.3 2000 48.7 43.8 51.1 35.0 35.0 50.8 2010 58.9 44.2 68.6 42.4 42.4 54.4 Figure 19 describes the oil output path under the tilted price scenarios. When the oil price is "Low First, High Later," we have just a magnification of the incentive toward future extraction found on the "High Price After 2000" scenario. In the other cases there is only a small bias toward earlier extraction, as initial prices increase relatively more (or fall relatively less) than future prices. The different price scenarios have however a significant impact on real consumption, which is of course our welfare indicator (see Figure 20). If prices are Low First - High Later, that is bad on both counts: oil can be sold abroad only at a low price, and it becomes expensive when it must be imported. That shows up in the figure, where real consumption is seen to be more than 12% below its Base values, in the midle of the "temporary-cheap oil" - 79 - Figure 16 Alternative Oil Price Scenarios After Year 2000 BASE PRICE -HIGH PRICE AFTER 2000 ...... ... LOW PRICE AFTER 2000 00- m~~~~~~~~~~~~~~~~ LI~~~~~~~~~~~~~~ 0 . tJ' #~~~~~~~~~~~/ CD, 19m).O l9m.( 2000.0 2010.0 2020.0 - 80 - Figure 17 Tilted and Permanently Lower Price Scenarios BASE PRICE --…- LOW PRICE ........... LOW FIRST, HIGH LATER HIGH FIRST, LOW LATER o 0............. o / 0 '9~~~~~~~~ 19g0.o 1990.O 0D.0 20X0.0 2020.0 - 81 - Figure 18 Oil Output Under Different Price Scenarios After Year 2000 BASE PRICE ----------- HIGH PRICE AFTER 2000 ........... LOW PRICE AFTER 2000 C ....................... o0 0- I .0 .4 O ~~~~~~~~. Pa'~~~~~~.44 ...a J°_ / CD 10 -J~~~~~~~~~~~~~~~~c 00 0lgo. 2 .4 2000 a - 82 - Figure 19 Oil Output Under Different Price Scenarios BASE … - - - - LDOW PRICE .......... LOU F fRsr, HJAH LflTER -Q -. * *HIGH6 FIRST,LOW LATER 1 C1) 1'L( O O U) D c) M~~~~~~~~~~~~~~~~~~~~~~~~~~ 0W II~~~~~~~~~~~~~~~~~~C q X 6 ~ ~ ***. \ C.C C 9 0) M 2980.0 1990. 0 2e00. 0 2010.0 2020.0 - 83 - Figure 20 Real Consumption Under Different Price Scenarios BASE …---------- LOW PRICE ........... .LOW FIRST, HIGH LATER HIGH FIRST, LOW LATER G -- - U, _ cnu,~~~~~~~~~~~~~~~L Log C-, 0 e~~~~~~~~~~~~~~0 _) o o G- 63 _ _ Cl ~ ~ ~ ~ ... a:.-- /. C.3 ~ ' o3 CD. zr Ea 9.0 20. 000 2 . - 84 - period. The reallocation of oil output is substantial, actually allowing continued exports within the period of high prices: Peak consumption is actually higher than in the Base Case during that period of additional exports (2000-2010). Things are much brighter when cheap oil is permanent ("Low Price" scenario), as better future perspectives makes it advisable to compensate lower prices with larger output, generating only a small fall in real consumption. The last price scenario (High-Low) is good for the same reasons that the Low-High case in bad; the magnitude however is modest given the size of the price differential with the Base Run. Figures 21 and 22 describe the allocation of the capital stock and the price of gas, in reference to the Base Price scenario. The stock of capital in the traded sector is seen to follow the same pattern as real consumption, and for the same reasons. With permanently low prices, more production of tradeable goods is required initially (1980-2000), to compensate for lower oil revenues, but later more of the capital stock can be allocated to the production of nontraded goods, since energy imports are cheaper. This second period of relaxation is absent in the Low-High prices scenario, so that oil/gas reserves are exhausted at a lower rate, requiring a larger proportion of the capital stock in import substitution/export production (which in our model are the same thing). An interesting result is how the strong temporal bias from combining low initial prices with high prices after year 2000 completely eliminates the period of "cheap" gas. The reason is that, from the perspective of the later 1980's, higher future prices and lower current ones increase the value attached to keeping gas in the ground: its rental value increases, and production is shifted towards the high price future to such an - 85 - Figure 21 Capital Stock in the Traded Sector Under Different Price Scenarios | . W~ESE . -- -....... Low FrRST, HIGH LR1flR C,o) -C-* HICH FIRST,LOW LMEER e) -_ q-4 V) ¢~~~~~. *j*:. ic 0 * 99~~~~~~~~~~~~~~~~~~~* cr: . . C:; 'C - (0 C) N w-4, a ,_0.1. -%. ~ - * O. \ 202 Cc c C) Cs ' * ~ 6~~ 19BO.0 1990.0 2000.0% 2000 22. PRICE OF GAS OVER PRICE OF OIL 0.5. D.6 D.7 .8 0.9 1.0 co 0 .o X ~~...................... =.;,,,. .,i o o c)4 D.5 D.6 0.7 0.8 0.9 00 :1~~~~~~~. ro I 0.6~~~~~~~~~~~~~I 0002.8 - 87 - Figure 23 Price of Gas Under Different Price Scenarios D.SE ---- LOW fRIC ............ LOw FRST, HICH LRTER O -*-* HIrHFrRS7,LDW LRO C: C o ,0 9o- 0~~~~~~~~~~~~~~~~~~0 EcJ 00 tLi 0~~~~~~~~ Cs Ln In~~~~~~ Cs1! (90LJ99W0 2000 200. 02. - 88 - extent that the demand constraint is not binding anymore and the wedge between the fuel price equivalent and the shadow price of gas disappears. Substantially higher initial effort in the production of non-oil tradables is required with pessimistic price projections. Still, that effort is not enough to compensate for lower oil export earnings, so that total exports, imports and GNP grow at a slower rate than in the base case (see Table 8). The optimistic view on the backstop (Low Price after 2000) reduces the need for rapid growth in the next century, as the imported energy bill does not grow so fast. 5.1.6. Sensitivity to the Growth of Domestic Demand for Energy While varying projected reserves and oil prices, we have so far kept the behavior of domestic energy input coefficients constant across experi- ments. Specifically, we assumed that starting in 1986/87, energy input coefficients would slowly decline, reflecting more energy efficient investments and a gradual increase in domestic energy prices. Domestic energy prices in Egypt in the base period are extremely low, less than 20 percent on average of the corresponding world price levels. These low prices, combined with rapid economic growth in the 1970's and rapid urbanization, have led to extraordinarily high growth rates of domestic demand in the late 1970's and early 1980's. In the recent past total demand for energy increased at rates averaging more than 10 percent per annum. It is extremely unlikely that such rapid growth will be sustained beyond the mid-1980's. Economy wide growth was exceptionally high (averaging almost 10 percent) during the 1977-1980/81 period. No matter which scenario we choose, the future will be characterized by more moderate overall growth, - 89 - Table 8 The ITapact of Alternative Oil Price Scenarios on the Growth of Macroeconomics Aggregates (Percentages) High Low High Low Prices Prices First First Base After After Low Low High Case 2000 2000 Price Later Later GDP 1982-1985 8.7 8.7 8.7 7.5 7.5 8.8 1986-1991 5.1 4.7 5.7 4.5 1.7 5.4 1992-1995 2.7 3.0 2.2 3.2 5.5 2.8 1996-2001 2.7 3.0 2.3 3.4 5.0 2.4 2002-2011 3.0 2.7 3.6 3.7 3.4 3.0 Consumption 1982-1985 7.6 7.6 7.6 6.8 6.8 7.7 1986-1991 4.7 4.3 5.1 4.2 1.8 4.9 1992-1995 2.7 2.9 2.2 3.1 5.0 2.7 1996-2001 2.8 2.8 2.4 3.2 4.6 2.3 2002-2011 2.0 2.3 3.5 3.5 2.9 3.0 Total Exports 1982-1985 10.2 10.2 10.2 8.8 8.8 10.4 1986-1991 15.3 5.6 7.3 5.5 1.7 6.8 1992-1995 2.8 3.1 2.2 3.4 6.5 3.0 1996-2001 83.2 8.8 7.7 7.9 8.2 7.9 Non Oil Exports 1982-1985 7.7 7.7 7.7 11.9 11.9 7.2 1986-1991 11.6 12.4 10.6 10.4 15.0 11.3 1992-1995 1]L.8 11.1 12.8 10.6 7.1 11.9 1996-2001 10.5 10.2 10.9 9.5 7.3 11.1 2002-2011 8.6 9.1 7.9 8.0 8.6 8.4 - 90 - translating into more moderate domestic demand growth. More important in this context however, is the fact that Egypt is gradually moving towards higher domestic prices and is also contemplating major efforts to improve the energy efficiency of the domestic economy. There are some very large and very inefficient energy users in domestic industry. There is therefore the potential for considerably reducing energy coefficients as higher prices and efficiency-improving investments show their effect. To explore the impact of different assumptions about domestic energy efficiency on the nature of the optimal growth path, we have compared the base case to two different scenarios derived from the assumptions presented in Table 9. It shows how the unit energy requirement in each sector is assumed to change through time in the different scenarios. The Base Case postulates a decrease of 13% in the quantity of energy used per unit of output between the present and the year 2000, as a result of domestic pricing policy, changes in technology and the composition of output, during a period of increasing energy prices. The High Energy Efficiency case, assumes that these savings can be taken further, so as to reduce the energy/output coefficient by the end of the century to 78% its initial value (and to 63% by 2010). In the Low Energy Efficiency scenario no conservation effort is made, domestic users of energy are subsidized, etc. so that energy use per unit of output increases (by 10% near the year 2000) over its initial levels. In addition to the behavior of the energy coefficient, a major determinant of the domestic demand for energy is the composition of output. In particular, the nontraded sector is significantly more intensive in the use of energy, as it includes some of the big users, e.g. transportation, electricity. That somewhat moderates the effect of reduced energy efficiency when the economy's net oil importer, since - 91 - more demand for energy requires more production of tradable goods to pay for them, what reduces the average energy/output coefficient. But as Table 10 makes clear, continuing low energy efficiency nevertheless puts the economy under considerable stress, reducing by almost one half the annual rates of growth in GDP at the end of the planning period. The large energy imports requirements also calls for even more rapid growth in the production of tradable goods. Table 9 Alternative Scenarios for Domestic- Energy Efficiency Index of Energy Requirements per Unit of Output High energy Low Energy Year Base Case Efficiency Efficiency 1985 1.0 1.0 1.0 1990 96.3 93.8 103.1 2000 86.9 78.7 110.9 2010 77.5 62.5 118.8 Table 10 The Impact of Different Energy Efficiency Scenarios on production structure, energy use and GDP growth rates Ratio of Production GDP Real Domestic Use of Tradable to of Energy Nontradable Goods High. Low High Low High Low Base Eff. Eff. Base Eff. Eff. Base Eff. Eff. 1982-85 8.7 8.7 8.7 6.5 6.5 6.5 .8 .8 .8 1986-91 5.1 5.3 4.6 4.9 4.3 6.6 1.3 1.2 1.7 1992-95 2.7 2.8 2.6 4.5 3.8 6.2 2.3 2.2 2.6 1996-2001 2.7 3.0 1.9 4.3 3.5 6.1 2.3 2.0 2.9 2002-2011 3.0 3.5 1.7 4.2 3.2 5.9 1.8 1.5 2.7 - 92 - Figures 24 and 25 present the same story, in comparison with the Base run. The cumulative impact of higher energy intensity on real consump- tion is substantial, particularly in the period of net energy imports. On the other hand increases in energy efficiency could increase real consumption by near 10% toward the later parts of the planning horizon. The magnitude of the reallocation of capital between sectors with changing energy efficiency (Figure 25) appears as modest: the total capital stock in the traded sector is only about 6% higher (toward the end of the planning period) with Low Energy Efficiency than in the Base Case. But two things should be pointed out. One is that this reallocation comes on top of a substantial reorientation of tradable goods, so it may not be so easy to accomplish. The second is that the lower increase in KT is partly due to the significantly lower energy intensity of traded goods with respect to nontraded goods, and that may represent a further challenge, when the mean of tradable goods expands beyond those currently produced. 5.2 Experiments with Endogenous Investment 5.2.1 A Fundamental Optimality Condition Throughout the analysis in the preceding sections, the rate of growth of investment, and therefore the pace of capital accummulation was fixed at 5.5 percent per annum. This specific growth rate was chosen because it appears reasonable as a long-run average, in light of Egypt's development objectives as well as the experience of other developing or semi-industrial economies. The growth of investment was kept constant, however, only as a first step in the analysis of the growtlh and structural adjustment process. - 93 - Figure 24 Real Consumption Under Different Scenarios for the Efficiency in the Domestic Use of Energy BASE ,-s _ _ - - - ~HIGH ENERGY EFFICIENCY LiG ...... .... LOW ENERGY EFFICIENCY ta: - - { -I - _ CO o - _. .,- - a: '@@ *. 0~~~~~~~~~ 0 aZ' G 9O0 90.ao1.0 2 .ci 0 U) . z 0~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ c;Q C:) a: 1980.0 1990.0 2000.0 2010.0 2020.0 (Ii - 94 - Figure 25 Stock of Capital in the Traded Sector Under Different Scenarios for Efficiency in the Domestic Use of Energy fL-i . BASE HIGH ENERGY EFFICIENCY aGm o ............