Report No. 2204-PE Peru: Long-Term Development Issues Volume 11 The Main ReDort April 13, 1979 Country Programs Department I Latin America and Caribbean Regional Office FOR OFFICIAL USE ONLY Document of the V4brd Bank This document has a restricted distribution and rnay be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank- authorization. CUpRENICY EQUIVALENTS The exchange rate is being adjusted daily roughly in line with the difference between domestic and international inflation. The exchange rate and currency equivalents in effect on March 29, 1979, were as follows: Currency Unit - Sol (S/.) Calendar 1978 March 29, 1979 US$1 S/. 160 S/. 210 SI. 1 - US$0.0062 US$0.0047 - 35 - FOR OFFICIAL USE ONLY TABLE OF CONTENTS VOLUME II: THE AAIN REPORT Page No. INTRODUCTION ........................................... 40 PART I: THE SEITING: NATURAL RESOURCES, POPULATION, ECONOMIC HISTORY AND GOVERNMEaNT STRATEGY .... ...... 43 CHAPTER I: GEOGRAPHY, NATURAL RESOURCES AND DEVELOPMENT ......... 43 A. GEOGRAPHY AND CLIMATE ....... ....................... 43 B. NATURAL R£ESOURCES ... .............................. 44 a. Mining ....................................... 44 b. Fisheries ...... .............................. 45 c. Agricultural Land .............................. 46 d. Water ........................................ 48 e. Eaergy ............*a ....................................... ... # 51 C. CONCLUSIONS ... .................................... 58 CHAPTER II: POPULATION: GROWTH AND SPATIAL DISTRIBUTION 59 A. DEMOGRAPHIC TRENDS; Characteristics and Projection 59 a. Brief Historic Overview, 1530-1940 . .59 b. Population Growth, 1940-72 . .60 i. Overview .60 ii. Components of Growth: Birth and Fertility Rates .60 iii. Components of Growth: Mortality Rates 65 c. Magnitudes and Consequences of Future Population Growth . .66 B. LABOR FORCE TRENDS AND PROJECTIONS ............. ... 72 a. Past Trends and Characteristics .............. 72 b. Determinants of Future Growth of the Labor Force and Alternative Projections, 1972-2000 77 C. SPATIAL DISTRIBUTION AND MIGRATION .... ............ 81 a. General Characteristics of Urbanization and Migration .... ......... .. .. 81 b. Regional Disparities and Causes of Urbani- zation and Migration ..... .................. 93 c. Pattern of Urbanization and Spatial Distribution ... . . ....... 114 D. CONCLUSIONS ....................................... 123 CHAPTER III: AN OVERVIEW OF ECONOMIC HISTORY 126 A. THE INCA EMPIRE ................................... 126 B. IRE CULONIAL PERIOD ............................... 127 C. THE GUANO ERA ..................................... 128 0. IHH TWENTIETH CENTURY ............................. 133 Thiostumen hu a restricted distribution and ray be ud b redpients only in the peoforance Of thew oRii duti.tconten may nOt otbenrue discld wthoug Word Bank autho=n. - 36 - Page No. CHAPTER IV: RECENT GOVERNMENT STRATEGY .......................... 139 A. THE VELASCO GOVERNMENT ............................. 139 a. Objectives and Strategy . ............... 139 b. Achievements ................. 141 c. Problems and Shortcomings . .......... 144 B. THE MORALES GOVERNMENT ................. .......... 154 a. The Tupac Amaru Plan ............... ......... 154 b. Macro-financial Management and the 1977-79 Stablization Program ...................... 159 PART B. MACROECONOMIC.ISSUES: EFFICIENCY SAVINGS AND CYCLES .. 163 CHAPTER V: MACROECONOMIC DEVELOPMENTS AND LSSUES ...... ......... 163 A. MAIN FEATURES AND CYCLICAL BEHAVIOR OF PATTERN OF GROWTH ..................... .................... 163 a. Growth and Structure of Output . ........... 163 b. Resource Use ............... .. ............... 166 c. The Role of Exports in the Economy .......... 169 d. Cyclical Fluctuations .......... .. ........... 170 B. INVESTMENT AND SAVINGS ........................... 174 C. FINANCIAL INTERMEDIATION, INFLATION AND SAVINGS .. 182 a. The Financial Structure of the Economy ...... 182 b. Monetary Policy Investments ................. 189 c. Monetary Behavior and Inflation ............. 193 i. Honey Supply ............................ 193 ii. General Model of the Demand for Real Monetary Balances and the Effect on Inflation ...... 198 iii. Requirements for Balanced Financial Growth ........ 199 -Implications for Money Supply Policy ... 199 -The Role of a Realistic Interest Rate Policy ................................................... 201 CHAPTER VI: THE ROLE OF THE PUBLIC SECTOR ...................... 206 A. IMPORTANCE AND INSTITUTIONAL ORGANIZATION OF THE ECONOMIC PUBLIC SECTOR ......................... 206 B. PUBLIC INVESTMENT AND EFFICIENCY OF RESOURCE USE 212. a. Magnitudes and Characteristics of Public Investment in 1971-76 ...... ............... 212 b. Achievements and Probles . ................... 217 C. Conclusions ............. .................... 226 C. FINANCING OF INVESTMENT AND PUBLIC SAVINGS ....... 226 a. Public Enterprises as Savers .... ............ 230 i. Price Policies ..... ..................... 231 ii. Financial Management ..... ............... 236 iii. The Fishing Industrv ..... ............... 237 b. The Social Security System as Saver .... ..... 238 c. Local Governments and Decentralized Agencies as Saver ....... ............................ 241 d. The Central Government as Saver .... ......... 242 - 37 - Paae No. D. MANAGEMENT CAPACITY OF THE PUBLIC SECTOR ......... 252 a. Aggregate Development Planning .............. 252 b. Controls of Economic Activity ............... 257 E. CONCLUSIONS AND FUTURE PRIORITIES . ............... 259 CHAPTER VII: THE TAX SYSTEM .................................... 260 A. AN OVERVIEW OF ISSUES IN THE TAX SYSTEM ......... . 260 B. ISSUES IN MAJOR TAXES ................................. 265 a. Taxes on Income and Profits ... .............. 265 i. The Personal Income Tax ................ 265 ii. The Corporate Income Tax ......... . ....... 273 b. Property Taxes ............... . .. 281 c. Taxes on Goods and Services ................. 285 i. The General Sales Tax ................... 285 ii. Selective Excise Taxes ... ............... 288 d. Tariff Policy and Taxes on Imports .......... 289 C. TAX ADMINISTRATION ............................... 291 D. CONCLUSIONS ...................................... 294 PART C: DEVELOPMENTS. POLICIES AND PROSPECTS IN MAIN ECONOMIC - SECTORS .... .. ...................... 296 CHAPTER VIII: AGRICULTURE AND RURAL DEVELOPMENT 296 A. BACKGROUND ........................... . . 296 a. Resource Base . .296 b. The Agrarian Structulre and Rural Poverty 297 S. RECENT TRENDS IN OUTPUT AND FOREIGN TRADE 307 C. RECENT GOVERNMENT STRATEGY AND POLICIES . .310 a. Land Reform: Achievements and Problems 312 i. Main Characteristics of the Government's Model of Agrarian Structure . .312 -The Agrarian Production Cooperative (CAP) ..314 -The Agrarian Society of Social Interest (SAIS). 316 -Cooperativization of Peasant Community . 317 -Cooperativization of Small Farmers 318 ii. Effects of the Land Reform . .320 -Effects on the Political Structure 320 -Effects on Income Distribution and Employment . .320 -Effects on Prc'duction . .325 iii. Performance and Issues in Agrarian Reform Enterprises (AREs) .326 -Initial Conditions and Location 327 -Adequacy of the Organization of the AREs 328 -Acceptance of Cooperative Organization of Production .329 - 38 - Page No. -Size and Economies of Scale ............ 329 -Conflicting Interest Between Members.... 330 -Mechanisms of Resource Allocation and Labor Discipline ...................... 332 -Management of the ARE s. ................. 333 b. Production Planning and Marketing and Price Policies .................................. 333 i. Production Planning ..................... 333 ii. Price and Marketing Policies ........... . 336 c. Credit Policies ....... so .... ............... 339 i. Sources of Credit .. ........0 ...... 339 ii. Reaching the Small Farmers .. ............ 341 iii. Interest Rate Policies ... ........ ....... 342 d. Public Investment in Agriculture ....... .. .... 344 D. FUTURE DEVELOPMENT PRIORITIES .................... 347 a. Expansion of Output ........................ 348 b. Rural Development Programs in the Sierra .... 349 E. CONCLUSIONS ........ ..................... .. *..... ........ 350 CHAPTER IX: MINING .......... . ................................... 360 A. OVERVIEW OF PAST DEVELOPMENTS AND GOVERNMENT POLICIES ................................. ........ 361 B. OPERATIONS AND INVESTMENT BY THE PUBLIC SECTOR ... 366 a. General Policy Issues ........ .............. 366 i. Planning ........................ ....... . 366 ii. Financing ............................... 367 iii. Marketing ................... .............. 367 iv. Manpower and Training ............ ....... 368 v. Environmental Considerations ............ 368 b. Operations and Investment Programs of the State-owned Mining Enterprises ........ ..... 369 i. CENTROMIN .............. .. ............... 369 ii. HIERROPERU .............................. 373 iii. MINEROPERU .............. .. .............. 373 iv. Concluding Overview ............ .......... 378 C. CONCLUSIONS ...... ................................ 379 CHAPTER X: INDUSTRY ............................................ 380 A. INTRODUCTION ................. .. .................. 380 B. OBJECTIVES AND POLICIES IN THE INDUSTRIAL SECTOR . 381 C. ACHIEVEMENTS AND PROBLEMS OF THE INDUSTRIALIZATION STRATEGY .......... ............................. 389 a. Output and Growth and Investment .... ........ 389 b. Structure of Industry ..... .................. 392 c. Foreign Trade and Foreign Dependence ........ 397 d. Industrial Concentration, Small-scale Enterprises and Employment .... ............ 399 e. Conclusions ........ ......................... 