- LOW ENERGY EFFICIENCY o E,L_ O-.,,,-_ G .' O ,: (C) b (10~~~~~~~~~~~~~~~~~( 0 CD { J s al~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ E 19B0.0 1990.0 2000.0 2010.0 2020.0 - 95 - In fact, one of the key questions facing the investment planner is precisely how to determine the optimal investment rate. This section attempts to contribute to a better understanding of the issues involved. How should investment behave on an optimal growth path? Should it be a rising, constant or declining fraction of total resources? If more oil and gas is discovered, should the economy-wide investment rate rise or decline? What is the sensitivity of the optimal investment rate to different oil price scenarios? To alternative assumptions about social time preference or the pace of technical progress? These questions are, of course, linked to the issue of optimal foreign borrowing. In this section we analyze optimal investment behavior keeping the current account fixed exogenously before, in a later section, allowing capital flows to become endogenous. Full endogeneity of capital flows is not a very realistic assumption in Egypt's case since both domestic and international political factors are very important in determining their possible range of varialtion. Before turning to the numerical results derived from experiments with the model, it is worth discussing in a qualitative way the important factors that will influence the behavior of the optimal savings and investment rate: (i) Traditional national accounting treats income derived from the extraction of a natural resource just like any other kind of income generated in the economy. This similarity is quite misleading. The income derived from a textile plant or from a cotton fielcd is different from the income derived by operating an oil well. I]n the former case, the generation of net income (after taking account of capital depreciation) does in no way subtract - 96 - from the stock of wealth available to the economy. But in the case of an exhaustible resource the generation of current income is accomplished by running down the country-s wealth. It is intuitively clear, therefore, that the optimal proportion of savings in national "income depends on the nature of that "income". To the extent that income generation implies natural resource decumulation, it would appear that this decrease in national wealth should be compensated by a higher investment rate that builds up alternate forms of wealth. Another way of looking at it is by noticing that the only way the exhaustible stock of oil can be shared with generations to come after the exhaustion date is to transform it into a form of reproductible wealth. (ii) With investment able to vary endogenously, the economic return of investment will always be equated to the consumption rate of interest on an optimal growth path. The economic return on investment is measured by the increased consumption stream made possible by the additional investment and/or any increase in the value of capital goods. As long as that economic return exceeds the CRI, it is worth increasing the investment effort until the marginal return on investment inclusive of capital gains equals the marginal cost of postponing consumption. (iii) In a country with exhaustible resources, the required equality between the consumption rate of interest and the marginal productivity of investment extends to the return on "investment' achieved by keeping the natural resource in the ground. The net economic return on the natural resource must also equal the consumption rate of interest. The time path projected for the - 97 - world price of oil is, of course, an important determinant of the economic raite of return on holding reserves in the ground. With higher expected future real prices, it is worth keeping more oil and gas in the ground. First, oil production costs depend on the extraction to reserve ratio. The rate of return or limiting extraction reflects therefore both the expected price appreciation and the decrease in production costs achieved by reducing the future output to reserve ratio. Second, while the income derived from extraction could be consumed, a large fraction of it will or should in faLct be invested in acquiring alternative forms of wealth in thle form of capital goods, either through imports or domestic production. This brings us back to the factors discussed in (i) and (ii) above. In fact, on an optimal path, investment and saving must be such that at every point in time, the following equality is satisfied. The Economic Rate of The Social The Economic Return on Return of Investment = Rate of 5 Holding Natural in Capital Goods Interest Resource Reserves This equality reflects the essential efficiency condition character- izing an optimal growth path. While the basic concepts are quite clear, in practice one has to be very careful when interpreting this optimality condition. It strictly holds only if each of the three "rates of return" is expressed in terms of a common unit of account and is defined to include any "capital gains" or "capiltal losses" that may arise because of relative price changes. The model used for the endogenous investment experiments is similar to the one used in the previous section, with two differences: first, we do not distinguish between oil and gas, as that distinction does not seem - 98 - particularly relevant for the issues discussed in this section; second, we incorporate the real world fact of absorptive capacity constraints on capital accumulation in the form of a diminishing marginal efficiency of investment. We assume that capacity constraints apply only to aggregate investment, on an economy wide level, reflecting shortages in such crucial factors as management capability, special skills, port capacity and the transportation network. As explicit and detailed modelling of these factors is outside the scope of the aggregated model, we have adopted a shortcut to incorporate them. Up to a cut-off level XLIM no absorptive capacity constraints are operative; XLIM is set at 10 percent above the actual increase in capital generated by last year-s investment. After that, the extent to which resources devoted to investment actually generate usable capital goods declines exponentially on the margin. The existence of such a constraint does not affect the basic optimality condition equating economic rates of return but care should be taken in defining the marginal productivity of devoting one unit of income to capital formation when such constraints are binding: resources devoted to investment will not generate capital on a one for one basis when the constraint is binding, so their marginal productivity should be adjusted downwards. For example, the costs of a steel mill or a land reclamation project are of two kinds. First, there are the energy, labor, capital and raw material costs that are traditionally taken into account in project evaluation. Second, there are the more subtle and indirect costs imposed on the economy by projects that use the particular scarce factors that are responsible for the overall absorptive capacity limits on the economy. - 99 - The declining marginal efficiency of investment schedule is anl attempt to capture in the model these very real yet often neglected costs of rapid capital accumulation. On the other hand, this treatment also implies that projects that lead to a relaxation of such constraints should carry a premium over the value of capital if only future outputs of the project itself are considered. 5.2.2 The Base Run The Base Run values once again vividly demonstrate the major trans- formation Egypt will have to go through in the next decade or so. Consider for example Fig. 26 which gives the composition of wealth as a function of time. The value of oil in the ground (including anticipated future discoveries) currently accounts for 65% of total wealth while the value of the capital stock in the traded goods sector is about 19% of total wealth. Twenty five years from now however, oil wealth will have fallen to 18% of the total, while capital in the T-sector then makes up 56% of total wealth, if at least optimal extraction, saviLngs and allocation policies are followed. This major shift towards capital in the T-sector is needed to replace oil revenues as a source of foreign exchange. The share of oil in total wealth in the fixed investment case is much higher 20 years from now (26%), reflecting the fact that extraction of oil and capital accumulation are considerably higher over the next 25 years if capital accumulation is left free to vary over time. Around that time (2005) rent income and "exogenous" resource transfers will only make up 33% of total income, as opposed to 48% in the base year (Figure 27). - 100 - Figure 26 Composition of Wealth, Base Run SHARE OF OIL RESERVES IN TOTAL WEALTH ...... CAPITAL STOCK IN TRADABLES OVER TOTAL WEALTH 1.0- - 1.0 0.8" O.8a - 0.8 0.6- 0.6 0.4- 0.4 0.2 - 0.2 0.0- 0.0 1980.0 1995.0 2010.0 2025.0 2040.0 TIME - 101 - Figure 27 Oil Revenues and "Exogenous Transfers" as a Share of Total Income 0.50 - 0.50 0.45 -0. 45 0.40- - 0.40 0.35- - 0.35 0.30 - - 0.30 0.25 0.25 0.2 4 0.20 1980.0 1995.0 20;0.0 20Z5.0 2040.0 TIME - 102 - One of the more important indications to come out of this run is that the rapid depletion of oil wealth necessitates a faster rate of capital accumulation than is currently undertaken. In our base run first period gross investment takes 29% out of GDP as opposed to the 27% currently devoted to it because absorptive capacity constraints are binding (more on that in the discussion of the ARI) but that share quickly rises to 33% in 1983-84 and continues to go up to reach a peak of 43 in the middle of the next decade. If we look at net investment this translates into a savings rate out of Net National Income (inclusive of exogenous transfers) of only 22% because absorptive capacity constraints cause a shortage of profitable investment projects and accumulation of Foreign Assets is ruled out by assumption. But the optimal savings rate is clearly higher during the period of high oil revenues, reaching a peak of 37% out of National Income in 1990 before it settles down to about 30% in the post-oil period around the turn of the century. This comes down to a marginal savings rate out of oil income of not less than 51.4% during the peak year in 1990! All this implies that over the immediate future (1981-86) capital accumulation should accelerate at more than 15% as absorptive capacity constraints losen up rather than the 6% currently envisaged. Also an increasingly large share of that investment program, starting out at 57% but increasing to 71%, will have to be channeled into capital formation in the traded goods sector. The net effect of the decline of oil revenues and build up of capital in the T-sector is of course a rapid increase in both exports of non- oil traded goods and import substitution. The volume of exports should grow considerably faster than GNP over the planning horizon (for 1981-86 the numbers are 14.7% and 9.0% respectively.) - 103 - Due to the anticipated deterioration in the terms of trade these numbers look less dramatic in value terms (exports grow 13% in value terms over that period). Also the share of domestically produced traded goods in the consumption of traded goods in general will rise considerably from around 63% now to 74% in 2005. To make all this possible a major shift of non-oil production towards traded goods, away from NT goods, is called for when oil revenues start declining. Figure 28 gives the proportion of the labour force to be employed in the T-sector as a function of time. A gradua:L appreciation of the real exchange rate (price of NT goods in terms of domestically produced T-goods, the relevant definition for domestic resource allocation) of 7% during the period with oil (the next 20 years) is called for, to be partially reversed later on (the real exchange rate eventually drops nearly 3% from that peak level), to sustain this reallocation. The rapid increase in exports can only be achieved if Egypt's external competitiveness; improves, the "warranted growth rate" of exports is higher than that of world trade. Accordingly, a considerable deterioration in the terms of trade is necessary over the next 25 years (10%). The net effect of all this on GNP is a heady 9% real growth projection for 1981-86 and 8% for 1986-1992. These percentages are somewhat lower if oil revenues are not included, non-oil GDP will grow at 8% and 6.6% respectively during 1981-1986 and 1986-92. After that, the decline in oil revenues slows down GDP growth to a rate below population growth until about 30 years from now. Non-oil GDP of course grows much faster after 1992 at 5.2% but it takes several decades before that component starts dominating the declining oil component, it is not until 2010 that the model predicts real - 104 - Figure 28 Fraction of the Labor Force Employed in the T-Sector 0.6 0.6 0.4- . 1 , , - 0.4 19O.0 1995.0 2010.0 2025.0 2040.0 TIME - 105 - growth rates of 3.5% and higher for total GDP. Part of that reflects the decline in the terms of trade (the growth measures mentioned refer to GNP with world traded final goods as numeraire). In constant prices the post-oil decline is less dramatic (Table 11). Table 11: GDP Growth (Percentage) A. Numeraire Imports Years Oil Non-oil Total 1986-1992 9.7 6.6 7.8 1992-1996 - 2.5 5.2 2.6 1996-2002 - 2.8 4.3 2.2 2002-2012 - 9.0 4.1 2.1 2012-2020 -13.8 4.3 3.5 B. Constant Prices Years Oil Non-Oil Total 1981-1986 9.1 7.4 8.0 1986-1992 6.5 8.7 8.0 1992-1996 -5.1 7.7 4.2 1996-2002 -5.1 5.6 3.8 2002-2012 -10.8 4.8 3.7 2012-2020 -15.0 4.3 4.0 Finally the energy side. Oil output under this scenario will peak 10 years from now at around 70 m ton but will decline gradually thereafter (in volume terms). Energy use is however predicted to rise rapidly at slightly less than 10% over the next decade (in value terms), although the growth rate - 106 - gradually declines from there on to settle down at approximately 5% in the next century. In volume terms domestic energy use will grow at 7.2% over the next decade, as our oil price scenario implies an average growth rate somewhat below 2.5% over the next 10 years. The 7.5% growth rate for energy use in volume terms is more than a full percentage point below the corresponding rate for GDP, mainly because traded goods production grows several percentage points faster than non-traded goods production and has a much lower energy input coefficient (the difference is due mainly to the inclusion of electric power generation in NT production). Under this scenario Egypt will become a net importer of energy at the end of the century. Finally what may be the most important part of this analysis, the analysis of the intertemporal project analysis parameters the ARI and the CRI. Two major factors dominate the behaviour of the Period to period ARI. -1. The first one is that absorptive capacity constraints rule out an immediate jump of the level of investment from the current level to the one indicated by a proper intertemporal optimizing analysis. This constraint will be relaxed over time which in turn leads to a value of capital (in terms of "real consumption") that falls over time and accordingly a higher ARI now than later on, after the Absorptive Capacity constraint ceases to be binding. The existence of Absorptive Capacity constraints (and their anticipated relaxation over time) indicates that less investment should be undertaken than without those constraints; it is very important however that this does not simply imply the use of a higher ARI for all projects. This would not only lead to less investment but would also unduly bias its composition towards quick- yielding projects. What such a situation will lead to is a high period-to- period ARI initially but a lower one later on. It can not be stressed enough - 107 - that the use of simple Internal Rate of Return criteria in such circumstances is extremely misleading. Consider an extremely simplified example. Let us assume two projects, both of which occur outlays up front of $1 million. Project one yields all its benefits two years from now. Those benefits are $1.15 million, giving an internal rate of return of 15% (Fig. 29). The second project has the same cost structure but yields no benefits until 5 years from now, to the amount of $2.01 million. This project also has an internal rate of return of 15%, so according to IRR methodology they would both be accepted or rejected depending on whether the "Cut-off rate" is below or above 15%. At the current IBRD rule of thumb (10% cut off rate) both would be accepted. Now consider what happens when the time path of the discount rate changes over time in a manner similar to what we described before: Year: 1 2 3 4 5 ARI1 10 10 10 10 10 ARI2 17 15 10 10 10 Of course at the flat ARI1 pattern both projects have a positive Net Present Value (NPV). At the second ARI pattern however project one has a NPV of - $17,000 and should therefore not be undertaken, while the second project has a NPV of $120,000 and should therefore still go through. Under the second scenario project one should be rejected and project 2 accepted; nevertheless, they have the same internal rate of return of 15%. Clearly IRR calculations give misletading information when comparing projects with different time-phasing of net benefit streams. - 108 - Figure 29 Costs and Benefits of Two Hypothetical Projects 6 US* I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2~~~~~ ~~ _____ i n -11 -- - 109 - The value of capital goods in this model has two components: the first major component is simply the discounted value of all future output to be produced by the "marginal" machine. The second is related to our assumption that the Absorptive Capacity constraint is related to past capital accumulation. More investment now will accordingly relax the AC constraint in the future. Accordingly, if the AC constraint is anticipated to be binding tomorrow, capital formation will be more useful than when it is not because of this externality. In our model all capital formation has this property; in the real world this is of course not the case. Accordingly, we separated these two components of the value of capital and also derived an ARI that does not incorporate this e.xternality (*RI in Fig. 30). Clearly, ARI > *RI because the value of this externality declines over time. This does NOT mean that when two projects with identical cost-benefit structure but the first one with this externality (road building project) and the second without (candy factory) are compared, the first one has to be rejected because of the higher ARI. One should also iLncorporate the higher value of its output: during the period of binding capacity constraints capital goods (roads) are more valuable than consumption goods (candy). It is easy to show with the numbers coming out of this exercise that when two of these projects have exactly the same net benefits over time the road project will nevertheless have a higher Net Present Value the higher ARI and similar net benefit streams not- withstanding. The valuation difference dominates the ARI difference. In our case the value of capital including this externality (roads) in terms of capital excluding this externality falls from 1.5 to 1 between now and 1990. - 110 - The second factor determining the ARI is the existence of a major temporary component in current income (oil revenues). As we discussed before in the discussion of the optimal savings rate, this leads to temporarily high savings and investment programs. The implication of that is, once again, not a lower ARI for all projects, but a period-to-period ARI that is low during the period of high but declining oil revenues and high after revenues have stopped flowing in.A "flat" ARI pattern would again produce a biased composition of aggregate investment, in this case against quick-yielding projects. The combined result of the gradually relaxed AC constraints and the decline of oil revenues gives a U-shaped ARI pattern (Fig. 30). The CRI pattern does not follow the ARI pattern exactly because the relative value of capital in terms of the price of a "representative" bundle of consumption goods changes over time. Initially the value of capital falls faster than the value of an additional unit of consumption because of the gradual relaxation of the absorptive capacity constraints and the rapid accumulation there after so that the CRI is lower than the ARI during that period (see Fig. 30). After about 12 years however there are no significant changes in the relative value of capital goods in terms of consumption goods so that the CRI equals the ARI. After the effect of the binding absorptive capacity constraint has died out the CRI also follows a U-shaped pattern because of the currently high but declining pattern of oil revenues. The story is really the same as the one told for the ARI but seen from a different angle. Remember that the only endogenous component in the CRI is the growth rate of per capital consumption. This growth rate is falling initially, in line with the gradual decline in oil - ill - Figure 30 ARI, *RI and CRI in the Base Run ACCOUNTING RATE OF INTEREST (ARI) ...... .ACC. RATE OF INT. IGNORING ABSORPTION CAPACITY (*RI) …- CONS0UMPTION RATE OF INTEREST (CRI) 0.20- 0.20 0.1 0.15 0.12 -0.12 0.08- ~~~~~~~~~~0.08 0.04 - T-0.04 I gs0. a 1995.0 2010.0 2025.0 2040.0 TIME - 112 - revenues, making for a gradually declining CRI. After the oil has run out the decline stops and the growth rate gradually climbs back up towards it steady state value, which is reflected in a gradual increase in the CRI in the second half of the planning period. 5.2.3 Variants on the base run (a) No Absorptive Capacity Constraints The impact of the absorptive capacity constraint on the ARI is most clearly seen by comparing the base run time pattern of the ARI with that obtained in the first variant presented: this is a run made under the assumption that there is no absorptive capacity constraint at all, resources devoted to capital formation lead one for one to increases in the capital stock. The ARI is more than 8% lower in the first period, a difference that takes about 10 years to fade away. From there on the ARI is the same. Of course Tobin's q, the value of capital over its reproduction costs, is always equal to one in this run. There are some relative price changes between capital and consumption, but not enough to drive a significant wedge between the CRI and the ARI. Initial investment is unbelievably high, the optimization exercise indicates that in the absence of absorptive capacity constraints a full 54% of GDP should be devoted to capital formation in the first period. Absorptive capacity constraints bring that down to a more realistic 29%. The value of capital is naturally much lower initially, in fact only half of the value in period 1 in the base run. One third of that difference is caused by the lower future marginal productivity in period 2 and later (the capital stock is bigger in those periods because of the high initial investment). The remaining two thirds of the difference is due to the - 113 - fact that new capital has beneficial effects in the base run because it loosens up future AC constraints, benefits that are absent in this variant. GNP growth in the first 6 years is more than 2% higher each year because of the very high initial level of investment but later differences are minor. Consumption grows nearly three times (2.8 to be precise) faster but from a much lower base, first period consumption is 28% below the base run value. Of course a substantially higher level of exports is needed to cover the import requirement of the accelerated investment program, non-oil exports are 44% higher than in the base run case. In the short run oil output cannot be increased much, because of capacity limits; accordingly, the extraction pattern is very similar. In the absence of binding capacity limits on oil output in the first coupLe of periods extraction would of course have accelerated. (b) Different oil reserve scenarios Consider now some variants where the AC constraints are reimposed, but where the reserve level is varied. The reserve level used in the base run is the best current estimtate, but considerable uncertainty is attached to it. Explicit incorporation of that uncertainty in the optimization will not be attempted here. Instead we present some runs under alternative scenarios. Consider for example the case where discoveries will fall off much faster, say that total discovered and un-discovered reserves are only 1 billion ton. Oil output will then peak earlier and at a lower level and the share of oil in wealth and of oil revenues in income will decline considerably faster (see Fig. 31 and FLg. 32). - 114 - Figure 31 Composition of Wealth, Low Reserve Case SHARE OF OIL RESERVES IN TOTAL WEALTH ...... . CAPITAL STOCK IN TRADABLES OVER TOTAL WEALTH 1.0- 1.0 0.5 0.8.. . 0.6 - * 0.6 0.4- - 0.4 0.2- - 0.2 0.0- -- 0.0 1980.0 I95.0 2010.0 2025.0 2040.0 TIME - 115 - Figure 32 Share of Oil, Oil Revenues and Exogenous Transfers in Total Income, Low Reserve Case BASE CASE ..... . LOW RESERVES D. 50 - 0.50 0.45- 0.45 0.40 -* 0.40 0.35 - 0.35 0.30 ~~~~~~~~~~0.130 0.30-\ OD 0.25 - ..* 0.25 0.20- - - 0.20 19B0.0 1995. 0 2010.0 2025.0 2040.0 TIME - 116 - Twelve periods from now oil will represent only 7.6% of total wealth as opposed to 18% in the same period in the base run. The capital stock in the T-sector will now constitute 65% of the total in period 12 instead of 56%, the base run value in that period. Oil revenues and "exogenous resources" will then make up only 28% instead of 33% of total income. Egypt will turn into a net importer 4 years earlier than in the base run case. GDP will grow much slower over the 1986-1992 period (4.8% instead of 7.8%; between now and 1986 the rate of growth is the same), but non-oil exports will have to grow considerably faster (nearly 4% points a year) during that crucial transition period (the second half of the next decade). The savings rate must be nearly 3 percentage points higher during the peak period of oil revenues 8-10 years from now because the decline will be much steeper. The real appreciation is somewhat smaller in the beginning, but the differences are small because of the flatness of the transformation curve. Of course more or less the opposite occurs when discoveries turn out better than expected. Under the high reserve scenario (total level 2.1 billion tons). Investment, after the AC constraint ceases to be binding, will take a percentage point less out of GDP initially (about 4 percentage points less when compared with the Low Reserve case). Egypt would remain a net oil exporter 10 years longer than in the Low Reserve case and 6 years longer than in the Base Run case. GDP would continue to grow at 8.3% during 1986-1992 before slowing down gradually. Oil would still make up more than 30% of total wealth (30.8%) two and a half decade from now, while oil revenues and exogenous reserves would make up 37% of total income at that time. Of course non-oil exports do not have to grow as fast as in the base case (2 percentage - 117 - points less in volume terms over the next 6 years). Accordingly, the terms of trade deteriorate slighltly less (2%) over the next 25 years. (c) Alternative oil price scenarios All this is based on the assumption that oil prices, after a flat period between 1981 and 1984, will grow at a rate that gradually declines from 2% to zero 60 years hence. Two alternative growth rate scenarios were considered. Under the Unfavourable scenario oil prices are lower during the time that Egypt is a net exporter, and higher thereafter (see Figure 17). The impact on extraction of oil is fairly predictable: there is a shift away from current towards future extraction. The initial value of oil reserves is considerably lower (11%). Total wealth falls a little bit less, with about 3%. GNP growth is substantially slower under this scenario, 7% p.a. instead of 9% over the 1981-86 period and 5.9% instead of 7.8% over the 1986-1992 period). Non--oil exports of course have to be higher and grow faster (3 percentage points over the first six years in volume terms) because of the shift in the extraction pattern. Initial consumption falls nearly 1%, a difference that increaLses over time commensurate with the 3% drop in total wealth. The real exchange rate (price of non-traded over traded goods) depreciates gradually under the combined impact of lower oil revenues and a supply response to lower oil prices based on the higher energy intensity of Non-Traded goods, but the differences never reaches more than 2%. The second alternative scenario is one favourable for Egypt: prices are high during its period as a net exporters and low (compared with the Base Run) during the periods in which they are a net importers (see again Figure 17). - 118 - The shift in extraction pattern now goes the other way of course, you want to be left with less oil when the growth in oil prices stops (Figure 33). Note by the way that we may have a zero growth rate in the real price of oil and at the same time a finite extraction rate with some oil kept in the ground because of the dependence of extraction costs on reserve levels: more oil left in the ground lowers future extraction costs. It is this link that leads to finite extraction rates even during the periods of zero growth in oil prices without capacity constraints. The new scenario leads to a small increase in wealth of nearly 2% and a comparable upward shift of the consumption pattern over time, but somewhat tilted towards the future. GNP growth is substantially higher than in the unfavourable case, and a little bit higher than the Base Run values (9.2% for 1981-86 and 8% for 1982-92). (d) Alternative utility function parameters The final two runs vary parameters of the utility function. First we made a run equal to the Base Run in all respects except for a significantly lower rate of pure time preference (2% a year instead of 4.5%). The most striking effect is on oil extraction (Fig. 34). Capacity limits on output are never binding, and, as theory would predict, a major part of oil extraction is shifted towards the future. In 2005 oil wealth still makes up 25% of total wealth, seven percentage points more than in the base run. Non-oil exports accordingly grow at a much faster pace initially (a stunning 21% in volume terms over the next 6 years, a Korea like performance) if oil production is slowed down that dramatically. To make room in world markets, several percentage points of further increases in competitiveness are called for over the next two decades. - 119 - Figure 33 Oil Extraction Path in the Base Run and in the Favourable Price Scenario BASE CASE ...... .FAVOURABLE OIL PRICE 80.0- -_______________________________ - B0.0 70.0 - -7, 60.0 60.0 50 0 6~~~~~~~~~~0.0 50.0 50.0 40.0- / \ _ 40.0 30.0- 30.0 20.0- @1 ~20.0 10.0 10.0 0.0- ;'s-i-~ ,_,5.0 2010.0 - 0.0 1 980.0 1995.0 2010.0 2025.0 2040.0 TIME - 120 - Figure 34 Oil Extraction in the Base Run and Low Time Preference Case BASE CASE ......... .LOW RATE OF TIME PREFERENCE 80.0- - G0.0 70.0- 70.0 60.0- 60.0 50.0 \.. 