404 - 39 - Page No. D. PUBLIC SECTOR INVESTMENT ............ 405 E. INDUSTRIAL FINANCING ... ..................... ... . 410 a. Sources of Finance : the Domestic Financial System ................ 410 b. Sources of Finance: The Role of Foreign System .... ............. 416 F. THE ANDEAN PACT ...... ............................ 418 a. Trade Liberation Program ..... ............... 419 b. Common External Tariff ....................... 420 c. The Sectoral Develolpment Programs ............ 422 d. Harmonization of the Andean Pact . e. Conclusions and ImpLications for Peru .... 424 G. INDUSTRIAL EXPORTS .............. .. ............... 426 a. The Role of Exports in the Present Industrial Policies . ........... .......... se.... 426 b. Past Export Performince ......... ......... 428 c. The Role of Industrial Exports in Peru's Long-term Developmnent .................. ... 429 d. Export-oriented Policies and Programs ....... 434 H. CONCLUSIONS ...................................... 437 TECHNICAL NOTE I: Determinants of Investment and Savings ....... 440 TECHNICAL NOTE II: Inflation, Real Interest Rates and the Demand for Money in Peru ................ 451 I Page 1 of 2 pages COUMTRY DATA _PER ARLEA 2 ______ATION -DENSITY 1285.2 ka thousand 16.5 million (mid-l977) 13 per kn Rate of Growth: 2.9 (from1l9TO to19T7) 54 per kmo of arable land POPULATION CEUAACTR$STICS (1977) HEALTI (1977) Crude Birth Rate (per 1,000) 441.0 Population per physician 1760 Crude Death Rate (per 1.000) 11.9 Population per hospital bed 500 3/ Infant Mortali_y (per 1,000 live births)65.1 TNCOME DISTRIBUTION (19Tl-72) DISTRIBUTION OF LAND OMNERSEIP Z of national income, M8106T, quintile 58.6 Z owned by top 10% of owners lowest quititlse 3.1 % owned by smallest 10Z of owners ACCESS TO PIPED WATER (1975) ACCESS TO ELECTRICI T (1972) Z of population - urban 72.0 Z of population - urban 349 - rural 15.0 -rural NVTRITION (1974) EDUCATION (1976) Calorie intake as Z of requirmemnts 100.0 Adult literacy rats Z 72.0 Per capita protein intake 61.7 Primary school enrollment % 111.0 l/ GNP PER CAPITA in 1977: US $ 830 GROSS NATIONAL PRODUCT IN 1977 ANMNAL RATE OF GRIOITH (M. constant Prices) uR $ Mln. % 71965-70 107_7i 1975-7 197q GNP at Market Prices 12422.1 100.0 5.4 5.6 0.0 0.1 Gross Domestic investment 1933.6 15.6 -1.5 15.8 -17.4 -15.4 Gross National Saving 1007.4 8.1 10.3 9;8 -4.9 44.5 current Account Balance -926.2 -7.5 Exports of Goods, NFS 1725.6 13.9 2.9 -7.1 7.7 17.0 Imnorts of Goods, NFS 2164.0 17.4 -1.0 10.0 -9.6 -31.1 OUTPUT, LABOR FORCE AND pRcDucTIV}y TR i976 {it 1970 pricesJ 2/ Value Added Labor Force V. A. Per Worker Ub S Mln, %. Thni % a S 7 Agriculture 1138.9 13.6 2109.4 43.4 539.9 31.5 industry and Mining 3167.7 38.0 1019.1 21.0 3108.3 181.2 Services 4030.9 48.4 1731.6 35.6 2327 8 135.7 Unallocated Total/Average 8337.5 100.0 4860.1 100.0 1715.5 10n0. GOVERNMENST FINANCE General Goverrueat Central Government qoles Btn.) % oE GDP (oles B In.) % of GDP 197T 1977 1972 1978 1978 1975-7A Current Receipts 190.1 18.1 19.1 264.2 15.5 15.2 Current Expenditure 219 7 20 9 17.4 2ed 7 17.3 17.0 Current Surplus -29.6 -2.8 1.7 -24.5 -1.5 Capital Expenditures 44.9 4.3 4.9 57.6 3.4 4.3 External Assistance (net) 14.9 0.9 2.3 1/ The Per Capita GNP estimate is at 1977 market prices, calculated by the saem conversion technique as the 1978 World Atlas. All other converxioms to dollars in this table are at the average exchange rate prevailing during the p-ri.od covered. 2/ Total labor force; unemployed are allocated to sector of their normal occupation. "Unallocated" c3nsists mainly of unemployed workers seeking their first job. 3/ 1975 estimate. not available not apolicable Page 2 of 2 pages COUNTRY DATA - PERU MONEY. CREDIT and PRICES 1973 1974. 1975 -1976 17 iI (Million outstanding end period) Money and Quasi Money 87506 107012 126174 151062 1988(0 320900 Bank Credit to Public Sector 20811 19773 43567 104900 198900 289800 Bank Credit to Private Sector 60361 73152 95788 134900 173800 238200 (Percentages or Index Numbers) Money and Quasi Money as % of GDP 24 4 23.9 22.8 19.7 18.9 19,2 General Price Index (1977 - 100) 37.8 44.1 53.4 71.9 loo.o 157.5 Annual percentage changes in: General PriL': Index 14.2 16.7 21.1 34.6 39.1 5T.5 Bank credit to Public Sector 77.0 -5.0 120.3 106.9 69.o 45.7 Bank credit to Private Sector 18.0 21.2 30.9 37.2 49.4 37.1 BALANCE OF PAYMENTS MERCHANDISE MEPORTS (AVERAGE 1976-78) US$ Miln a 1976 1977 1978 Fish meal 1T2T8 10.9 (Millions US $) Coffee 155.2 9.3 Other Agricultural goods 179.7 10.7 Exports of Goods, NFS 1744.6 2137.8 2399.9 Copper 342.6 20.5 Imports of Goods, NFS 2628.0 2694.4 2070.1 Silver 17L.8 10.4 Resource Gap (deficit -) *7_5. -32-9. Other Minerals 317.0 18.9 Petroleum, Crude and products 95.1 5.7 Interest Payments (net) -324.0 -371.9 -493.2 All other commodities 228. 0 13.6 Workers' Remittances - - - Total 1675.2 100.0 other Factor Payments (net) -42.4 -54.5 -84.5 Net Transfers 579 S 8 _.6. 0 gXTTRNAL DEBT, DECEMBER 31. 1978 Balance on Current Account -1191.9 -926 .2 -191.9 Direct Foreign Investment 170.8 54.1 25.0 Net MLT Borrowing 2/ Public Debt, icl. guaranteed 5046. Disbursement 1144.4 i026.4 866.5 Non-Guaranteed Private Debt 378. Amortization -305.4 -406.4 -478.9 Total outstanding & Disbursed 5425. Subtotal 839.0 620.0 387.6 l/ Capital Grants - - - DEBT SERVICE RATIO forg978 Other Capital (net) -338.3 -97.6 -86.7 Other items n.e.i 0 . _S5 n increase in Reserves 2/ 520.4 349.7 -82.0 PubLic Debt. incl. guaranteed 33. Non-Guaranteed Private Debt 4 Net Reserves (end year) 2/ -355.3 -705.0 -623.0 Total outstanding & Disbursed 36 RATE OF EXCHANGE IBRD/IDA LENDING, ( Dec ic78 ) (Mtllion US S) Averares - 1977 IBRD IDA US $ 1.00 = 81.97 Soles Soles 1.00 - US $0.01220 Outstanding & Disbursed 185.6 - Undisbursed 234.5 - Average - 1978 Outstanding inl. Unsiabursed 420.1 - us $ 1.00 = 160.00 Soles Soles 1.00 - US $0.00620 1/ Ratio of Debt Service to Exports of Goods and Non-Factor Services. 2/ AdJusted for Balance of Payments support loan. not available t not applicable I - 4h - PERU: LONG-TERM EiEVELOPMENT ISSUES INTRODUCTION 1. The most salient feature of Peru's socioeconomic development is the contrast between regions, between groups of population and between periods in its history. The Andes Mountains divide the country into three regions with distinct geographic and climatic features, ranging from desertic coastal plains, to cold Andean highlands and t:o tropical humid jungle lowlands. Since the beginning of colonization in the early 1500s, Peru's socioeconomic devel- opment has superimposed additional segmentations to the basic geographical diversity. Growth and development have followed a pattern of periods of boom, based on the concentrated exploitation of one or two primary products, followed by periods of stagnation and decline when the product was exhausted or its world demand collapsed. This poattern has concentrated modern economic activity in a few growth poles in the Coast and particularly in the capital, Lima, while the rest of the country has lagged behind. On the other hand, during the colonial administration a slocial structure was imposed -- more rigid than in most other countries in Latin America - that remained basic- ally unchanged until the first part oi' the century. At the top of the social pyramid were the descendants of the colonial administrators who benefitted from landownership and the export boomLs but did not evolve into a broadly based entrepreneurial class; most of the entrepreneurial talent in Peru has come from relatively recent immigrants during the last century. In contrast, the native population has remained concentrated in the Andean highlands and has benefitted little from the country's development. This geographic, economic and social segmentation is a basic characteristic which the Govern- ment is facing in fostering more balanced development in Peru. 2. Within this long-term perspective, the years since 1950 have been in many ways a successful period in Peru's economic history. Governments have become increasingly aware that the lonig-term prospects of the country depend heavily on changing the development pattern to reduce dependence on the vagaries of one or two primary export products and to increase the participa- tion of a larger share of the population in the development benefits. Average output growth has surpassed a respectable 5.5% per annum and the export struc- ture has been diversified. Governmenit action in providing basic education and health services has been noticeably intensified and a sweeping land reform has profoundly changed land ownership patterns. Finally, through massive migration into the coastal cities, a Larger share of the population has been incorporated into the modern sectors of the economy. 3. Rowever, the last 25 years lhave also brought into focus various critical issues that will constrain flture development. Sharp cyclical fluc- tuations of economic activity have continued. This has partly been caused by the export structure, which, although more diversified, remains concentrated in primary products. Not only are these products subject to violent fluctua- tions in world market conditions, but they are also associated with large, lumpy projects that introduce strong swings in investment activity and export volume; this is particularly true in the mining and petroleum sectors. Of equal importance has been the deterioration of the savings performance since the early 1960s, which has constrained the investment efforts and rapidly i I - 41 - increased the external indebtedness of P'eru. While the role of the State in the economy and the share of resources :.t manages have increased enormously, public savings have remained very low. At the same time, the-average savings rate of the private sector has also declined since 1960, partly because of inadequate price policies, but probably more so because of the uncertainty created by the Government's efforts to change the basic structures of the economy. The results from these basic changes has also highlighted the diffi- culty of attacking the poverty problem :Ln Peru. Since 1969, the Government has redistributed almost half of the cultivable lands, yet this favored only a fifth of the rural population; the problem of the minifundista with very little and poor land and of the landless migrant laborer remains. Similarly in urban areas, wage legislation and ownership-sharing schemes have redistri- buted income to the permanent workers, but these are only a small share of the labor force. At the same time, the efficiency of resource use in the productive sectors remains low (particularly in manufacturing) and the lack of management capacity has increasingly become a problem (particularly in the public sector). 