50.0 40.0 / 400 30.0 - 30.0 20.0- 20.0 10.0 - 0O.0 0.0- 0.0 19B0.0 1995.0 2010.0 2025.0 2040.0 TIME - 121 - The lower rate of time preference of course also translates into a faster pace of capital accumulation in the first part of the planning horizon (also cumulatively, the steady state capital stocks in period 30 are about 10% higher than in the Base Run). First period investment is 12% higher than in the base run, a difference that rapidly increases as AC constraints are relaxed; at the end of the planning horizon this is reversed. The savings rate out of net income starts off at 25.5%, 3% above the base run level, and reaches a peak of 46.5%, a full 10 percentage points above the corresponding base run values. The ALC constraint is binding for two more years and the higher pace of first period investment results in a further 6% of efficiency loss on top of the 6% lost in the first period in the base run (6% of total resources devoted to capital accumulation). The ARI starts out higher, the initial value of capital is very high because the AC constraint is tighter, and declines more rapidly because of the faster pace of capital accumulation (which pushes out future AC constraints), so that the ARI is higher initially. It falls quickly to a rate considerably below Base Run values, as one would expect given a rate of time preference difference of one and a half percentage point (Fig. 35). Of course the *RI (which is based on a value of capital that does not take into account the effect of current capital accumulation on future AC constraints) is uniformly lower than in the base run case. The net effect on GDP is a considerable smoothing of the growth rate. Total consumption growth is lower initially (first 12 years) but consistently higher from there on. In a final run we kept the low rate of time preference, but increased the curvature of the utility function substantially. The precise value of the elasticity of marginal utility with respect to per capita - 122 - consumption was chosen to insure the same steady state capital stock in the base run. This results in an elasticity of 1.017 instead of 0.5. Table 12: Growth Rate of GDP and Consumption in Base Run and runs with alternative utility function parameters GDP Total Consumption p=1.0404 p=1.0404 Year Base Run(*) p=1.0404 1-nr1.017 Base Run p =1.0404 1-rj=1.017 1981-85 9.0 5.3 8.9 4.5 0.2 3.9 1986-91 7.8 8.6 7.0 4.9 4.7 3.8 1992-95 2.3 4.8 3.2 4.9 6.3 4.7 1996-01 2.2 3.8 2.9 3.2 4.8 4.0 2001-11 2.1 3.2 2.9 3.0 3.9 3.8 2012-20 3.5 3.4 3.3 3.7 4.0 4.1 (*) Base run: 1-n = .50, p = 1.093 (4.5% on annual basis) The growth rates of GDP and consumption are even smoother now than in the low time preference case (Table 12). Oil extraction is slower than in the Base Run, but not as dramatically as in the previous run. In 2005 oil still makes up 21% of total wealth, closer to the equivalent Base Run value of 8% than the p=1.0404 value of 25%. The savings pattern is also intermediate between these two: higher than the base run (the savings rate out of net income starts out at 24% and reaches a peak of 41% before tapering off slowly to around 30%) but not as high as in the low time preference run. *RI, BASE CASE ARI, BASE CASE ...... *RI, LOW TIME PREFERENCE ....ARI, LOW TIME PREFERENCE 1.25 [ 1 0.25 D IU v 0.20 \ P .00-. . -5 0.15 0.07 - 0.06 1 2 TIME T1Mw 0.01 'Dli I 0 '.00- -~~~~~~ ~~~0.00 0_ __ __02_ _ __ __-__ _ __ _ 190.001.0 205. 2400 980. 0 995. 0 2030.0 202U5. 0 2040. T IME TI MI1 - 124 - 5.3 Foreign Borrowing In the experiments described above the current account has been kept exogenous; it follows such a time path that per capita debt is constant throughout. Accordingly optimal foreign lending or borrowing was not considered. Of course both capital and oil are assets that can be used to smooth consumption expenditure, but both these assets have obvious limitations if used for that purpose. In this section a foreign capital market is introduced, with a constraint on terminal debt that insures a final period debt - GNP ratio equal to the one Egypt has now. The base run incorporates the real world feature of a substantial wedge between borrowing and lending rates; for technical reasons that transition was made a continuous function of per capita commercial debt rather than a discontinuous jump at zero. The model is in all aspects identical to the one used in 5.2 except for the CA constraints, the period by period constraint of 5.2 are now replaced by an intertemporal budget constraint. Equation 10 of page 25 is replaced by the following set of equations: PW M + PW Mot < PW3t (E3t °t X3t) -tEt t - rt CD t1/Gt_l + CADt (1Oa) where rt CDt l/Gt_l reflects interest payments on beginning of period debt and CADt is the current account deficit during period t Current account deficits lead to higher debt: - 125 - CDt = CADt+ CDt_1 /Gtl (1Ob) Final period debt CDt has to be below a value CDt chosen in such a way that the period T Debt-GDP ratio equals the one at the beginning of the planning period. The cost of borrowing, rt, depends on beginning of period debt CDt_l and, in some runs, on the debt service ratio. Introducing the debt service ratio as an argument in the cost of borrowing function did not lead to qualitatively different results, so these runs are not reported here. So for the runs analysed here, lOc holds: rt = h(CD t) , lim h = r, .lim h = r (10c) CD+- CD+= r was taken to be 6% per year (keep in mind we are talking about the real rate of interest), and r as 3%. This spread between borrowing and lending rates is substantially larger than those observed in practice, but is meant to capture other, non-interest costs of borrowing such as commissions, front-end fees, etc. The precise functional form of h( ) is given in Appendix A (page IV, V). 5.3.1 The Base Run with Foreign Borrowing and Endogenous Capital Accumulation Capital accumulation is predominantly constrained by absorptive capacity constraints in the first few periods, not by availability of savings; accordingly the introduction of foreign borrowing does not lead to significant changes in the investment path and the ARI. The average cost of borrowing is - 126 - between 6 and 3 percent a year (real rate) where the high number applies to positive commercial debt positions and the low number to negative commercial debt positions (the real rate of return of net lending is assumed to be 3%). The difference for most variables is minor compared with the no borrowing base run, except for the time path of consumption and therefore, given production, such variables as non oil exports, etcetera. Consumption now is smoothed considerably, although not completely (Table 13). Table 13 Consumption Growth Rate with and without Foreign Borrowing Periods 1-3 4-6 7-8 9-11 12-16 17-20 No borrowing 3.5 5.7 6.1 3.8 3.3 3.7 Borrowing 3.2 4.6 4.7 3.3 3.1 3.7 The uniformly higher or equal growth rate in the no-borrowing case reflects the fact that initial consumption is substantially higher in the borrowing case (7% in the first period): no borrowing implies a higher growth rate from a lower level. An interesting feature is that the consumption growth rate is about 1.3% higher during the period of gradual real appreciation than during the periods with a real depreciation. This is related to the fact that foreign loans are denominated in foreign traded goods while consumers also consume non-traded goods. A gradual real appreciation then implies that the - 127 - economy is borrowing in terms of goods that are becoming cheaper over time and accordingly should speed up consumption now and slow it down in the future (see Martin and Selowsky (1981) and Dornbusch (1982), for an elaboration of this point). The reverse holds during a gradual depreciation. The increase in initial consumption is larger in traded goods than in non-traded goods, so not surprisingly the volume of non-oil exports is substantially below the no borrowing base run case (31%) although it grows substantially faster in the first six years (25% instead of 14.7%). The lower volume of non oil exports should not come as a surprise, there is now an alternative form of foreign exchange (borrowing). A more macro-view would be that borrowing allows higher consumption now (which of course will result in lower consumption in the future) given that oil revenues are increasing initially and borrowing against future income is now possible. The time path of per capita debt is straightforward, an increase initially before oil income reaches its peak (borrowing against future income) with nearly 30% betweea now and 1900-1990, a decline thereafter when oil revenues peak (per capita debt falls off 10% from its 1900-1990 peak level), after which an increasingly fast accumulation of debt takes place as oil revenues dwindle to finally hit its exogenously imposed terminal constraint. That constraint is set such that the terminal debt/GNP ratio is the same as the initial one. Finally, the smoothing effect of borrowing can be seen by comparing the CRI in the case with and without borrowing: the "dip" around the period where the absorptive capacity constraint ceases to be binding has disappeared (Figure 36). - 128 - Figure 36 Various Discount Rates, Base Run with Optimal Borrowing CONSUMPTION RATE OF INTEREST (CRI) - ACCOUNTING RATE OF INTEREST (ARI) ..... ACCTG. RATE OF INT. IGNORING ABS. CAP (*RI) 0.20o- 1 I l 0.15- 1.15 9 .. I .O- ...v 1.30 1960.0 1995.0 =1W.o 202s.a 2040.0 - 129 - 5.3.2. An Alternative Oil Pice Scenario The "Base Run" oil price scenario used in most of the analysis up until now was rapidly made obsolete by the events of late 1982-early-1983.We therefore present an alternative run, with the most general version of the model, under the assumption of a substantial downward shift in the entire schedule of oil prices (Figure 37). The welfare effects of such a relative price change (oil is cheaper in terms of everything else) depend on your net export position in oil; the situation is complicated in the case of Egypt because it is exporter for the next decade at least, but will turn into a net importer after that. The question then comes down to whether the benefits of lower oil prices during the future net importer phase outweigh the losses during the current net exporter phase. An alternative but equivalent way of looking at it is by looking at the gross flows (oil output and oil use) and see what changes in the value of those flows do to the value of the assets making up Egypt's wealth: oil in the ground, physical capital and labour stock. The value of oil in the ground, according to the model, falls with nearly 25%, the rent on a barrel of oil kept in the ground falls from $14.60 to $11.16. On the other hand lower oil prices lead, ceteris paribus, to more value added left over to pay capital and labour. The value of capital (discounted value of all its future marginal products) indeed increases with slightly above 5% as derived from the dual of the opt:Lmal solution under the new price scenario. The entire path of the marginal product of labour ("shadow wage") also shifts up around 5%. The net effect can not be signed on a priori grounds; the numerical outcome indicates that future benefits in fact dominate current - 130 - Figure 37 Different Oil Price Scenarios - BASE CASE ----ALTERNATIVE OIL PRICE SCENARIO (US$/61) ao.0 5~~~~~~~~~~0.0 7.0- 70.0 60.0- 60.0 50.0 , - 50.0 n 40.3- / ,' ~/ 40.0 30.0- o 30.0 20.0- a I - 20.0 1§O. 0 199S.O 2010.0 2025.0 204e.0 - 131 - losses: the discounted utility of current and future consumption in fact increases under the low price scenario, although not much, with 1.5 percentage point. The higher marginal productivity of capital at lower oil prices calls for more investment although that increase is kept in check during the initial few periods by ithe absorptive capacity constraints. Over the first six years investment should, at lower oil prices, be about 2% higher than in the base run, a difference that will increase as absorptive capacity constraints cease to be binding further in the future. Current GNP faLlls by one percentage point (in fact the growth rate will also be lower by an average 1.5% over the first six years), and investment should go up; should one conclude that since wealth has in fact increased a little bit, aggregate consumption should also rise, necessitating an increase in the optimal CA deficit and an increase in foreign debt? Higher or equal wealth coupled with lower current income would argue yes; however, the gradually increasing loss in oil revenues leads to a gradually increasing real depreciation in comparison with the base run, indicating that loans taken out now will have to be paid back in traded goods that are becoming more and more expensive (see our discussion of the real cost of borrowing and gradual relative price changes in the previous paragraph). This argues for less borrowing rather than more and a "tilt" of the consumption pattern towards less consumption now and more in the future, and accordingly less foreign borrowing and a lower CA deficit. The numerical solution indicates that the second ("cost of borrowing") effect dominaLtes, the increasing real depreciation over the first couple of years calls for a slowdown in consumption and initial decrease in - 132 - the CA deficit that is gradually reversed over time. From the period where the real exchange rate starts catching up again, (the level remains permanently below the Base run path, but the difference starts to decrease again from 1990 onwards), foreign borrowing increases above the base run level, as the real cost of borrowing and wealth effects now work in the same directions. See Figures 38 and 39. The accompanying optimal current account deficits are presented in Table 14. Table 14: Optimal Per Annum Current Account Deficits Under the Base Run and an Alternative Scenario Oil Price Scenario (Billion US$) Case 1981-82 1983-84 1985-86 1991-92 Base Run 3.46 2.22 1.88 1.81 Alternative 2.05 1.60 2.48 3.27 Oil Price Scen. In looking at those numbers one should take the indicative nature of such results into account, the patterns produced by these runs will come out under a wide variety of assumptions, but specific numbers are rather sensitive to things we do not know much about. Finally, maybe it is worthwhile to point out that the lower oil prices call for a significant shift towards non-oil exports, which should be substantially higher in the low oil price scenario (nearly 30%), necessitating both a decline in consumption and a shift away from traded goods (hence a real depreciation). Clearly the composition of investment also has to change in favour of capital in the traded goods sector. ,-133 - Figure 38 The Real Exchange Rate (NT/FT) Under Different Oil Price Scenarios ------- REAL EXCHANGE RATE IN THE ALTERNATIVE PRICE SCENARIO AS RATIO OF BASE CASE VALUES 1.O3JS- _____.___-____________ 1.000 -_1. O 0.99- 0.995 0.990 0.990 D. 9S - 0. 