4. This brief introduction provides an idea of the enormous task the Peruvian Government faces in fostering balanced development. Nonetheless, the Government also has important assets at its disposal, the most important of which are the large mining resources and the tenacity of the Peruvian people. Comparing Peru of 1976 with that of 1950, much progress can be recorded, although much more will have 1:o be done before the conditions for long-term, self-sustained growth can be attained and before the fruits of development reach the poorest half of the population. 5. This report aims at providing background for understanding Peru's development problems and analyzes some of the issues listed in the above paragraphs. The focus of the report is primarily long term: the issues that will affect Peru's development over the next 25 years rather than the details of the past or future 2-3 years. The short term is analyzed only when it affects the long term. Put another way, the focus is understanding the past and present to better design future programs with long-term consequences. The report is far from comprehensive and the relative weight given to specific issues may not reflect the Government's priorities, since it tries to empha- size issues in which the World Bank has little information. 6. The report is divided into three parts. Part A defines the setting in which development has to take place. It first reviews the natural resources and population of Peru, emphasizing the opportunities and constraints they will impose on development in the next 25 years. The way these resources have been used to foster development in the past and the legacies that history has left are reviewed next. Finally, the Government's strategy of the last eight years is reviewed in more detail, assessing briefly the successes and short- comings and the priorities set for their future. Part B analyzes the main macroeconomic characteristics of Peru's past development and the main con- straints that are likely to recur. Finally Part C presents the main issues and prospects in the principal economic sectors: agriculture, industry, and mining. 7. The purpose of the report is primarily to provide a long-term frame- work and reference for World Bank staff working with Peru, although it hope- fully should also be useful for other foreign lending agencies in their design - 42 - of programs for Peru. Finally, although it is exceedingly difficult for a foreign group to understand in depth the large complexities of the Peruvian economy, the Mission hopes the report will be in some way useful to the Peruvian Government as an independent view of the country's-development issues. The report does not question the validity of government objectives, nor does it address directly the political dimension of Peru's development process. However, it does address some issues which will have to be faced whatever political choices will be made in the future. - 43 - PART I: THE SETTING; NATURAL RESOURCES, POPULATION, ECONOMIC HISTORY AND GOVERNMENT STRATEGY - 8. Part I discusses the basic setting to understand Peru's development problems and prospects. Chapter I reviews the physical natural resource base, its consequences in shaping the basic parameters within which development has to take place and the policy implications. Then, Chapter II discusses the basic characteristics of population growth and its spatial distribution and assesses how future population growth and policy will affect development. Chapter III presents a brief overview of Peruvian economic history highlighting the main legacies that it has left in shaping the Peru of today. Finally, Chapter IV reviews the development strategy of the Government. CHAPTER 1: GEOGRAPHY, NATURAL RESOURCES AND DEVELOPMENT A. GEOGRAPHY AND CLIMATE 9. Peru, the fourth largest country in Latin America, 1/ is divided by the Andes mountains into three regions with extreme contrasts in topographic and climatic features. Between the 3,000 km of coastline along the Pacific Ocean and the Andes is a narrow belt covering 11% of the country (the Costa). It is made up mainly of dry, flat plains and sand dunes below 300 meters and the equally dry Andean foothills sharply eroded by the rivers flowing west. The climate in the Costa is dominated by the effect of the cold von Humboldt ocean current. 2/ From April to December most of the Costa is clouded, with high humidity but no rain. The Andean Highlands (the Sierra) above 2,000 meters, covering 26% of the country, are composed of steep mountain slopes and high valleys located between the mountain ranges. The latter are rela- tively flat plateaus, the broadest and highest area of which is the Altiplano in the south. The Sierra climate varies widely during the year and at 1/ Peru's area totals 1,285,000 km2 equivalent to the European Common Market minus the United Kingdom. 2/ From above the von Humboldt ocean current, cold moisture-laden air is blown into the Costa by winds froti the Southwest. When the lateral move- ment of the winds is inhibited by the Andes, they form a static cloud layer that covers most of the Costa fromi April to December causing low tempera- tures with high humidity, but no rain. During the rest of the year, with the southern movement of the sun, the cloud layer disappears, bringing sunshine and a lowering of the high pressure. _44- different altitudes. 1/ East of the Sierra are vast tropical lowlands cover- ing the remaining 63% of the country (the Selva). The Selva is subdivided into the Ceja de Selva (or Piedmont) and the Low Selva (below-600 to 700 m). The former is made up of alternating hills and valleys in the eastern Andean slopes and has continuous rainfall of 1,000 to 3,000 mm with average daily temperatures of 18 to 21 0C. The Low Selva is a very hot and humid plain (30 to 33 0C and around 2,000 mm of rain well distributed over the year) of dense vegetation consisting predominantly of rain forest. 10. The rugged topography limits trade and interdependence between the three regions of Peru and modern economic activity has concentrated in the Costa, particularly in Lima. The ease of land transportation within the Costa and the existence of various good natural harbors is a partial explanation for this. Transportation and communications between regions and within the Sierra and the Selva are extremely costly and difficult. Constructing roads or railroads outside the Costa is a major and often uncertain engineering challenge. 2/ Roads deteriorate rapidly -- and sometimes disappear --- and maintenance is expensive. Still today there is no all-weather road connect- ing the Costa and the Sierra with the Selva, and roads within the Sierra are scarce and rudimentary. The lack of transportation leaves many areas isolated from potential markets and constrains the economic development of new areas in the Sierra and the Selva. B. NATURAL RESOURCES a. MininR 11. The unusual topographic and climatic features also result in a varied natural resource base. The largest asset is the vast and varied deposit of 1/ The Sierra climate varies widely with arid western slopes having rain up to 200 mm in the December-April period, little cloud coverage and tem- peratures varying from 13 to 20°C, depending on altitude. In the Central Sierra, rain is concentrated in the December-April period when it ranges from 600 to 1,000 mm (although with wide annual variations). Temperature fluctuates between 5 and 200C. The Altiplano in the south is cold and humid owing to both the presence of Lake Titicaca and the altitud.e. Precipitation of 600 to 1,000 mm falls from December to May, some in the form of snow and hail. 2/ To illustrate the difficulties in building east-west roads in Peru, the Central Highway, connecting Lima with La Oroya in the Sierra and Pucallpa in the Selva (now in construction), climbs from sea level to about 4,800 m at the highest point (there is only one crossing of the Peruvian Andes in the north at 2,200 m; all other crossings are higher than 4,0CO m). Then, after crossing three major mountain ranges with deep canyons and torrential streams, it reaches the Ceja de Selva at about 600 to 1,000 m. Even there, heavy rainfalls and difficult soil conditions threaten the stability of this road. - 45 - minerals located mainly in the Sierra and the southern coast. Table 1 gives an idea of the enormous proven reserves of the four principal minerals now exported, although large areas of the country have not yet been geologically explored. 