9g o.s«,-- - a.3 0.9?5- - 0.975 .9 - 0.970 1953.13 290.0 2200.0 2010.0 - 134 - Figure 39 Optimal Borrowing Levels (Comm. Debt) Under Different Oil Price Scenarios | - - - -DEBT INT ALTERNATIVE PRICE SCENARIO | ~~~AS RATIO OF BASE CASE VALUES 3.5 3.5 . ~~~~~~~~I .1t 3.0- 3.0 2.5- 2.5 2.0 - 2.0 l.5 / ~~~~~~~~~ 1~~.5 .0DI / -1.0 0.51 0.5 C.0- I I - 0.0 0SSO.0 2CO0.0 20l0.0 - 135 - 6. Conclusions We have constructed and used a highly aggregated intertemporal optimizing model to explore the sensitivity of optimal policies with respect to such exogenous variables as world oil prices, level of reserves, energy efficiency, rate of time preference, etc. under a variety of assumptions about capital accumulation and foreign borrowing. Those assumptions range from both being completely exogenous to both being endogenous and incorporated in the optimization process. Rather than presenting a summary of the many experiments performed, it may be more useful to devote this section to discussing two general classes of results. The first refers to a set of Egypt-specific policy conclusions, in the form of the time patterns obtained for such aggregate variables as oil revenues, savings, investment and foreign debt, which are robust across all scenarios for exogenous parameters and prices. These patterns could serve as inputs in medium term, dissaggregated consistency models. The second class of results relates to the practice of shadow pricing and project evaluation: we obtain implications for the realtive price of traded and non-traded goods, the relation between relative price changes and foreign borrowing, the path of the accounting rate of interest, and the pricing of natural gas which are of substantial practical significance but very hard to derive without a model like the one used here. 6.1. Some Robust Policy Conclusion The long-run optimal growth model developed in this paper was used to analyse some of the fundamental macroeconomic challenges implicit in Egypt's economic structure. While there is a great deal of uncertainty - 136 - relating to oil and gas discoveries, world prices and other exogenous variables, such as long term growth of remittances, it is nonetheless possible to derive some robust conclusions that can serve as guidelines to development strategy in Egypt: (i) Revenues from oil and gas are a very special kind of income. They imply decumulation of wealth unless they are converted into other productive assets. It is misleading not to differentiate GDP in the non-oil economy from this "income" derived from resource depletion. Workers' Remittances, Suez Canal earnings and foreign aid are also "exogenous" resources that should be viewed as a "transfer" to the economy not based on domestic production. In an economy where a large function of current expenditure is financed by decumulation of an exhaustible resource or out of transfers that may not be available in the future, there is a very serious need for structural adjustment necessitating a high rate of investment directed towards the traded goods sectors. (ii) An optimal growth strategy will therefore require very high investment rates during the 1980s. The analysis suggests that investment rates between 30 and 35 percent of total resources are desirable targets for economic policy. A high proportion of this investment should be directed into export expanding and import-substituting activities. Over the next two decades the growth of non-oil merchandise exports will have to be in the 9 to 13 percent range and imports will have to grow more slowly than domestic production. These are difficult targets to - 137 - achieve and sustain over an extended period. They reflect a very substantial structural transformation of the economy that must begin very soon and will reach its most difficult stage some time during the 1990s. (iii) While Egypt may experience short term macroeconomic difficulties of a "cyclical" nature in the mid-1980s, the contribiution of the oil and gas sector to total resources will continue to be large for the next 5 to 8 years. The big challenge lies further in the future, when some time during the 1990s the share of "exogenous resources" will start showing a rapid and sustained decline as Egypt adjusts to a new type of economy where growth comes mainly not from increases in oil and gas production but from productivity growth and factor accumulation in the domestic non-oil economy. (iv) Egypt's foreign debt is not particularly high in view of the cost of borrowing on the one hand and the marginal productivity of capital, if well deployed, on the other. Current Account deficits in the range of 3% of GDP are indicated as reasonable in the light of our experiments, although they should decline and may be even reversed during the peak period of oil revenues in the early nineties. 6.2 Shadow Prices and Project Evaluation While the results discussed in 6.1 are very Egypt-specific, there are other aspects of our results which seem of much wider relevance. (i) When analysing relative price adjustments required by the long- - 138 - run structural adjustment process, it is important to distinguish between the relative price of domestic tradables vis-a-vis domestic non-tradables, and the relative price of domestic goods in general (tradables and non-tradables) vis-a- vis goods produced by the rest of the world. The former relative price, which we could call the internal real exchange rate, will have to change to bring about the required expansion in production of tradable goods only to the extent that this sector uses a mix of factors of production very different from that used to produce non-traded goods. The external terms-of-trade issue should be distinguished from the tradables/non-tradables relative price issue. Even if there were no change at all in domestic relative prices, there would still be a need for a depreciation in the external terms of trade, or a decline in the relative price of domestic goods in terms of foreign goods, because exports cannot expand without an accompanying increase in external competitiveness. While the first kind of real exchange rate adjustment depends on the "degree of difference" between domestic tradables and non-tradables, the extent to which an external terms of trade adjustment is needed depends on the extent to which a country must accept lower net export prices (net of transportation costs) when it wants to expand its world market share. ii) One of the interesting results of the foreign borrowing analysis is the interaction between changes in the real - 139 - exchange rate and optimal Current Account deficits. A gradual depreciation (traded goods becoming more expensive in terms of non-traded goods) makes foreign borrowing, for any given rate of interest, more expensive in terms of consumption foregone to repay the loan. In one particular case, the current account response to a downward shift of the entire schedule of current and future oil prices (in effect the shift from the "median" predictions mid-1982 to the median prediction mid-1983) was reversed because of this effect. Lower oil prices are bad for Egypt now (they are net exporters) and good in the future (when they are net importers). On balance this shift made them better off. Also current income went down, so any one-good borrowing model would have implied a larger CA deficit early on. However the decline in oil prices also necessitates a steep depreciLation over the next few years which was shown to reverse that conclusion: the optimal response involved a smaller, not a bigger CA deficit. (iii) Another important conclusion relating to the shadow price covering from the optimizing model, relates to the structure of interest ratEts and particularly the time path of the accounting rate of interest (ARI). The common practice of using a constant period to period ARI was shown to be seriously misleading in these circumstances. If absorptive capacity constraints are tight now but are gradually relaxed over time the value of capital declines more rapidly than it will when - 140 - they cease to be binding, indicating that the period to period ARI should be high now but low later. Use of a high ARI throughout would produce an undue bias in favor of quick- yielding projects. It cannot be stressed enough that Internal Rate of Return criteria (current bank practice!) are extremely misleading in such circumstances. A similar effect is caused by high but temporary oil revenues. This calls for a low ARI now, but a higher one in the future, when revenues tighten up. Use of a uniformly low ARI would now unduely bias against quick yielding projects. All this has its counterparts in the valuation of capital goods, another useful output one obtains from a model like this. (iv) The final price of interest is the shadow price of natural gas. When exports would not be viable (the net back value assuming perfect substitution with oil in the user countries would be far too low) and domestic substitution possibilities are exhausted, a wedge may open up between the shadow price of gas and its fuel-oil equivalent. The way that wedge moves depends on the intertemporal pattern of energy demand and gas supply and the accounting rate of interest. (The ARI influences the extraction pattern and therefore the supply at any given moment). Without binding supply constraints, the opportunity cost of producing gas, and hence its price, is just the cost of extraction/distribution plus the rent to gas reserves. Low gas reserves reduce the size of the wedge and - 141 - can eliminate it completely. Rapid increase in domestic demand for energy also tends to close the wedge, by pushing gas demand and extraction to its physical limits. - 142 - Footnotes _/ It is best, in this context, to think of human resources as real capital assets. Human capital can be increased by real investment, in a manner similar to physical capital formation. 2/ If nN and N are the compensated price elasticity and income elasticity for nontraded good, and eN the supply elasticity, the change in the real exchange rate, PT - PN' when there is a transfer of F as proportion of GNP is given by - NF/(rN + eN) * It can be shown that when both factors are mobile, eN is directly proportional to the average of the elasticities of substitution and inversely proportional to the square of the difference in labor intensity in both sectors. 3/ For the exact specification see Appendix A. 4/ That is f4(K4t) = min lb4K4t, H4tI* 5/ The reason is that we do not want future generations to be penalized for living in a more populated society: this formulation gives equal weights (before discounting) to future and present representative families. 6/ The steady state is reached after oil and gas reserves have been exhausted, of course. It also requires a constant real price for oil after T, which we assumed in the base case to be $73. The steady states values were recalculated as needed for different scenarios. Their precise definition is spelled out in Appendix D. 7/ As we mentioned before, n is a function of the volume of exports and is very large for exports below a "normal" level. That normal level - 143 - itself grows over time. See Appendix B for the exact functional form. 8/ See Squire and Van der Tak, (1975). 9/ That is Q3,t = P3,tx3,t at/R 3,t-1 10/ Note that with constant costs of extraction it would seems that case (ii) above could not hold for more than one period. On the one hand we know from Hotelling's rule (equation 13a in section 4.3) that the rental value of gas grows at the rate of interest, so the price of gas net of extraction costs should do that too as supply constraints earn zero rent in this scenario; on the other hand the price of gas should also equal its fuel oil price equivalent in case (ii). At constant discount rates either one of these conditions cannot hold for more than one period unless oil prices net of gas extraction costs grow at that constant rate, an implausible scenario. In our model however the rate of growth of the unit value of reserves is an endogenous variable through the dependence of the discount rate on relative price changes (see also section 5.3 for a discussion in the context of foreign borrowing). That allows for more flexibility than in traditional one (final) good optimal extraction models. 11/ It is misleading to use a single ARI in an economy facing major future changes in income streams (exhaustion of oil). We will discuss the pattern of the period-to-period ARI(t). To discount back from t1to 0 one forms a discount factor ((1+AR(1) (1+-AR(2))..(1+AR(t1)). - 144 - References Buiter, W. and D. Purvis (1983), "Oil, disinflation and export competitiveness: a model of the Dutch disease," in J. Bhandari and B. Putnam (ed.), The International Transmission of Economic Disturbances under Flexible Exchange Rates (MIT Press, forthcoming). Corden, M. and J.P. Neary (1982), "Booming sectors and De-industrializatin in a Small Open Economy," Economic Journal, 92 (December). Dixit, A. H. (1976), The Theory of Equilibrium Growth, (Oxford U. Press, 1976) Dornbusch, R. (1983), Real interest rates, home goods and optimal external borrowing, Journal of Political Economy, 91 (February). Martin, R. and M. Selowsky (1981), "Energy prices, Substitution and Optimal Borrowing in the Short Run," World Bank Staff Paper #460, forthcoming, Journal of Development Economics, 1983. Squire, L. and H.G. van der Tak (1975), Economic Analysis of Projects, The Johns Hopkins University Press. van Wijnbergen, S. (1980), "Inflation, Employment and the Dutch Disease in oil exporting countries: A Short Run Disequilibrium Analysis," World Bank mimeo, forthcoming, Quarterly Journal of Economics, 1984. (1981), "Optimal Capital Accumulation and the Allocation of Investment between Traded and Nontraded Sectors in Oil Producing Countries," World Bank mimeo. World Bank (1983), Arab Republic of Egypt: Issues of Trade Strategy and Investment Planning. - 145 - Appendix A: Summary of the Egyptian Model 1. Variables (i) Primal (Activity Level) Cit = Final consumptions good i Xit = Gross domestic output of good i Kit = Capital stock used to produce good i Lit = Employment in sector i Eit = Exports of good i Mit = Imports of good i Ft = Net exogenous inflows of foreign exchange (aid, remittance, etc.) Zit = Intermediate use of good i Yit = Effective investment in sector i H3t = Maximum procluction of oil per period H4t = Maximum production of gas per period It = Total investment * = Level of investment at which absortion capacity becomes a binding constraint HEt = Level of exports at which price starts to decline ut = Utility level (i.e. real consumption) in period t R3t = Reserves of oil at the end of period t R4t = Reserves of gas at the end of period t CDt = End of Period Commercial debt CADt = Current Account Deficit - 146 - DSRt = Debt Service Ratio in period t rt = Cost of Borrowing All the above variables are measured in per capita terms. Lt = Labor force in period t (ii) Coefficients aij = Unit requirement of good i to produce good j PWit = International price of good i Pi = Substitution parameters 6 = Discount rate = 1 + Elasticity of the marginal utility of income g = 1 + rate of growth of the labor force gi = 1 + rate of labor augmenting technical change, i = 1,2 ai = Share parameters in the production functions bi = Capital-Output ratio in sector i Si = Share of sector i in the production of new capital di = I - depreciation rate n = price elasticity of exports (iii) Dual (Shadow prices) xt = Marginal utility of real consumption Pit = Unit value of final good i vit = Unit value added in sector i - 147 - rit = Marginal Productivity of Capital in sector i qit = Unit value of Capital Goods in sector i S3t = Unit value of reserves of oil S4t = Unit value of reserves of gas PIt = Unit value of Investment goods Wt = Marginal Productivity of labor The sectors are: 1 - Traded Goods (exported and used domestically) 2 - Non-traded Goods 3 - Oil and gas sector 4 - Gas sector (measured in tons of crude equivalent) 5 - Non Competitive Imports Sector 6 is used as an aggregation index for sectors 1 and 5, defining a composite importable good, which is used in consumption and production. 