1/ Mining has been and will clearly remain one bf-the main pillars of Peru's growth, both by supplying savings and foreign exchange and by fostering technological development. However, mining provides little direct employment per dollar invested and will have for a long time few linkages with the rest of the economy. Because of their strategic importance to the economy, during the 1970's the Government has taken over all the large mining companies, except one, and intends to maintain a substantial state participation in the sector's future development. The capacity of the new state mining corporations to prepare, implement, finance and manage new large projects will be one of the most important factors in determining the rate at which the sector can grow. Chapter IX deals in more detail on these and other major mining sector issues. Nsble 1: MERV OF D": p!£.'.ALZ (volum in al2Ioas of tom1i ezceW silTer in million croy oz) Froduation Mr:izated P-oveo ?e-prres I/ Val=uin v a13 ia =izsber of yeurs 19T4 st 197.4 pro4dAton leel Copper 0.223 27.13 121.1 Zin 0.39T ?,,3 16.4 Lad 0.207 2.7-3-2 13--15.5 frog 9.473 T80 82.3 Silver !40c 500-TW 12.3-17.2 v/ aesieres vhtxe exis:ence is pram and could be 1L.ned eco-azically at toda s prices. b. Fisheries 12. The second major natural resource asset is the large fishing poten- tial. 2/ Fishing gives three importani: opportunities to the Peruvian economy; it provides foreign exchange, 3/ it has developed important backward linkages 1/ Other minerals with substantial reserves and now in production include antimony, bismuth, cadmium, indium, mercury, molybdenum, gold, selenium, telerium and tungsten. In addition, reserves totalling 350 million tons of phosphates and 60 million tons of low-quality coal have been proven. 2/ Peru enforces a 200-mile limit to protect its fishing resources. It was one of the first countries in the world to do so. 3/ The share of fish products in total exports has fluctuated widely. On average, the share was 12% in 1960-62, 26% in 1965-68, 34% in 1970-71 and less than 14% in 1976. -46- with the rest of the economy and is a relatively labor-intensive activity, 1/ and it provides a large potential source of protein for human consumption. 2/ However, the magnitude of the fishing potential remains uncertain and subject to violent fluctuations of ecological conditions. The cold von Humboldt Current contains an abundance of fish food which make the offshore waters one of the world's richest fishing grounds, while the climate facilitates fishing operations throughout the year. Anchovy catch, the most important fishing resource, boomed during the 1960s because of the growing world demand for protein-rich animal feed and the invention of nylon nets which lowered costs of production. 3/ Anchovy catch peaked at 11 million tons in 1971 (equivalent to well over half of the world traded production), but fell drastically in the next two years and has not yet fully recovered. For still unexplained reasons, at least every 7 years warm surface currents that normally turn seaward in Ecuador push far south. The warming of the water adjacent to Peru then results in decreased anchovy catches. This phenomenon has been recurring in recent years. All indications are that Peru has large resources of other species, particularly hake, bonito and tuna. However, there is little precise information on amounts that can be fished on a sustained basis; recent expe- rience has been disappointing and the danger of changing ecological conditions and/or of overfishing remains large. The State now controls all the processing and foreign marketing of fishmeal, 4/ and has made major investments in developing fishing ports. c. Agricultural Land 13. In contrast to mining and fisheries, the availability of land suit- able for agriculture is extremely limited (see Table 2). Most of the soils suitable for intensive agriculture (less than 3% of the country) are already under crops; most of the lands not in farms and with agricultural potential are either in the Ceja de Selva (with high costs of access and clearance) or are in the marginal class. The ratio of croplands actually harvested to rural popula- tion is less than 0.35 ha and varies widely from 0.5-1.0 ha in the Costa to as low as 0.1-0.2 ha in the major Sierra departments (Puno, Cuzco and Cajamarca). 1/ Relatively efficient industries building boats and producing nets are examples of the beneficial linkages created by the fishing sector. 2/ This is particularly important since Peru is expected to continue facing serious food shortages. 3/ The cotton nets used previously were weak, subject to rot, heavy, awkward to handle and lasted only 2-3 weeks. The nylon net solved all of these problems, lasted 2-5 years and its cost was only double that of cotton nets. (See Roemer, M., Fishing for Growth, Harvard University Press, Cambridge, 1970, p. 83.) 4/ Between 1973 and 1976 the State also owned all fishing operations for fishmeal production. These were returned to the private sector i.n mid- 1976. - 4T - Agriculture in the Costa prospers in :he small irrigated areas where, with the use of relatively advanced technology, yields are high and over 40% of the country's gross value of crop productLon is generated. Although the harvested area in the Costa could be expanded by the rehabilitation of around 200,000 ha of land (that now suffer from poor druinage and salinity) and by bringing new land into cu tivation with high cost Lrrigation infrastructure, most of the production increases will have to come from higher yields. Table 2: LAM POT=MIAL AED ITS USE F-OR AE:CUVLTUS (in millioan hectares and perceat) Total Costa Sierra SeIl Lazd Suitable for: 1/ Tota2 127.8 100.0 Izenire Agriculture 3.5 2.T Permnent AFicutur i/ 16.0 13.0 Marginal Agiculture 58.5 45.6 lat Suizable i 49.8 38.7 Land Used ror: 1/ Total 127.8 100.0 13 100.0 33.2 100.0 81.6 100.0 croplad 3.7 2.9 0.8 6.2 2.3 6.9 0.6 O.7 Saji-fed (2.4) (1.9) (0.1) (0.8) (1.8) (5.4) (0.5) (0.6) rr±gtstd (1.3) (1.0) (0.7) (5.4) (0.5) (1.5) (0.1) (0.l) Other Land in Ia&= 19.8 15.5 1.0 T.6 17.2 51.8 1.6 2.0 Sot in farms 104.3 81.6 11.2 86.2 13.7 41.3 79.i& 97.3 Area Earvested 1.9 1.5 0.5 3.9 1.1 3.3 0.4 0.5 I/ Srr7eyed in 197. 2/ inclites fruitsa pastwr ad forestry 3/ Usable for pastures aud forestu7 i/ Not suitable for an rorm. of aiculture, forestry or pastursge 1972 Census & EaturaL grslanda, foreat and woodlands, mud other. Sou-e-: 0EW. Capacidad da Usa in los Sualos del Pard (Tercers Avroxi- ci6) :un, 1972 14. The Sierra accounts for well. over half of the cultivable land in Peru but, except for a few valleys with permanent streams, it is a very unfavorable environment for agriculture. Before colonization, the Incas had terraced large areas, but after the conquest, terrace maintenance was abandoned. This, and subsequent deforestation, 1I accelerated erosion and large areas have been lost permanently for agriculture. Less than 7% of the lands are used for crops; over half of the remainder is used for grazing, mostly under marginal conditions since steep slopes prevail over most of the Sierra. Crop farming is extremely difficult, risky and limited to one short growing season from 1/ Timber was used for mine tunneling and for fuel. - 48- about November to March. Besides severe topography, rainfall is limited (distributed erratically within and between years) and temperature changes impose a constant risk of frosts. With the low and gradually deteriorating soil fertility, fallow periods have lengthened and now reach-up to 5-7 years. At altitudes above 3,700 m conditions toughen further in the Sierra. Yields of the major crops I/ fall by as much as 40% and only sheep and Criollo cattle can withstand the altitude. Beyond 4,200 m only llama, vicuna and alpaca can survive. Around one-third of Peru's population makes a living in the rural Sierra -- of which over one-fourth live at altitudes above 3,500 m -- prac- ticing a type of farming which is often referred to as primitive, but could equally well be qualified as ingenious, given the precarious resource base. Most of the increase in production in the Sierra will have to come from improved pasture and livestock management, more and better managed small-scale irrigation works and the increased use of fertilizers and improved seeds; there is little scope for putting new lands into production. 15. Land in the Selva has limited potential for agriculture, except in some parts of the Ceja de Selva. ONERN, the Government's natural resource evaluation agency, has classified the soils in the lower Selva as marginal and mostly suitable only for extensive grazing and forestry. The soils generally have low natural fertility levels and under the heavy rainfall and temperature regimes, after cultivation for a single season uses up the humus, further soil deterioration is rapid. The limited production is confined almost entirely to seasonal crops grown on the mud flats exposed by the receding annual flood waters of the Amazon River and its major tributaries. Soils in the Ceja de Selva are generally of better quality, and although communications with the rest of the country are exceedingly difficult, the area does offer a limited potential for expansion of the agricultural fron- tier -- in areas like the Huallaga Central and the Palcaza-Pichis settlement area. 16. The relative scarcity of the agricultural land resource base has serious implications for Peru's long-term development. As is discussed in more detail in Chapter X, if major efforts are made to expand the agricul- tural land frontier and to improve yields, output may grow by 2-3 percent annually in the long run. Barring a major technological breakthrough, which at present is unlikely, a higher growth rate will be difficult to achieve. This has two implications. First, if rural incomes are to increase oa a sustained basis, the population there will have to stop growing or decline. And second, it will be difficult for Peru to reduce its dependence on imported foodstuffs, which has been growing rapidly in the last decade. In fact, depending on the growth of the population, the food gap may increase further with serious consequences on the balance of payments. d. Water 17. Peru has abundant water resources, but their geographical location is unbalanced and the flow of rivers fluctuates much throughout the year. The 1/ Potatoes, quinua, maize. - 49 - Andes divides rivers into two systems. To the east a complex and extensive river system flows into the Amazon Basia. They have an immense water flow and provide for a large share of Peru's untapped water resources. To the west of the Andes, about fifty rivers flow throuigh the coastal region into the Pacific. As shown ia Table 3, most of these rivers are seasonal streams that swell in the summer months to 10-20 times their mininum flows. These rivers provide the bulk of water needed for irrigated agriculture and urban domestic use in the Costa. However, the seasonal fluctuations substantially reduce the use of the river resources -- particularly for expansion of irrigated agricultural land -- without difficult and costly flood control and catchments. 18. Since the migration trends are for a continued transfer of the population from the highlands to the dry Costa, it is important to assess Tal I P1: PDUms WAT5ZII LANCmS IN OAETAL A1A, 2000 (VaptioUMSa tbouasnd. an1 othert in ,?Isec) Domestic DopaAmut Projeote4 water Amiilable river vacer Potential deficits or City urba popul-4 n.ed a/ A="r.66 Average Xi. From avorae From minit Tuhes 131 .25 105.01 9.1 Pils 1,208 2.38 101.1) 6.0 Piurt 348 .68 13.1). .68 SullA 212 .41 68.13 6.o TalS - 34 .07 - .h -07 LaabWq-e 1.236 2.43 b4.o 7.2 chi..ayo 643 1.27 2a. 3 5.0 L Libertad 1.949 3.83 53.D 3.2 o.63 rujiLlo 1.3;S 2.59 8.3 0.2 2.57 Altub 1,156 2.27 1149.D 2T.8 Chimbote 763 1.50 l44.o 27.T LisA 10,896 38.26 179.0 47.6 3nACbo Si .16 24.0 9.4 Lims 9,737 36.06 25.0 11.8 11.06 24.26 Ica 901 1. T 60.0 1.36 .43 la 267 .53 8.0 .53 Pi-so 128 .25. 26.0 1.2X Cincha AItd 176 .35 21.0 0.08 .27 Areqqipa 1,185 2.33 125.0 36.2 Aequiys. 949 1.&7 11.0 4.6 bq-ueaua 326 .64 . - .64 .64 T . 253 .50 .80 .46 .o4 235 .46 .80 .416 T*thl Coat&L Urba 19- 54.t _/ For aszu=;Czions see Chi2tcr 1_ /. For Esl:izr.s see tecx.. I .krarae 163-S72 Source: 023-11, 2E =r ' issica aszi--^tes. - 50 - whether water supply will become a constraint to urban development and the source of large investment requirements. 1/ Table 3 attempts a rough projec- tion of availabilities and possible water requirements in the main urban areas for the year 2000. 2/ While these estimates are rough, in-the absence of other more comprehensive studies of overall water balances and projections for Peru, they do give an overview of the potential existence of problems and directions for investment. The table suggests that only two major areas, Lima and Trujillo, stand to have significant problems in domestic water supplies. Other minor deficits may arise along the coast, but they can most likely be filled from wells. Given the topography of the Peruvian west coast, several cubic meters per second of ground water seem to be continuously available, 3/ thus making it likely that outside of the two large metropolitan areas perhaps only Piura, where the mountains are a good distance from the coast, will have an unfilled deficit. To eliminate this deficit, work is currently utderway to transfer water from the Chira to the Piura river. 19. In the Lima and Trujillo areas, there are water surplus rivers relatively close, but the transfer requires investments of close to $0.5 billion at 1976 prices. To satisfy Lima's requirements, for which a shortage is expected in the 1980s, first the very high leakage will have to be reduced 1/ Rivers and availability of water have been important determinants for the location and growth of Peruvian cities in the Costa (see Chapter II). 2/ The projections focus on domestic consumption in the urban coastal areas since: (i) all indications are that east of the Andes there is surplus water; (ii) the rural population generally can derive their water from small streams and wells; and (iii) general trends in other countries indicate that domestic water requirements displace agricultural require- ments as an area develops. The projections are further based on the following assumptions: (a) the projections of distribution of population by the year 20)00 by cities explained in Appendix Table 1.70. (b) domestic consumption of water in liters/head/day (LHD) will be 419 at Lima (assuming that with the reduction in leakage from 50 to 25% the domestic consumption will fall from 558 LHD at: present to 419 LHD) and 170 LHD at other cities (similar to that of other medium-sized cities in Latin America, for example Cali is 170 LHD and Managua is 160 LHD); and (c) that 100% of the population will be covered with piped water systems. The national average for urban areas, according tc' the 1972 Census, is 55%. 3/ At present Lima derives 5.3 m /sec from wells, while in the Moche Valley (at Trujillo) the figure is 1.6 m3/sec. - 51 - from 50% to a more normal 25%. After that water resources from the east side of the Andes will have to be tapped. A recent study by the National Electricity Institute concludes that the cheapest source would be the Mantaro River, provided it is freed from mining pollutants. At an estimated cost of $250 million, about 15 m3/sec could be transferred, which would suffice until late in the century. Beyond that point, it is estimated that the demand for hydroelectric power will grow more quickly than that for water, and as new hydro schemes are added (the eventual maximum available from the Mantaro, however, would be 43 m3/sec) water would automatically be available. The future costs of the transfers of the Mantaro waters would be much higher, if it were not because the tunnel through the Andes already exists 1/ and, since Lima is downhill from the water surplus area, energy requirements are minimal or negative (as the water runs through hydro-generators on its way downhill). 2/ To cover the deficits in the Trujillo area, water can be transferred from the Santa River 3/ through a canal of approximately 150 km. The cost of this project, including the required catchments, would probably be close to that of the Mantaro Project. 4/ 20. In conclusion, water resources in Peru seem to be adequate to cover requirements for a long time, but their management will require heavy invest- ment to maximize the use of rivers flowing west and transfer water to the Costa from rivers flowing east. An important policy issue is water pricing. At present, water tariffs for both irrigation use and domestic consumption are very low. In Lima, for example, despite their quadrupling in 1975-76, tariffs still barely cover operational costs (but not maintenance or capital costs) and are only two-thirds of what they sre in Bogota. 5/ Bank staff estimate that by the early 1980s tariffs will have to be increased 3-4 times to cover costs fully (at 1976 prices). A more adequate tariff policy would discourage wastage and reduce the long-term expansion requirements of water supply. e. Energy 21. Potentially, far more serious than the water shortages are the growing energy deficits Peru may face in coming decades. Up to the early 1960s Peru was a net exporter of energy; since then, net imports have been growing rapidly and by 1974 accounted for almost 30% of domestic energy con- sumption (Table 5). Over the past 25 years, energy use has grown at an 1/ Only a widening of the tunnel is necessary. 2/ This is in sharp contrast to a country like Mexico where the dry areas are uphill from the water and where by the year 2000 the pumping energy required will be considerable. 3/ The Santa River is the largest oii the western coast with an average flow close to 150 m3/sec. 4/ At 1976 prices, it is estimated that the cost of transporting 5 m /sec is $1.5 million per km. 5/ The minimum charge for water is $0.04 per m3 in Lima and $0.07 in Bogota. - 52 - average of 6.3% p.a., with an elasticity with respect to real GNP of about 1.2 (about average compared to other developing nations). Energy production, on the other hand, grew by only 2.5% p.a. over the same period. 22. Petroleum has become Peru's main energy source and was a major foreign exchange earner up to the early fifties. The bulk of oil reserves were located in the northern coast. 1/ However, since there was practically no exploration in the sixties, production grew little. After the creation of PETROPERU in 1969, the Government stepped up exploration in the eastern Amazon Basin, encouraged by positive geological studies and major discoveries in the adjacent Ecuador Amazon Basin. 2/ Then in 1971-72 both PETROPERU and Occidental discovered oil in their first drillings. The 1973 quadrupling of world oil prices in 1973 also increased the attractiveness of jungle explorations despite their extremely high cost. 3/ These positive factors encouraged the Government to intensify exploration efforts on its own 4/ and to embark in the construction of an 835-km pipeline from the jungle to the coast (with an initial capacity of 200,000 barrels/day) at a cost of over $750 million. This was considered the only way of bringing out the petroleum if large quantities were found. 5/ The petroleum sector became the 1/ About 65% of oil reserves used in the past have been located at l'alara in the northeastern coast (where production started in 1863). AIL addi- tional 30% has come from the continental shelf in the northeast (where operations started in 1956). 2/ In the 1971-75 period an estimated 58,000 km of seismic lines were shot that identified some 159 possible oil-bearing structures in the Peruvian Amazon Basin. The investment was over $180 million. 