2. Objective Function T c v = E 6 tUt t=1 (At) Ut= (y C - +Cu (l-y )C u u (1) t 2 2t 2 6t 3. Constraints (i) Production functions (V t) > ) lt-p I l -pi -,/p2 (vi)>0 it < A1(a~Klt + (1-oti~)(Lltg) I) (2) - 148 - -P2 ~ ~ )t )P2 )2(3 (v2t °O) x2t < A2(a2K2t + (1-a2)((l-Llt )g2) (v3t a 0) X3t < Min {K33t/b3 H3t (4) (v4t >O) X4t < min {K34t /b3, H4t} (5) K33t 34t 3t (ii) Material balances (pit 0) Zt + Eit + SIIt xit (7) (p2t ) c2t + s2It + a21xIt + a22x2t + a23 3t 2t (8) (p3t °) E3t +a31xIt + a32x2t 4 x3t + x4t (9) (PGt 0) X4t - a41Xlt a42X2t 0 (10) -P6 -6 -'P (p6t °) 6t 61 it a62x2t + a63x 3t < A6a6Zit + (l-a6)Z5t 6 (11) (pt 5t 5 tt S t Iit+ 3tE3t + Ft 6(R3,t-l)3tX3t 12) - 149 - In the version with endogenous foreign borrowing, (12) is replaced by (P5t °) m t tElt + L3t(E3t -e(R3t-.1)x3t) + Ft - rtCDt_ /g + CADt (12a) rt+l = ELIBOR(t)/(l+A ( -A1CDt))+ A2 DSRt /Min(.00l,DSUB-DSRt) (12b) is the cost of borrowing; ELIBOR(t) is the LIBOR rate (6% p.a. in section 5.3); p ) is the Cumulative Normal Distribution and AO, A1, A2 are parameters (equal to .0:3, .7 and 0. in the runs reported here). The Debt Service Ratio (DSRt) is defined as real amortization (considering a 5% rate of world inflation and an average maturing of loans of 20 years) plus interest, divided by total export revenues. The change in the stock of Debt from period to period is given by: (q7t °0) CDt = CDt-1/g + CADt (12c) since, CDt' as all other variables in our model is defined in per capita terms. (iii) Investment (qt ° 0) Kit diKi,t_1/g + Yi t-1/g i = 1,2,3 (13) 3 It if I ( Y (PIt ° 0) X Yi = I t t (14) i=1 g(It, Yt t >t * The funtions Ytand g are defined in Appendix C. - 150 - Rt0 R R X (15) (qRt) 3,t-1 3,t/g 3t qG R =R /g (16) (Rt) R4 1 t = R3/g (iv) Terminal Conditions (qiT )iKiT iT i = 1,2 (17) iT > O) KiT ~KT The steady state capital stocks KiT are defined in Appendix D. With endogenous borrowing, there is also a condition on terminal debt: (q 7T 0) CDT 4 CDT (17a) where CDT is a value which produces approximately the same Debt/GNP ratio that Egypt had at the beginning of the planning period. 4. First Oder Condition for Maximization The Lagrangian function for this problem is: T -Pu -PU 1/ t=l t t[Ut - (Y2C2t + (1-y2)C6 u 2 -Pi -Pi -1/p1 l vt t- 3tK3 b 3-v 2H3t I- v, vit[Xit-Ai(ai Kit + (1-ai)Lit ) ] - [v3tX3t - v3itK33t 3 32 3t] - [v4tX4t - v4ltK3t/b3 -v42tH4t] - 151 - Plt[Zlt+Elt+Slt Xlt: P2t] C2t+S2 t+a2xit +a22x 2t +a23x 3t+a24x4t-X2t] 33t [E 3t+a 31 Xt+a 32X2t-X3tjX4t - PGtI[X4t-a 41Xlt-a 42X2 ] p6t[C6t +61 Xit 62x2t 63 3t 64X4t -p 6 p l6 -e(E-R -A6(a6Zit +(l-a6)Z5t ) ] p5t[Zst+ssIt-1tE - 3t(E3t (E3t e(R3,t-l )X3t Ft] 3 0 itK it[ diKit-/g Yi,t-1/g] 3 * (pitX Yit PIl,t It PI2,tg(,t1Yd)] s3t(R3,tR3,tl /g -FX ) - s4t(R4t - R4,t-1/g+ 4t )i + qi i=I 3 The dual equations (first order conditions for maximization) are: (i) Allocation of consumption 1+pu (C2t ° 0) XtY2(UIt/C2t) P 2t (18) 1- p (C6t > 0) At(1-Y2)(Ut/C6t) 4 P6t (19) (ii) Value added (Xit) vlt = - p2ta21 - p3ta31 - p6ta61 (20) - 152 - (X2t) v2t P2t - p2ta22 - p3ta32 - p6ta62 (21) (X3t) v3t P3t - p2ta23 - p6ta63 - q3t P5t n3t t (22) (X4t) v4t P4t - p2ta24 - PGta64 - q4t (23) where P4t ' p3t - PGt (24) (iii) Optimal production conditions (Ll) wt = v (1ai)(Aigi) iLitx i - 1,2 (25) it (Kit) r A (-X)l Pi i -1,2 (26) it ~ ~~t itiAi' Kit v3~~~~ =v31~~~~~~~ ~(27) 3t ' 31,t +v32,t (7 -V +V ~~~~~~~~~~~(28) v 4t w 41,t 42,t (K3t) r3t - 31tb - V41 l/b4 (29) (iv) Value of the stock of capital and Value of Reserves (Kit) qit rit + diqi,t+l/g i - 1,2,3 (30) (R3t) 53t 83 t + p56l3tE36 t (31a) - 153 - where 0wt 3 a t,X3t 3t is the increase in the share of foreign oil companies as extraction costs increase due to depletion of reserves (R4t) 4t 84,t+l/g i = O,G (31b) (v) Value of new capital (Y it °O) qit/g+ PI2,t+1 pIt i 1,2,3 (32) with PIt PIIt + pI2,t (33) and gt+l ' 9(It+i' Y- the absortion constraint function defined in Appendix B. (it > 0) PIJL,t + PI2,t agt/Iat < Slpit + S2P2t + S5p5t (34) (vi) Exports P5t if Elt < HEt (Elt) Pit (35) pst(Elt/HEt) 1 otherwise. (E3t) P5t 3t P3t (36) (vii) Aggregation of imports and traded goods (Zit ° 0) '6p6tA6 (X6t/Zit) i 4 pit - 154 - p6 ~l+p6 (Z5t 0) (1-a )p A6 (X /Z ) 6 < 5t (38) (Z 6 6t 6 6t 5t 5 5. National Aggregates Per capita gross national product is given by any of the following expressions: t it lt 2t 2t (v3t + S3t )X3t + (V4t + s4t)X4t + (p5t1IIt - pit)Elt Y = (r tKltr 2tK 2t+r 3tK3t)+wt+v32,t 3t+S3tX3t+V42,t H4t 4tX4t+(p5tt-pit )Eit Yt= (P2tC2t+P4tC4t) + (slPlt+S2P2t + 5P5t)It + (p,tE,t + P3tE3t - PstM5t) The first expression is the usual definition as sum of value added in all sectors plus rents (plus the implicit optimal tariff to exports). The second expression defines income as factor payments to capital, labor and rents in the oil and export sector. Finally we show the decomposition in consumption, investment and current account surplus. - 155 - Akppendix B: Calibration of the Model Table 1 presents the basic data about the Egyptian economy in 1981/82 which was used to calibrate the parameters of the model 1. The input-output coefficients are obtained as the ratio of inter- mediate use to gross output (e.g., aNT = 2.617/16.056 = .163, etc.), and they are: Y .163 .205 10013 (aid) = .324 .601 0. .382 .245 .0017 2. The share and scale coefficients for the production functions were obtained from estimates of the (constant) elasticity of substitution and initial _P _P ~-1/ p factor shares as follows. Given the production X = A(aK P+(1-a)L ) if S = WL/VX is the share of labor in value added (V is the value added coefficient W the wage rate), we have cV(l-a) = (K/L)PS/(1-S). The results are: Table 1: Base Year Economy TMillion of Egyptian Pounds Oil E Total Net Gross Unit Sector Traded Nontraded gas Composite Interaediate Export Consumption Investment Output Price Traded 0 0 0 13.225 13.225 2.219 0 .612 16.056 1. Montraded 2.617 3.218 0.47 0. 5.882 0. 7.773 2.041 15.696 1. Oil 10.656 0. 0. 32.5 .1728/ton (tons) 5.205 9.438 0. 0. 14.643 Gas 0. 0. 0. 2.8 .1728/ton Traded Composite 6.133 3.846 .060 0. 10.039 0. 10.872 0. 20.911 1. Inports 0. 0. 0. 7.686 7.686 -10.136 18.645 2.45 0. 1. Value Added 6.407 7.001 5.993 0. 39.3623 -4.3475 18.645 5.103 58.7628 Un Gross Output 16.056 15.696 6.07 20.911 Capital 20.002 17.779 .198 0. 37.979 1. Labor 5.828 5.233 0. 0. 11.061 Investment 1.755 3.258 .09 5.103 Notes: (1) Foreign exchange is converted at the rate of 1.25 US$/LE (2) Base price of oil is 30$/b, which translates to (30$/b) x (7.2b/ton) x (.8LE/$) - 172.8 LE/ton. (3) The external accounts were balanced as follows: Exports minus imports (-4.3475) + Remittances (2.4) + Suez, Non Factor Income, Direct Foreign Investment (2.1432) - Share of foreign oil companies (1.20976) + Grants and others (1.014) - 0. (4) Capital goods are estimated to be a fixed proportion mix of nontraded goods (40%), domestically produced traded goods (12x) and imports (48x). - 157 - Table 2 Substitution Share Scale Sector parameters Share parameter parameter (p/l) (S) (a) (A) Nontraded 1. / .5 0.55 .2625 1.4384 Traded 0. /1. .40 .4 1.3146 Composite -0.5/2. .36758 .4331/ 1.9647 Notes: 1/ Corresponds to share of imported goods in production of the composite. Note that although the share and scale parameters in Table 2 are estimated under the assumption of coefficient use of factors within each sector, there is nothing to guarantee an efficient allocation between sectors. In fact, the first period of optimizing model shows a significant reallocaiton of labor toward the nontraded sector. 3. Extraction cost of oil. The average cost of extraction of oil is assumed to increase as the ratio of output to existent reserves increases. These costs reduce the egyptian share of the avlue of oil output, since the foreign companies receive more as cost recovery. The function used to determine the share of foreign companies (Dt) is: Dt = .17232 + .69591.RO(O)XO(t)/(RO(t-1) + .001) where RO(t), XO(t) are reserves and output of oil in period t. Initial reserves are introduced to calibrate the initial share to the base data (.2154). - 158 - Appendix C: Absorptive Capacity Constraints B.1 Define effective capital accumulation as Y and resources devoted to capital accumulation as I. Then 4 )-it = y<~it i-lYi Yt< It Absorptive capacity constraints are assumed to affect aggregate investment and to apply to large increases with respect to previous periods. We assume that investment below a certain cut-off limit Y goes through without efficiency losses: yt =It if It < Y Y is a mark up on last periods effective capital accumulation. In that way the ACC will not be binding in steady state at replacement (in efficiency units) investment rates: 4 y ( X y )* 1.2 i-l it-i i.e., an annual increase of 10% gives no problems. The marginal efficiency of the excess of Y over Y* declines exponentially: I > Y => dY = exp[- a(I - Y )]dI - 159 - Simple integration then leads to the following formula: It if It < Yt Yt Y + a (1 - exp(- a(I-Y*))) if I > Y a is set equal to 2.3. This value implies that the marginal efficiency declines to 10% when I = 2 Y A natural measure of efficiency losses equals (I - Y) - 160 - Appendix D: The Steady State Capital Stocks The structure of the model is such that if all variables are defined in terms of efficiency units of labor (remember technical change is labor augumenting), it has a steady state solution - provided the real oil price is constant and the exogenous inflows of foreign exchange are also constant in terms of efficiency labor. With an infinite horizon, the economy would tend to that steady state, and even with a finite, but large, horizon most of the growth path will be "close" to it (for rigorous discussion of this and related propositions see any text on growth theory, e.g. (Dixit) 1976)). The equation defining the steady state capital stocks Ki in efficiency units equals 3 I = [GL * GLT - DEP] * I Ki (1) i-l (1) incorporates our assumption of equal depreciation rates across sectors. DEP is one minus the depreciation rate, GL and GLT are one plus the rate of labour growth and technological progress respectively. The rate of return in each sector is defined as R* 3 * RVi aFi i i i * * ( where Vi equals per unit value added in sector i and ) Si Pi represents the reproduction costs of capital, all calculated at steady state prices. - 161 - In steady sta,te the following holds for each ri : ri = (DELTA x(GLT) n_ -DEP) (3) where DELTA is 1 plus the rate of time preference and 1-n equals the elasticity of marginal utility of consumption with respect to consumption.**** World Bank Aggregate Demand and Capital Market Imperfec- Macroeconomic Imbalances tions and Economic Publications in Thailand: Simulations Development of Rela ted |with the SIAM 1 Model Vinayak V. Bhatt and os ^^e.a Waflk Grais Alan R. Roe .Interest Focuses on the demand-side adjust- World Bank Staff Working Paper ments of the Thai economy to lower Mo. 338. July 1979. 87 pages agricultural growth and to higher (including footnotes). energy prices. Discusses policy. measures and structural changes that Stock No. WP-0338. $3.00. might enable the economy to over- come these problems and continue to maintain high GDP rates of growth. The Changing Niature of World Bank Staff Working Paper No. Export Finance and Its 448. April 1981. 70 pages (including Implications for Developing 2 appendixes). Countries Stock No. WP-0448. $3.00. Albert C. Cizauskas World Bank Staff Working Paper No. 409. July 1980. 43 pages (including An Analysis of 3 annexes). Developing Country Stock No. WP-0409. S3.00. Adjustment Experience and Adjustment Experiences in Growth Prospects of the the 1970s: Low-income Asia Semi-industrial Countries Christine Wallich Compounding and Dis- Frederick Jaspersen This background study for World counting Tables for This background study for World Development Report 1981 examines Project Evaluation Development Report 1981 examines low-income South Asias adjustment J. Price Gittinger, editor the successful process of adjustment to the external shocks of the 1970s, Easily comprehensible, convenient to extemal "shocks" of the 1970s especially those factors that helped tables for project preparation and (rising prices of oil imports, reduced make the effects of these external analysis demand for exports, slower economic developments less severe in the a growth In the OECD countries) in the region than in other parts of the The Johns Hopkins University Press, semi-industrial developing countries. developing world. 1973; 7th printing, 1982.143 pages. Presents an analytical framework for World Bank Staff Working Paper No. LC 75-186503. ISBN 0-8018-1604-1, quantifying the effects of demand 487. August1981. iv + 39 pages $6.00 paperback. management and structural adjust- (including references). ment in forty-two countries, with par- Arabic: World Bank, 1973. (Available ticular reference to Uruguay, Brazil, Stock No. WP-0487. $3.00. from ILS, 1 715 Connecticut Avenue, Republic of Korea, and Turkey. N.W., Washington, D.C. 20009, U.SA.) World Bank Staff Working Paper No. Aspects of Development $4.00 paperback. 477. August 1981. 132 pages (Including Bank Management French: Tables d'int&ets composes et 3 appendixes). William Diamond and d'actualisation. Economica, 4th print- Stock No. WP-04 77. $5.00. V. S. Raghavan ing, 1979. Deals exclusively with the manage. ISBN 2-7178-0205-3, 36 francs. A41ustment in ment of development banks. The Spanish: Tablas de interes compuesto y Low-Incomne Africa book Is divided into eight sections, de descuento para evaluaci6n de proyec- Robert Liebenthal each dealing with one aspect ofto.diraleco,17;4hpn- ThIsobackground siebentudy forormanagement of Its problems, and tnosg Edtorial 9 ecnos0.973. de. print- This bckgrond stuy forWorldof the various ways of dealing ig 90 Development Report 1981 analyzes with them. ISBI 84-309-0716-5, 380 pesetas. the adjustment to external shocks during the 1970s made by a group EDI Series in Economic Development. of middle-income and low-income The Johns llopkins University Press, A Conceptual Approach to African countries, with particular 1982. 311 pages. the Analsi of External Debt reference to Kenya. Tanzania, of the Developig Countries SenegaL, and Sudan. LC 81 -481 74. ISBM 0-8018-2571-7,ofteDvlpn Cutrs $29.95 hardcover; ISBD 0-8018-2572-5, Robert Z. Aliber 486W August 1981. ri + 56 pages $12.95 paperback. World bank Staff Working Paper no. (Including bibliography). 421. October1980. 