3/ The cost of onshore seismic work and drilling is by far the highest in the world. 4/ To promote petroleum explorations the Government concluded 16 production- sharing agreements with foreign companies. The contracts (known as "Modelo Peruano") had the following features: the land is owned by the State; the gross volume of oil produced is split between PETROPERU and the contractor (usually on a 50-50 basis); the contractor carries all the risk and pays all development and operating costs but can export its share without surrendering the foreign exchange to the Government (if the Government purchases petroleum from the contractor, it has to pay in foreign exchange); all taxes are paid by PETROPERU from its share; and the contractor has a specified maximum time to find petroleum (or he loses the concession) and once in production, the contract lasts around 35 years. In 1974 the Government decided to stop granting contracts and do any new explorations on its own. In July, 1976 this position was reversed. 5/ The original cost estimate in 1974 was about $340 million. The pipeline (36 inches in diameter at the widest point) will have an initial capacity of 200,000 barrels per day which can be boosted to 500,000 barrels daily by the addition of pump stations. - 53 - first priority for the Government; by 1974-76 it accounted for 37% of the public investment program. Unfortunately, although the pipeline construction is completed, explorations have yielded very limited results so far. 1/ Follow- ing the early successes, no new large commercially viable oil fields have been found as yet. 2/ The total proven reserves in the jungle were in 1977 around 550 million barrels. Over half of the jungle area is still unexplored and the geological studies continue to suggest the possibility of oil formations, but the past low successes in drilling does not provide a basis for much optimism. Recently, renewed efforts have been made to search for petroleum on the con- tinental shelf; the proven reserves there and in the coast stood in 1977 at around 250 million barrels. Table 4 shcws that under the present Government targets of producing 195,000 barrels per day by 1980, the presently proven reserves would be exhausted by 1990. 3/ 23. Hydroelectricity is the seconcd major potential source of energy in Peru. Although during the last decade poroduction has been growing at almost 9% annually, hydroelectricity still accounts for less than one-tenth of energy consumption and for three-fourths of electric generation in the country. While present installed capacity can generate the equivalent of 9.2 thousand barrels of oil per day, the estimated total hydroelectric potential is for at least 20 times more. 4/ The Government intends to further reduce the dependence of electricity generation on petroleum and is now preparing proj- ects to rapidly increase the hydroelectric installed capacity. However, a realistic long-term target for the yearLy growth of hydroelectric capacity is not more than 9-10%, given the structurie of demand for energy and the insti- tutional constraints. 5/ Peru also has vast reserves of coal (but they are 1/ The issue of the justification of the pipeline is highly controversial. In this respect it is important to note that the largest share of the costs are the laying of the pipeline; the tubes themselves account for less than 15% of total costs. Thus, a pipeline of lesser capacity would not have been substantially cheaper, and the possibility remains that larger quantities of petroleum may be discovered. 2/ In the first fourteen structures drilled, the success ratio (number of structures that find oil over number of structures explored) was 60%. However, in the next 27 structures drilled the success ratio fell to 11%. 3/ Since the foreign contractors are to keep around half of the oil they produce, the maximum domestic supply for which foreign exchange does not have to be paid is around 137,500 barrels per day for the 1980-89 period (see footnote 2 to Table 4). 4/ The present installed capacity guarantees around 7-8,000 Gwh per year. The estimated potential is for 109,000 Gwh. 5/ Even the target of 9-10% yearly g:rowth of installed hydroelectric capac- ity may be highly optimistic, sinc:e demand projections indicate a need for much less electric energy (se,a footnote I of Table 6). 54- - Table 4: PERU; INDICATIYVE PROJECTIONS OF PTTIOLEM PEODUCTION AND CONSUMPTION, 1974-1990 1974 1977 1978 1980 1985 1990 A. Per day - in thousands of barrel, Production 74 t9 116.5 170.0 195.0 195.0 11.0 Jungle - 47.5 100.0 125.0 125.0 11.0 PETROPERU C - ) (35.0) (45.0) (45.0) (45.0) (11.0) Contractors ( - ) (12.5) (55.0) (8o.o) (80.0) (-) Other areas 74.9 69.o 70.0 70.0 70.0 Less: share of contractors - 22.3 45.0 57.5 57.5 Total domestic suoply 74.9 94.2 125.0 137.5 137.5 11.0 Domestic demand 3/ 110.6 120.0 129.0 150.0 205.5 265.3 Balance -35.7 -25.8 -4.o -12.5 -68.0 -254.3 B. In millions of barrels Reserves stock (beginning of period) Jungle 500 550 533 460 232 4 Other areas 250 280 255 229 101 - Production 27.3 42.5 62.1 71.2 n.2 4 Domestic supply 27.3 34.4 39.2 48.4 48.4 2 Deficit L3.0 9.4 7.5 4.6 24.8 92.8 Deficit (in $ mill.) 143 104 16 50 273 1021 1/ Assumes projected level of production of PEIOPERU up to 1980. After that assumes sane level as in 1980 until exhaustion of economically proven res4rves. i/ Unfdr -: sent contracts, foreign contractors are entitled to keep 50% of their pro. ducsion -.rhich does not generate foreign exchange for Peru. The other half is tranz- rer--d to F7TROPERU. Calculated as 1/2 of contractors in the jungle and _14 of "cor nceas" (since 1/2 is by contractors). 3/ Vea: i. r 1985 and 1990 taken from Table 6 by subtracting from the energ require- n: :s on the 6% long-term growth Faternative, the production of non-petroleum energy. Fo- 1977 to 1980 interpolated taking into account the impact of the 1975-T7 recassion. 4( Deficit converted to dollars at 1976 prices of $11/barrel. Source: Mission estimates. predominantly anthracitic, a fuel which is difficult to burn 1J) anci some limited proven reserves of gas. 24. Unless new sources of energy reserves are found, energy could become a major constraint to Peru's development. Table 6 was prepared to illustrate broad magnitudes that the energy gap could reach by the year 2000. The annual 1/ Economically recoverable proven reserves are estimated at the equivalent of around 511 million barrels of oil. In addition, there are 500 million barrels that are proven but not economically recoverable at today's prices. Recently, coal mines in the Alto Chicamsa area were discovered capable of producing the equivalent of around one million barrels of oil per year for up to 50 years. - 55 - growth rate of energy requirements is projected to range between 6% (the historical rate) and 7.5% (equivalent to an elasticity of 1.3 with respect to real GNP 1/). The accumulated energy requirements for the-1975-2000 period would range between 3 and 3.9 billion barrels of oil equivalent. On the other hand, the prclected accumulated production of energy for the period -- even assuming an optimistic rate of growth of hydroelectric capacity -- is only 1.3 billion barrels (excluding the share of petroleum kept by foreign contrac- tors). This leaves an energy gap to be imported of around 1.7-2.5 billion barrels. At 1976 prices of approximately $11/barrel, this would cost between $19-28 billion. Achieving self-suffic:Lency in this period would require that between 3 and 5 times as much petroleua be found in Peru as already has been found in the Amazon Basin. The distribution of the gap during the 25-year period will depend on the yearly rate at which petroleum is extracted. Assuming the probable pattern of production shown on Table 4, net imports of energy will fall substantially in 1978-79, as jungle output becomes significant, but increase rapidly thereafter. In dollar terms (at 1976 prices) the annual energy gap could rise from around $50 inillion in 1980 to $1.0-1.4 billion in 1990 and $1.7-2.8 billion in 2000. 2/ 25. The most important conclusion suggested by these projections is that a coherent long-term energy strategy is urgently needed to minimize the burden of energy imports on the balance of payments. The strategy would have to aim at reducing the growth of consumption, maximizing the use of energy resources available domestically and developing new sources of energy. This raises two related policy issues: pricing and investment. In the past, Peru has followed a policy of cheap and subsidized prices for energy (particularly for petroleum products), even in the 1960s and early 1970s when the energy gap was widening. 3/ This had two important consequences. On the one hand, it 1/ The level reached by other developing countries with growth acceleration. 2/ It is useful to put the potential demand of household electricty into the perspective of Peru's overall, energy needs. Currently domestic electricity requires about 1,000 Gwh annually, the equivalent of 1.6 thousand barrels of oil per day, or about 1.2% of the country's total energy bill. It covers only 4.5 million people, about one-third of the population, giving each person who actually has electricity an average of about 200 kw per person annually. (This amounts to being able to use a one-hundred watt bulb for five hcurs each day.) Extending current levels of coverage to the entire population of 2,000 would entail only 6,500 Gwh annually (equivalent to 10.5 thousand bbls/day), and building into this complete coverage the natural levels of increased use that will come with increased income, brings the total demand for domestic use of electricity in the year 2000 to about 18,000 Gwh -- about one-third of the total potential that could reasonably be generated from hydro sources in that year. In terms of overall energry demand, universal household use would amount to less than 5% of total energy demand in the year 2000. 3/ Prices were either below world mirket prices or did not cover production costs. See Chapter VI for a discussion of the impact on public sector finances of the petroleum consumption subsidy. - 56 - was not conducive to an efficient use of energy resources and demand grew very rapidly (particularly for private transportation). On the other hand, the internally generated investable resources of PETROPERU and other enterprises in the energy sector shrank; this limited investment and caused increased external borrowing. 1/ More recently, the Government has taken important price action. Power tariffs in Lima were increased by 40% and in dollar terms the price of regular gasoline was tripled during 1975-77. 2/ However, even after these changes, petroleum prices remained well below other Latin American net importers of energy. With the completion of the pipeline in 1977, the flow of the jungle oil is reducing the short-term balance of payments pressure to adjust prices, but failure to raise prices would have serious negative consequences in the long run. 26. The second policy issue is how to achieve the extremely high level of investment required to develop hydroelectricity and explore for cil. The crucial aspects are financing, and management and technical capacity. The financial aspects depend on the adequacy of pricing policies to generate investable resources in the public sector and on the role of the foreign companies in petroleum explorations. The amount of resources and the risk involved in petroleum explorations in the jungle are so high, that it would be impossible for PETROPERU to carry all the burden. 3/ Concerning the managerial and technical aspects, development of the electricity sector is hampered by the relative inexperience of ELECTROPERU 4/ and the lack of a coordinated development strategy for the sector. 5/ Although PETROPERU has more adequate experience, it is probably already operating at peak capacity. 1/ See Chapter VI. In the 1972-76 period almost 45% of the total project- related foreign loan disbursements went to finance practically all of PETROPERU's investment expenditures. 21 In soles, the price of gasoline increased eight-fold in the May 1975- November 1977 period (from S|.9 to S/1.75 per gallon). However, in dollar terms (at current exchange rates) the price change was from $0.23 per gallon in May 1975 to $0.70 per gallon in November 1977. 3/ PETROPERU is at a disadvantage in petroleum explorations since it does not have alternative areas to explore and compensate risks. An interna- tional oil company minimizes overall risks by exploring in various areas in the world simultaneously. 41 ELECTROPERU was created in 1971 but has not yet implemented any large- scale project. 5/ Up to the creation of ELECTROPERU, 22 private and 3 state-owned corpora- tions accounted for 50% of the installed capacity. In addition., there were around 660 firms generating electricity for their own use (mainly mining companies) that accounted for 48% of the installed capac:Lty. The remaining 2% was generated by over 420 small state- or municipal-owned plants. Table 5: TREDS OF PRODUCTION, CONSU)nPTION ANID ITS IffURTb OI IllERGY BY TYPE (IF FUEL, 1950-74 1/ A. CONSUMPTION B. PRODUCTION a- , Crude 2 Crude met -57- Year Oil Coal Gas Electr.- Total Oil Coal Gas Electr.- Total Imports A. Ia 000s of barre]Ls per day of oil equivalent 1950 23.4 1.9 T.2 - 32.4 42,5 2.6 7.2 - 52.3 -19-9 1964 68.6 .9 8.0 4.0 T 64.3 2.0 8.0 4.0 1974 110.6 4.1 9.4 9.2 133.2 74.9 1.1 9.4 9.2 9g - - -3 1950-54 27.9 2.2 7.3 0.4 144.9 2.7 7.3 0.4 55-3 -17.6 196o0-6k 56.5 2.1 7.6 3.3 57 .T 0 2.1 7.6 3.4 701.0 1970-T74 100.0 3.7 8.9 7.7 120.2 68,2 1.4 8.9 7.5 2 B. Annual Averatge Grovth Rate (5) 1950-6c 7.3 0.5 0.3 - 63 2.1 -1.7 0.3 - 2.2 D.8. 1960-70 T.0 T.4 1.3 8.4 7 3.1 -0.5 1.3 S.4 3.1 0.& 1970-T4 4.3 - - 9.9 rT 1.4 -14.9 2.9 9.9 3/ 1.9 12.2 1950-7T4 6.7 3.3 1.1 8.2 T 2.4 -3.5 1.1 8.2- 2.5 n.8 C. Percentage Struc Lure (S) 1950-54 73.9 5.9 19.3 0.9 100.0 81.2 4.9 13.2 0.7 100.0 -46.7T 1960-64 81.3 3.0 10.9 4.7 100.0 8l.4 2.9 10.8 4.9 1oo.o - 8.6 6 19T0-7T4 83.2 3.1 7.3 6.14 10.O0 79.3 1.6 10.3 8.8 too.o 28.5 - y 811 SeLs converted into a common unit tbousanis of erude oil equialent per day 2/ Primary alectricity 1/1954-19T4 Percent of total consumption Source Appadix Table 11.1. Tab-e 6: PFM: INDICAT IE ESTIMAT OF EnEGY POTETIAL AND REWIREMENS, 1974-2000 (in barrels of oil equivalent) Accumulated Potential 1975-2000 * energy 1974 1985 9ggo 2000 (million)2- millions)! A. Enerxy prodtiona -(n/ Tota 946 242 4 814.1 184.. 1530 (excluding oil con- tractors sbare)lJ (94.6) (184.9) (84.1 (184.7) (1323) Petroleum 74.9 195.0 11.0 - 830 830 Gas 9.4 16.1 20.5 33.4 184 234 Bydro 9.2 26.2 42.3 log.6 403 4 1803 Present cap city (9.2) (9.2) (9.2) (9.2) () - (T) Untapped potential( - ) (17.0) (33.1) (100.4) (316) ' (1716) / CoaL 1.1 5.1 10.3 41.7 113 511 IO/ B. Energy reow rements Alternative yearly growth ratea: 6.0% V 133.2 252.9 338.4 606.o 3049 6.5s 8/ 133.2 266.3 364.8 684.8 3299 T.0o 133.2 280.4 393.2 773.5 3573 7.5% 2/ 133.2 295.1 423.7 873.2 3=2 1/ Grovth rates assumed were msxium reeliseticaly chievable without exhausting the econeoically reoverable proven reserves. 'Tbe yearly grovth rates assumed vere: coal 15%; gs 5% and electricity 10%. In the oao. of electricity. iC% is conidered the maxium rate at which the construction of hydlroelectric potential can expAnd. (This implie; over 60000 Gvh of power produced in 2000; devand projections indicate a need for no more than 37500 wtv in that year). The projectiont for coal at 15% p.a. is optimistic since most of the coal is anthracitic and past grotb has been negative. The petrole.m pro- jections are taken f Tasble 4. 2/ Sun of energy production and requiresents ov,r 26 years. I/ Total economically Proven energy reserves in 1976 4/ Present daily capacity x 365 days r 26 years 5/ Sum qver 26 years of daily production assuming that total installed capacity grovs by 10% a year. i/ 26 ti_m the production of maimu installahj.e capacity. (About 109000 Owh per year or 66 jillion barrels of oil equivalent) using nil bydro resources in Peru. Corresponds to the 1950-74 annual average grEnuth rate. o/ Corresponds to the historical elasticity of .mnergy consumption to GMP ( 1.2) assuming a DP Mgrowth of 5.5% p.a. 2/ Assuses GDP grovth of 5.5-6.0% P.-Aand an elsaticity of energr consmption to GDP of 1.3. 10/ 511 million are the proven reserves that are economically recoverable. In addition there are about 500 million of proven reserves, but not economically recoverable at today' s prices ad an estimated 33 billion of probable reserves. llj Equivalent to around 1/4 of petroleum reserves. See Table 4. Solrce: Mission estimates. - 58 - C. CONCLUSIONS 27. This chapter showed that Peru's varied natural resource base holds both major opportunities and major constraints to the country's future devel- opment. The largest asset is mining, which will remain the backbone of the economy as a source of savings and foreign exchange, although it provides little direct employment and may have, for a long time, few linkages with the rest of the economy. Fishing resources appear to be abundant and could provide foreign exchange, protein-rich food for domestic consumption, and some employment. However, their long-term growth potential is still uncertain and they fluctuate sharply from year to year. The largest constraint is the scarcity and poor quality of agricultural lands. The population pressure on the lands now in cultivation is high (particularly in the Sierra) and (aside from some valleys in the Costa) yields are low. Since the soils in :he Low Selva plains appear to have very limited long-term agricultural potential, the only new lands that could be brought into production are limited areas in the Ceja de Selva and, with expensive irrigation, some valleys in the Costa. Much can be done to increase land productivity with known technologies, but even then, the long-term average growth of food production may not be above 2-3 percent yearly (barring unlikely major technological breakthroughs). The second major potential constraint is the lack of energy resources. Even taking into account the hydroelectric potential and the proven coal reserves, Peru's dependence on imported energy resources will probably increase in the future -- unless a major breakthrough is made in the petroleum explorations in the Selva and offshore. The third potential natural resource problem is water. Although the country as a whole seems to have large water resources, satisfying the future long-term needs west of the Andes will require high investments in order to transfer water from the eastern side of the mountains. Finally, to increase the agricultural potential of lands in the Costa, better irrigation and drainage systems will be required to reduce the sharp seasonal fluctuations of water flows of coastal rivers. - 59 - CHAPTER II: POPULATION: GROWT]I AND SPATIAL DISTRIBUTION A. DEMOGRAPHIC TRENDS: CHARACTERISTICS AND PROJECTIONS 1/. a. Brief Historic Overview, 1530-1940 28. Demographic trends in Peru can be divided into three phases: rapid depopulation between 1530 and 1750; moderate growth up to the end of the 19th Century; and accelerated growth during this century. A major depopulation took place between 1530 and 1750, but there is considerable uncertainty on the number 2/ and the specific causes. 3/ .able 7: 30rYLA=01 =-==. L53;0.9i 1530 .5qI L?3 .3 676 19 .o~a1. 3.5-5.3 . a 2.T 6.T Some.: S. ht, at. at. ;.. 