25 pages (including Stock lo. WP-0486. $3.00. apnx WP-0421. Stock Mo. WP-0421. $3.00. Development Banks Developments in and Food Policy Issues in William Diamond Prospects for the External Low-income Countries Operating experiences that serve as a Debt of the Developing Edward Clay and others practical guide for developing coun- Countries: 1970-80 A background study for World tries, with a selected list and and Beyond Development Report 1981. Discusses summary description of some sicholas C. nope food distribution-especially its development banks. insecurity in the face of extemal This background study for World economic pressures and potential The Johns Hopkins University Press, Development Report 1981 analyzes conflicts with intemal production 1957; 5th printing, 1969.141 pages the debt situation and its implica- concems-in general and with (Including 2 appendixes, index). tions for future borrowing. reference to Bangladesh, Zambia, LC 57-13429. ISBN 0-8018-0708-5, World Bank Staff Working Paper No. and India. $5.00 (.3.50) paperback. 488. August 1981. 70 pages (including World Bank Staff Working Paper No. 2 annexes, references). 473. August 1981. vii + 115 pages. Development Finance Com- Stock No. WP-0488. $3.00. Stock No. WP-04 73. $5.00. panies: Aspects of Policy and Operation Energy Prices, Substitution, A General Equilibrium William Diamond, editor; and Optimal Borrowing in Analysis of Foreign essays by E. T. Kuiper, the Short Run: An Analpsis Exchange Shortages in a Douglas Gustafson, and of Adjustment in Oil- Developing Economy P M. Mathew Importing Developing Kemal Dervis, Jaime de Melo, The Johns Hopkins University Press, Countries and Sherman Robinson 1968.130 pages (including appendix, Ricardo Martin and Examines the consequences of alter- index). Marcelo Selowsky native adjustment mechanisms to LC-68-27738. ISBN 0-8018-0166-4, Develops a short-term model for foreign exchange shortages in semi- $5.00 (93.25) paperback. evaluating the adjustment (par- industrial economies. Compares french: Les societes financieres de ticularly, external borrowing) of oil- devaluation to two forms of import developpement: quelques aspects de importing developing countries to the rationing and flnds that adJusting by leur politique et de leurs activites. increases in oil prices during the. rationing is much more costly In leurapoi7able fetde loteuWorld a Banks 1970s. Discusses the borrowing terms of lost gross domestic product. (Available free from the World Bank. strategies that can be expected in the World Bank Staff Working Paper No. Washington, D. C) future and the demands that will be 44rl. January 1981 ing Pages Ncu Spanish: Las compafifas financieras made on multilateral institutions. 443. January 1981. 32 pages (includ- de desarrollo: algunos aspectos de su ing references). politica y de sus actividades. Editorial World Bank Staff Working Paper No. Stock No. WP-0443. $3.00. Tecnos, 1969. 466. July 1981. 77 pages (including 300 pesetas. footnotes, references). Stock No. WP-0466. $3.00. NEW Development PopcsGrowth and Structural of Capital Surplus Olp Exchange Rate Adjustment AGjustment In East Asia Exporting Countries: Iraq, under Generalized Currency Farvez Hasan Kuwait, Libya, Qatar, Floating: Comparative Saudi Arabia, UAE Analysis among Developing Analyzes the economic performance Rudolf Hablultzel Countries of the flive large market economies of Rudolf liabiutZel ~~~~~~~~~~~~East Asia,-Korea, Thailand. the This background study for World Romeo M. Bautista Philippines, Malaysia, and Development Report 1981 Examines the experiences of twenty- Indonesia-during the last two discusses the production strategies two developing countries In adapting decades; focuses on the key factors and the development policies of the to the generalized floating of the explaining their remarkable economic capital-surplus oil-exporting world's major currencies since 1973 and social progress; and identifles the countries. and discusses the implications that main economic issues for the 1980s. Worl Ban Staf WokingPape No. currency floating has on policymaking World Bank Staff Working Paper No. World Bank Staff Working Paper Nlo. in these countries and indicates direc- 529.1982. 42 pages. 483. August 1981. 53 pages (Including tions for further research. statistical tables). World Bank Staff Working Paper No. ISBN 0-8213-0102-0. $3.00. Stock No. WP-0483. $3.00. 436. October 1980. 99 pages (including appendix). Stock No. WP-0436. $3.00. International Adjustment In Motes on the Mechanics of ect appraisal. To give substance to the 1980s Growth and Debt the applied and policy dimensions, many of the readings are drawn from Vijay Joshi Benjamin B. King the experience of development prac- A backgound study for World A practical model to explore the way titioners and relate to such important Development Report 1981. Analyzes in which capital inflow from abroad sectors as agriculture, industry, the macroeconomics of international affects economic growth. power, urban services, foreign trade, adjustment. Highlights potential and employment. The principles market failures and areas for The Johns Hopkins UniuCFsi4 Press outlined are therefore relevant to a intervention. 1968. 69 pages (including 4 annexes). host of development problems. World Bank Staff Working Paper No. LC 68-8701. ISBN 0-8018-0338-I, $5.00 The Johns 11opkins University jPess. 485. August 1982. 57 pages. (£3.00) paperback. February 1983. About 304 pages. ISBN 0-8213-0062-8. $3.00. LC 82-7716. ISBN 0-8018-2803-1, The Policy Experience of $35.00 hardcover; ISBN 0-8018-2804-X, Twelve Less Developed $12.95 paperback. NEW Countries, 1973-1978 Bela Balassa Private Bank Lending to The iNature of Credit Uses the methodology applied In the Developing Countries Markets In Developing author's "The Newly Industrializing Richard O'Brien Countries: A Framework Developing Countries after the Oil for Policy Analysis Crisis" (World Bank Staff Working A background study for World Arvind Virnani Paper No. 473, October 1980) to Development Report 1981. examine the policy experience of Describes the evolution of relation- The central purpose of the paper is to twelve less developed countries In ships between private banks and analyze various forms of government the period following the quadrupling developing countries. Intervention in the loan market in of oil prices in 1973-74 and the World Bank Staff Working Paper No. terms of their effect on efficiency. world recession of 1974-75. 482. August1981. ul + 54 pages World Bank Staff Working Paper No. World Bank Staff Working Paper No. (including appendix, bibliography). 524.1982. 204 pages. 449. April 1981. 36 pages (including Stock No. WP-0482. $3.00. ISBN 0-8213-0019-9. $5.00. appendix). . Stock No. WP-0449. $3.00. Private Capital Flows to The Newly Industrializing Developing Countries and Developing Countries after The Political Structure of Their Determinatiohs: the Oil Crisis the Niew Protectionism Historical Perspective. Bela Balassa Douglas R. Nelson Recent Experience, and World Bank Staff Working Paper No. This background study for World Future Prospects 437. Octoberl980. 57pages (including Development Report 1981 Alex Fleming OctXo1 5 e lpresents a political-economic A b leging appendix). analysis of what has been called the A backgound study for World Stock No. WP-043 7. $3.00. "new protectionism:. Development Report 1981. Discusses the nature and determina- World Bank Staff Working Paper No. tion of recent private capital flows to Notes on the Analysis of 471. July 1981. 57 pages (including developing countries. Focuses on Capital Flows to Developing references). those flows passing through the Nations and the Stock No. WP-0471. $3.00. prospects for and d exatrines o he "Recycling" Problem developing countries' continuing Ralph C. Bryant access to the intemational capital A backgound study for World MEW markets. Development Report 1981 World Bank Staff Working Paper No. Summarizes and cridcizes the con- Pricing Policy for Develop- 484. August1981. 41 pages. venUtonal analysis of the interrela- ment Management tions between flnancial markets In Gerald M. Meier Stock No. WP-0484. $3.00. the Industrialized countries and capi- tal flows to the developing nations. Presupposing no formal training In economics, it explains the essential Private Direct Foreign World Bank Staff Working Paper No. elements of a price system, the Investment in Developing 476. August 1981. 67 pages. functions of prices, the various Countries Stock No. WP-04 76. $3.00. policies that a goverment might Cu Billerbeck and Y. Yasugi pursue In cases of market failure, .Blebc-n .Ysg and the principles of public pricing of World Bank Staff Working Paper No. goods and services provided by 348. July 1979. iu + 97 pages (Includ- government enterprises. It also pro- I 2 annexes). vides the would-be practitoner with an appreciaUon of the underlying Stock No. WP-0348. $5.00. logical structure of cost-beneflt proj- NEW Structural Adjustment Implications in terms of Income -PoUcies in Developing generation, external deficit and Short-Run Macro-Economic Economies inflation. Adjustment Policies In Bela Balassa World Bank Staff Working Paper tNo. South Korea: A Quantitative Examines structural adjustment 513. June 1982. 93 pages (including Analsis policies (policy responses to external appendix). Sweder van Wijnbergen shocks, such as the quadrupling of ISBN 0-8213-0023-7. $3.00. oil prices and the world recession of An analysis of the startling reversal of the 1970s) of developing countries. performance of the South ICorean Considers reforms in production World Debt Tables economy In 1979 and 1980 compared Incentives, Incentives to save aind to Acopltnofdtonhexera with the preceding flfteen years, and inetpbi ntvtmotaen o A compilation of data on the external with he pecedig fiteen ears and invest, public investments, sectoral public and publicly-guaranteed debt an explor-ation of the short-run policies, and monetary policies, and o 0 eeoigcutispu miacro-economic policy options comments on the Interdependence of eight evenladitiounaltables ofprivat available to Korea in 1981. Hlighlights eighteenlcymasrs n o f 0 adediptjionngal tables of private avaiableto Krea n 191. iighlghtS the various policy measures and on and nonguaranteed debt from the the role of commercial banks, foreign the international environment in Wrd n an tor Reportin capital Inflows, and money markets which they operate. World Bank Debtor ntpore ng and the use of credit obtained from System. Describes the nature con- these sources to flnance fixed and World Bank Staff Working Paper Nlo. tent, and coverage of the data; working capital. 464. July 1981. 36 pages. reviews the external debt of 101 countries through 198L contains World Bank Staff Working Paper No. Stock No. WP-0464. $3.00. tables on external public debt out- 510. November 1981. lv + 1 78 pages standing commitments, disburse- (Including 3 appendixes). ments, service payments, and net ISBN 0-8213-0000-8. $5.00. NEW borrowings of 101 developing coun- tries, by country, 1972-1981. Structural Aspects of (EC-167/81). December 1982. Annual. NEW Turkish Inflation: About 300 pages. 1950-1L979 ISSN 0253-2859. $75.00. State Finances in India M. Ataman Aksoy Computer tapes containing the data A three-volume set of papers that Inflation has been one of the major bases for the World Debt Tables are explores a range of issues relating to problems of the Turkish economy available from the Publications the nature of intergovemmental fiscal during the postwar period. This paper Distribution Unit World Bank. The relations in India. develops alternative inflation models tapes are available to international and analyzes their performance in agencies and official nonproflt agen. Vol : Sha light of the Turkish experience in cies of member governments at a VOL:I.- Ren Shaing order to provide a framework on nominal fee. For information concern- In India which a more realistic macro model ing fees for other organizations, Christine Wallich can be developed. please write to the addressee listed Deals speciflcally with the principles World Bank Staff Working Paper No. above. of revenue sharing in India. 540.1982. 118 pages. Supplements to World Debt Tables are ISBN 0-8213-0098-9. $5.00. issued periodically as information VoL 11: India-Studies in becomes available; the current State ,ltances updates are included with orders for Christine Wallich NEW World Debt Tables. Examines In detail the Implications of revenue sharing for project flnance. Thailand: An Analysis of Structural and Mon- The Impact of Contractual Savings on Resource Mobilization and Alloation: VoL 11k: The Measurement of Structural Adjustments The Experience of Malaysla Tax Effort of State Govern- Ame Drud, Waflk Grais, and Socil sccuflty Funds in Slngapore mcnts, 1973-1976 Dusan Vujovic and the Phippines: Ramifications of Raja J. Chelliah and This study was prepared as a Inetent PoUcdes Narain Sinha background paper for the preparation ivestments of Social Security Funds of a tructral-ajustmnt lon to In india and Sri Lanka: Loegislation Attempts to evaluate the tax perfor- of a structural-adJustment loan to and Experince mance of particular states in terms of Thailand and is a follow-up to a pre- Parthasarathi Shome and the average tax effort of all states. vious paper entitled 'Aggregate Katrine Anderson Saito Demand and Macroeconomic World Bank Reprint Series: Nlumber 144. World Bank Working Paper No. 523. Imbalances in Thailand:' Comparative Reprinted fi,m The Malayan Economic Review. uol. September1982. vol. 1, 85 pages, vol. II, statistics are used, within the frame- 23. no. 1 (April 19781.54-72: Labour and Society, 186 pages, vol. III, 85 pages. work of a four-sector macroeconomic wL. 5. no. I (January 1980J.19-30: and The Indlian ISBN 081003Xuo 00wl model to assess alternative ways of Journal of Economics, vol. 60, part 3, no. 238 ISBIY 0-8213-0013-X vol. 1, $5.00, vol. II f macroeconomic adjustment in the (January 1980).349-60. $5.00, vol. III, $3.00. Thai economy. Discusses speciflcally Stock nO. RP-0144. tree of charge. flscal policy interventions, manipula- tions of the exchange rate, and pro- ductivity Improvements and their Policy Responses to External Shocks in Selected Latin American Countries Bela Balassa World Bank Reprint Series: Number 221. Reprintedfrom Quarterly Review of Economics and Business. vol. 21, no. 2 (Summer 1981).131-64. Stock 11o. RP-0221. free of charge. Restructuring the World Econonsy: Round 11 ilollis Chenery World Bank Reprint Series: Number 204. Reprinted from Foreign Affairs (Summer 1981).:102-7120. Stock No. RP-0204. free of charge. Risk Assessments and Risk Premiums in the Eurodollar Market Gershon Feder and Knud Ross World Bank Reprint Series: Number 220. Reprinted from The Journal of Finance, uol'. 37. no. 3 (June 1982):679-91. Stock No. RP-0220. Free of charge. Italian: "Analisi economica dei progetti The Effects of Population di inuestimento"in Analisi dei progetti Growth, of the Pattern of di investimento: il metodo della Banca Demand, and of Technology Mondiale. Marsilio Editori, s.p.a., on the Process of Urbaniza- S. Croce 518/A, 30125 Venice, tion: An Application to India Italy. 1978. Rakesh Mohan ISBN 10-4108-8, L8,800. This paper uses a non-linear, three- Portuguese: Analise econ6mica de pro- sector, two-region wage-and-price jetos. LTC-Livros Tecnicos e Cientlficos, endogenous dynamic general S. A., Au. Venezuela, 163, 20.220-Rio de equilibrium model to study the effect Approaches to Purchasing Janeiro, R J, Brazil. 1979. of population growth, the pattem of demand, and of technological change Power Parity and Real ISBIY 85-216-0017-8, $8.00 equivalent, on urbanization in the context of a Product Comparisons paperback. low-income developing country start- Using Shortcuts and Spanish: Analisis econ6mico de proyec- ing at a low level of urbanization. Reduced Information tos. Editorial Tecnos, 1977. World Bank Staff Working Paper Sultan Ahmad ISBN 84-309-0719-X, 435 pesetas. No. 520.1982. 47 pages. World Bank Staff Working Paper No. ISBN 0-8213-0008-3. $3.00. 418. September 1980. ii + 60 pages (including 14 tables, bibliography). Economic and Social Stock No. WP-0418. S3.00. Analysis of Projects and of N_L_ W Price Policy: The Morocco Comparative Study of the Credit Proiect u India, 1975-1985: A Management and Organiza- Kevin M. Cleaver Sources and Uses of Funds tion of Irrigation Projects Approach Anthony E Bottrall World Bank Staff Working Paper No. Armando Pinell-Siles and World Bank Staff Working Paper No. 369. January 1980. 59 pages (includ- V.J. Ravishankar 458. May 1981. 274 pages (includi i n g This paper presents sources and uses 3 appendixes). Stock No. WP-0369. $3.00. of funds accounts integrated in a Stock No. WP-0458. $10.00. macroeconomic framework for Economy-Wide Models and analyzing financing patterns and Development Planning economic interdependence among, DevrelopmBitent Plnigsectors in India, contrasting the Economic Analysis Charles R Blitzer, period 1975-80 with the Sixth Plain of Projects Peter B. Clark and period. 1980-85. Lyn Squire and Lance Taylor, editors World Bank Staff Working Paper Mo. Herman G. van der Tak Surveys the specification and uses of 543.1982. 92 pages. Reconsiders project appraisal and medium-term and perspective ISBN 0-8213-0205-5. S3.00. recommends a more systematic and economywide planning models. consistent estimation and application Oxford University Press, 1975; 4th General Equilibrium Models of shadow prices and a calculation of pr,inting, 1982. 382 pages (including for Development Policy rates of return that take explicit selected additional readings, bibliogra- Kemal Dervis, Jaime de Melo, account of the project's impact on the phy, subject and author index). and Sherran Robinson distribution of income. adSemnRbno The Johns Hob pkins university LC 74-29171. ISBN 0-19-920074-2, Provides a comprehensive study of 1975 4thn Hprinti ng ve1981. 14Pagess $9.95 paperback. multisector, economywide planning 19 75; 4th printing, 1981. 164 pages pmodels with particular emphasis on (including appendix, glossary, issues of trade, distribution, growth, bibliography). and structural change. Theoretical LC 75-40228. ISBN 0-8018-1818-4, discussion of the properties of $6.50 ({2.75) paperback. multisector, applied general equilibrium models is combined with French: Analyse economique numerical applications to particular des projets. Economica, 1977; 2nd countries and problems. The models printing, 1981. considered range from input-output ISBIY 2-7178-0014-X, 29 francs. and linear programming to the more recent nonlinear computable general equilibrium (CGE) models. The authors consider how these models can be used to analyze questions of liuman Resource Develop- Phase : A System of Inter- growth and structural change, the ment and Economic Growth national Comparisons of selection of foreign exchange regime, In Developing Countries: A Gross Product and and the impact of alternative develop- ~ ment strategies on the distribution of Simultaneous Model Purchasing Power income. The empirical applications David Wheeler Irving B. Kravis, are based both on cross-country World Bank Staff Working Paper No. Zoltan Kenessey, Alan Hieston, analysis and on the experience of par- 407. July 1980.130 pages (including and Robert Summers ticular countries and demonstrate how such models provide a useful 8 appendixes, bibliography). * Establishes the methodology and pre- framework for policy analysis. Parthcu- Stock No. WP-0407. $5.00. sents comparisons of gross domestic lar attention is focused on the product per capita and currency problems of planning and policy for- purchasing power for ten countries in mulation in mixed-market economies 1970 and six of the same countries and on the nature of models required few in 1967. to capture the important mechanisrns' The Johns Hopkins University Press, that constrain policy markets. Incorporating Uncertainty 1975. 306 pages (including Cambridge University Press, 32 East into Planning of Industrial- glossary, index). 57th Street, [lew York, N.Y. 10022. ization Strategies for LC 7J-19352. ISBNI 0-8018-1606-8, 1982. xviii + 526 pages. Developing Countries $27.50 (f16.50) hardcover; LC 81-12307. ISBtl 0-521-24490-0, Alexander i. Sarris and ISBIl 0-8018-1669 -6, $8.95 (25.50) $42.50 hardcover, ISBN 0-521-27030-8, Irma Adelman paperback. $17.95 paperback. This survey of existing literature on planning under uncertainty focuses on issues of intemational trade and Phase 11: International Com- NYEW investment allocation. Various ways of parlsons of Real Product incorporating uncertainty into target- and Purchasing Power The Global Framework: planning models are discussed and living B. Kravis, Alan n eston, The GlobalFramework:proposals for possible empirical and Robert Summers An Update applications are outlined. Brian Nolan WrdUpdates Phase I and adds six new World Bank Staff Working Paper No. countries, comparing the flgures for This paper presents the methodologi- 503. January 1982. 58 pages (including the sixteen countries for the years cal background to the Global Model- appendix, references). 1970 and 1973. ling Framework used in the World Development Report exercises of the Stock No. WP-0503. $3.00. The Johns Hopkins University Press, World Bank. It gives an overall view of 1978. 274 pages (including this framework and discusses in l inde) some detail the data base supporting Interdependence in glossary, index). the global analysis and the methods Planning: Multilevel LC 77-17251. ISBI 0-8018-2019-7, of reconciling data from various Programming Studies of the $25.00 (f15.00) hardcover; ISBN sources. Ivory Coast 0-8018-2020-0. $8.50 (£5.00) World Bank Staff Working Paper Louis M. Goreux paperback. No. 533. 1982. 58 pages. Provides a system for analyzing each ISBN 0-8213-0047-4. $3.00. component of a country's economy Phase 111: World Product and independently and relates the interde- Income: International Com- pendencies between the components. parsons of Real GDP Human Factors The Johns Hopkins University Press, Irving B. Kravis, Alan Heston, Pojec rork 1977. 448 pages (including bibliogra- and Robert Summers and Francis J. Lethemphy, index). This report restates and extends the LC 77-4793. ISBN 0-8018-2001-4, methodology set out in the first two World Bank Staff Working Paper No. $27.00 (f16.20) hardcouer; volumes. Particular attention is given 397. June 1980. 85 pages (including ISBN 0-8018-2006-5, $9.95 to the problem of comparing services 3 annexes, 5 charts, bibliography). (26.00) paperback. and to the conflicting demands of regional and global estimates. Com- Stock No. WP-0397. $3.00. parisons are given of prices, real per The International capita quantities, and final expendi- Comparison Project ture components of GDP for thirty- Chreevompaiso that establish a four countries for 1975. By relating Three volumes that establish a worldwide system of intemational comparisons of real product and of the purchasing power of currencies. the results to certain widely available A Model of an Agricultural used to estimate the project's effects national income accounting data and lHousehold: Theory and on key national variables, thus per- related variables, the authors develop Evidence mitting a full social cost-benefit extrapolating equations to estimate analysis of the project. per capita GDP' for the thirty-four Howard N. Barnum ahealysis Hofkthe projest. jP-s countries for 1950 to 1978. In addi- and LynThe Johns Hopkins Uniersity Press. tion, the 1975 distribution of world Innovative model of short-run 1982. 336 pages (including maps and product by region and per capita behavior that combines production index). income class is estimated. The and consumption decisions in a LC 81-48173. ISBN 0-8018-.2802-3, 1975 results confirm relations theoretically consistent fashion for an $30.00 hardcouer. between both quantities and prices agricultural household. and per capita income found in the earlier volumes. The Johns Hopkins Uniuersity Press, Redistribution with Growth The Johns Hopkins Uniuersity Prss 1980. xi + 107 pages (including Holins Chenery, 1982. 398 pages. appendix, references). Montek S. Aheuwaiar LC 81-15569. ISBN 0-8018-2359-5, LC 78-21397. ISBN 0-8018-2225-4, Clive Bell John H. Duloy, $35.00 hardcouer: ISBN 0-8018-2360-9, $6.95 (i.475) paperback. and Richard Jolly $15.00 paperback. Describes existing inequality in The Political Economy of incomes in developing countries and Landsat index Atlas of the Specialized Farm Credit proposes a reorientation of develop- Institutions in Low-income meit policy to achieve more equitable Developing Countries distribution. of the World Countries Fourteen four-color maps, applica- J. D. Von Pischke, Oxford University Pess, 1974; 4th tions of Landsat imagery, reading and Peter J. Heffernan, and printing, 1981. 324 pages (including uses of the index maps, and pro- Dale W. Adams annex, biblography). cedure for securing imagery. World Bank Staff Working Paper No. ISBN 0-19-920070-X, $9.95 (i5.00) ihe Johns Hopkins University Press, 446. April 1981. iii + 99 pages. paperback. 1976.17 pages,131/2" x 21", spiral Stock No. WP-0446. $5.00. French: Redistribution et croissance. bound. Presses Universitaires de France, 108, LC 76-46190. ISBN 0-8018-1923-7, boulevard Saint-Germain, 75006 Paris, $12.00 (S7.25) paperback. NEW France, 1977. ISBN 22403102, 58.20 francs. Methodologies for Measur- Project Evaluation In Spanish: Redistribuci6n con creci- ing Agricultural Price Regional Perspective: A miento. Editorial Tecnos, 1976. Intervention Effects Study of an Irrigation ISBN 84-309-0624-X, 880 pesetas. Pasquale L. Scandizzo Project In Northwest and Colin Bruce Malaysia Clive Bell, Peter Hazell, and World Bank Staff Working Paper No. Roger Slade 394. June 1980. x + 96 pages This innovative study develops quan- (including 4 appendixes, references). titative methods for measuring the Stock No. WP-0394. $5.00. direct and indirect effects of agricultural projects on their sur- rounding regional and national economies. These methods are then applied to a study of the Muda irriga- tion project in northwest Malaysia. A linear programming model is used to analyze how a project changes the farm economy, and a social account- ing matrix of the regional economy is then estimated. This provides the basis for a semi-input-output model, which is used to estimate the indirect effects of the project on its region. Thereafter, a similar methodology is Risk Analyis in Techniques for Project WRINTS Project Appraisal Appraisal under Uncertainty The Meaning of Technologlcal Mastezr Louis Y. Pouliquen Shlomo Reutlinger h RKeatlion to Transfer of Technokogy Discusses methodological problemis Presents a method of evaluating risk World Bank Reprint Saries: Lrumyer 217. and the usefulness of simulation; in Investment projects and means for Reprinted from Annals of the American Acadmy Illustrated by three case studies. using quantitative measures of risk in of Political and Social Science, ol. 458 (tNouem- decisionmaking. ber 1981).12-26. The Johns HYopkins University Press, 1970; 4th printing, 1979. 90 pages. The Johns Hopkins University Press, Stock No. RP-0217. free of charge. LC 79-12739. ISBN 0-8018-1155-4, 19 70; 4th printing, 1979. 108 pages $5.50 (~.025) paperback. ' (including annex, bibliography). The model of an Agricultural LC 74 -9482 7. ISBN 0-8018-1154-6, Household in a Multi-Crop Econofmy: French: Lappreciation du risque dansp The Case of Korea I'evaluation des projets. Dunod Editeur, $5 95 (0-50) papeback. Choong Yong Ahn, Inderjit Singh. and 24-26, bouleuard de l'tl6pital, 75005 Lyn Squire Paris, France, 1972. World Bank Reprint Series: lumber 222. What Is a SAM? A Layman's Reprinted from The Review of Economics and 21 francs. Guide to Social Accounting Statistics. vol. 63. no. 4 (fNovember 191):20-25. Matrices Stock NIo. RP-0222. free of charge. Shadow Pces for Project Benjamin B. King The Reklvance of the Dual Economy Appraisal In Turkey World Bank Staff Working Paper No. Model: A Case Study of Thailand Afsaneh Mashayekhi 463. June 1981. 59 pages (including Trent Bertrand and Lyn Squire World Bank Staff Working Paper No. references). World Bank Reprint Series: Nlumber 219. 392. May 1980. 57 pages. Stock No. WP-0463. $3.00. (1980):480-rr. c Stock No. WP-0392. $3.00. Stock No. RP-0219. free of charge. Social Cost-Beneflt Analysis: A Guide for Country and Project Economists to the Derivation and Application of Economic and Social Accounting Prices Colin Bruce World Bank Staff Working Paper No. 239. August 1976. ii, iII + 143 pages (including 6 annexes). Stock No. WP-0239. $5.00. I WORLD BANK PUBLICATIONS ORDER FORM SEND TO: WORLD BANK PUBLICATIONS WORLD BANK PUBUCATIONS PCX BOX 37525 or 66, AVENUE D'IENA WASHINGTON, D.C. 20013 75116 PARIS, FRANCE U.S.A. Name: Address: Stock or ISBN # Author, Title Qty. Prce Total Sub-Total Cost: Postage & handling fee for more than two free items ($1.00 each): Total copies: _ Air mail surcharge ($2.00 each): TOTAL PAYMENT ENCLOSED: Make checks payable: WORLD BANK PUBLICATIONS Prepayment on orders from individuals is requestetd. Purchase orders are accepted from booksellers, library suppliers, libraries, and institutions. All prices include cost of postage by the least expensive means. The prices and publication dates quoted in this Catalog are subject to change without notice. No refunds will be given for items that cannot be filled. Credit will be applied towards future orders. No more than two free publiatons will be provided without charge. Requests for additional copies will be filled at a charge of US $1.00 per copy to cover handling and postage costs. Airmail delivery will require a prepayment of IJS $2.00 per copy. Mail-order payment to the World Bank need niot be in U.S. dollars, but the amount remitted must be at the rate of exchange on the day the order is placed. The World Bank will also accept IJnesco coupons. HG 3881.5 .W57 W67 NO.622 C.3 DERVIS, KEMAL. POLICY ANALYSIS OF SHADOW PRICING, FOREIGN BORROWING, The World Bank Headquarters European Office Tokyo Office 1818 H Street, N.W. 66, avenue d'Iena Kokusai Building Washington, D.C. 20433, U.S.A 75116 Paris, France 1-1 Marunouchi 3-chome Telephone: (202) 477-1234 Telephone: (1) 723-54.21 Chiyoda-ku, Tokyo 100, Japan Telex: WUI 64145 WORLDBANK Telex: 842-620628 Telephone: (03) 214-5001 RCA 248423 WORLDBK Telex: 781-26838 Cable Address: INTBAFRAD WASHINGTONDC ISSN 0253-2115/ISBN 0-8213-0277-9