2. > C La ?W.aaAda Pnrd. wp.2. an I7. T/ he glossary below defines demographic technical terms that are used in this Chapter: - Crude birth rate: number of live birth:; per 1,000 inhabitants - Crude death rate: number of deaths per 1,000 inhabitanits s - Rate of natural increase: the crude bi.-th rate minus the crude death rate - Total fertility rate: average number o: children per woman at the end of her reproductive age - Gross reproduction rate: average number of daughters per woman that will survive intil reproductive age given the present levels of mortality - Net reproduction rate: average number Of daughters per woman that vill survive until reproductive age given the present levels of mortality - Replacement level fertility: level of fertility when the net reproduction rate equals 1 - Child to woman ratio: children under 5 years of age per women between 15 and 49 years of age - Child dependency ratio: children under 15 years as a percent of the population in the 15-64 age group - Age dependency ratio: population in the over 65 age as a percent of those in the 15-64 age group - Total age dependency ratio: the child dependency ratio plus the age dependency ratio -,Sex xatio: number of males per hundred females. 21 One of the more conservative estimates, by S. Hunt, is that population fell from 3.9-5.3 million at the beginning of the conquest to 0.8 million in 1750.(See S. Hunt, Growth and Guano in the Nineteenth Century Peru, Discussion Paper No. 34, Research Erogram in Economic Development, Woodrow Wilson School, Princeton University, Princeton, New Jersey, February 1973, pp. 11-15). / S. Hunt quotes various authors as generally faulting "civil wars, epidemics, lawlessness, extortion and cultural shock". (See Hunt, op. cit., p. 26.) - 6o - 29. The estimates in Table 7 above show that population grew annually by roughly 1% in the 1750-1876 period and by 1.4-1.5% between 1876 and 1940, mainly because of changes in the mortality rate. 1/ During the first part of this century, with the beginnings of government efforts to expand health services and eradicate epidemic diseases, the death rate fell from around 32.5 per thousand to 27.1 per thousand in 1940. 2/ The estimated life expec- tancy at birth rose from 30 years in 1876 to almost 36 years in 1940. b. Population Growth. 1940-72 3/ i. Overview 30. Between 1940 and 1972, the annual population growth rate continually accelerated from around 1.8 to 2.9 percent and total population more than doubled (Table 8). Since the early 1960s, birth rates seem to be gradually declining, but not yet enough to offset the reduction in death rates, and to reduce the rate of natural increase. Population growth in Peru is now one of the fastest in Latin America. 31. With the acceleration in population growth, the median age has been falling (to an estimated 17.6 years in 1970) and the age dependency ratio has risen rapidly (Table 10). 4/ By now Peru has one of the highest child depen- dency ratios in Latin America (Table 9). ii. Components of Growth: Birth and Fertility Rates 32. Birth rate trends have been reflected in changes in fertility rates in practically all age groups increasing up to the mid-sixties and falling since 1/ Only in the beginning of this century did population resume the estimated levels of the precolonial period. 2/ Estimates of birth rates for 1876 and 1940 are unreliable. The National Office for Statistics and Census (ONEC) gives two different est:Lmates for 1876 in different publications: 44.5 and around 43.5; the :Latter estimate seems more plausible giving a natural rate of population growth of around 1.1% per annum. For 1940 the estimates for birth rates range between 45.1 and 45.4 per thousand. At the lower rate, the natural rate of population growth that year would have been around 1.8Z. 3/ Graph 1 illustrates the main changes in demographic trends since 1876. 4/ In the 1940-61 period, the group under 5 years of age rose the fastest because the fall in death rates and the increase in birth rates were predominantly determined by the lowering of infant mortality and better pre-natal medical care. Since then, the growth rate of that age group has declined somewhat (with declining birth rates), while the 5-14 age group has been increasing faster than total population (3.3% vs. 2.9% p.a.). - 61 - Table 8: POPULATION 2ROWT% AnD DISTR3IBTIOlI, 124072 Average Grovth Ye-ar to Percentage P ( dojble ln miflions Stnjcture , b190 I961J 1940 196i 1940 196 1972 1940 1f6l 1972 61 72 6l T2 Total o 7 10.3 14.l 100 100 100 2.1 2.9 34 24 Urban 1.7 4.0 7.2 27 40 53 4.2 5.6 17 13 (Li=) (0.5) (1.8) (3.3) (8) (19) (24) (6.2) (5.5) 12 13 (Other urban) (1.2) (2.2) (3.9) (19) (21) (29) (3.0) (5.7) (23) (13) Rura 4.5 5.9 6.3 73 60 47 1.3 O.6 54 116 cost& 1.7 3.8 6.2 28 38 46 3.8 4.7 19 15 Sierra 4.0 5.3 6.0 65 53 44 1.3 1.2 54 58 Solve 0.4 0.9 1.3 7 9 10 3.5 3.8 20 19 cansus -.f omission. 0.5 '0.4 O.6 - - - - - - - m mieludes cesus mission t/ imted census error Sorce: ONtEC Table 9: COOMPARI0S3 C! DENGRAPIC INDICATORS BErWEZN PI AID OTIER LATIN AlEiICAI CDUNITINE WITM BIJULIR PER CAPITA IN1COME 1/ (1975 or average 1970 - 75) Birth Death Rate of Life expee- Popul. 1975 Rate Rate atural tancy at under Country Popal. oIoo */O Growth Birth 15 unit mill. numbers nuser S S year % Latin Amrica- 290.3 38 9 2.9 62 42 Peru 15.6 44 14, 3.0 £ 44 Brazil 107.2 37 9 2.8 61 42 Chile 10.6 28 . 1.9 63 39 Costa Rica 2.0 28 2.3 69 42 )Amico 60.2 46 8 3.8 -63 46 Nicaragua 2.2 148 i4 3.4 53 148 Panam 1.7 31 'i 2.6 66 143 Urugay 2.8 21 10 1.1 70 28 Venezuela 12.0 36 i 2.9 65 44 Rink of Peru3/ 3 3 3 8 3 C/ Countriez of mainland Latin Americ:a vhich in 1970 had GNP per capita between US$ 375-1000, calculated with IBll) World Atlas methodology. i Mainland only. 3/ Rank of Peru vis-a-vis other couitries with a higher ralue for indicator. Source IBED World Atlas 1976; U3AID, Family Planning Service Statintics Annual Report. 1975; and Appendix Tables 1.1, 1.8 and 1.10. - 62 - Tablel0: AGE AND SE STRUCTURE OF POPULATION, 1876-19T2 (in percentages) Total Male Female 1 9 7 2 ___________ __________Urban Rural 1876 1940 1961 1972 1876 1°140 1°61 1972 1876 19410 1961 19T2 UT ra F T m A. Age Structure by Sex Totn. 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 0-4 14.2 15.4 16.9 16.3 14.5 15.8 17.1 16.3 13.9 15.0 16.7 16.2 15.1 15.3 15.0 17.9 17.9 5-14 22.7 26.6 26.5 27.6 22.9 *27.7 27.2 28.1 22.7 25.5 25.8 27.1 26.7 27.0 26.4 29.0 29.8 15-49 51.6 45.3 45.4 45.0 51.4 46.0 45.3 45.0 51.6 46.7 45.5 45.3 47.7 47.7 47.7 40.9 4O.7 50-64 8.8 8.3 7.3 7.1 8.6 6.9 7.0 7.1 9.0 7.7 7.6 7.2 6.8 6.7 6.9 7.6 7.5 65 and over 2.7 4.4 3.9 4.0 2.6 3.6 3.4 3.5 2.8 5.1 4.4 4.2 3.5 3.1 3.8 4.4 4.1 B. Sex Ratios by Age Total C00.8 9T.6 98.8 100.4 100.0 101.2 0.4 105.8 102.6 101.2 101.2 101.2 101.5 5-14 102.7 106.2 103.8 104.2 102.2- 107.1 15-49 100.0 96.2 98.3 99.8 100.2 99.0 50-64 95.6 87.2 92.1 98.5 97T.4 100.0 65 and over 92.9 89.2 77.3 87.9 81.3 87.0 C. Other Age Dependency ratio s : Total 65.6 86.6 89.8 91.J 83.0 105.8 Child -/ 61.1 78.4 82.4 84.2 76.6 96.7 Aged - 4.5 8.2 7.4 7.4 6.4 9.12 7 of males 15-64 60.0 52.9 52.3 52.1 54.4 48.2 Median age n.a. 19.6 18.9 18.5 n.a. 18.8 18.4 18.2 n.a. 20.4 19.3 1B.8 S See foot.note 1 on pagc 25 for derinitionz. Source: Appendix Tablec 1.5 and 1.6 - 63 - then (Table 11), although the age pattern of fertility has not varied greatly over time. 1/ Nonetheless, the total fertility rate remains high (at around 6.5 in 1970) and the net reproduction rate is well over twice as large as replacement level fertility. ;/ - 33. It is difficult to point out with any precision the causes of the changes in birth rates. The general hypothesis is that with better socio- economic conditions that reduce infant and maternal mortality, fertility rates will tend to increase at the beginning. Then, with further improvements in socioeconomic conditions, more education and a higher female labor parti- cipation rate, fertility will tend to irop. This .as been the general pattern found in other countries and the data from Peru support it. Fertility rates are higher in the Costa than in the poorer Sierra 3/ but decline sharply with city size. 4/ Thus, the shift of population from the Sierra to the Costa, 1/ The trends followed during the last century by the child to woman ratio is another indicator confirming the changes in fertility rates. The estimated child to woman ratio for 1876-1972 was: 1876 19A0 1.96± - 1972 1876 1940 .1961 Total (kban Rural 550 653 T38 T22 634 8T2 Source: Appendix Table 1.5 2/ See footnote 1 to para. 28 for definition of child to woman ratio, total fertility rate, net reproduction rate and replacement level fertility. 3/ The table below shows indicators of fertility differentials between regions and rural and urban areas, according to the 1961 Census: OthLer Paru Lima Costa Sierra Selv Child/Women Ratio: Total 735 614 T70 T55 927 }1e.i 816 u&8 3838 780 1032 TotaL Fertility Rate Total 6.a3 &.6i 6.'4 6.1 ?.i8 Orben S . 35 t.3 5. T8 6.-90 Rura.l 6.T0 6.62 6.86 6.31 7.56 Source: ONEC 4/ The more relevant concepts are probably level of income and access to education and other services, ral:her than city size. Various surveys in 1965 and 1971 by ONEC and the Miniistry of Labor estimated the gross reproduction rates to be 2.05 for Lima, 2.61 for Arequipa, 3.49 for Iquitos and 3.76 for Cerro de Pasco. The comparable national average for 1960-65 is estimated at 3.35. - 64- Tble 11: BM RATES AND AGE SPJCFIC FERILITr RATES, 1940-2000 retiimted Projected 1960 1965 1970 1975 1985- 1995 1940 1961 -65 -7C 1970 -75 -80 -90 2000 15-19 0.085 0.101 0.115 o.111 0.108 D.106 0.100 0.086 0.074 20-24 0.281 0.298 0.313 0.302 0.296 0.290 0.278 0.256 0.233 25-29 0.292 0.314 0.338 0.326 o.32 0.314 0.302 0.278 0.253 30-34 0.246 0.255 0.275 o.264 0.259 0.253 0.240 0.212 0.185 35-39 0.179 0.193 0.207 0.202 0.202 0.198 0.184 0.154- 0.126 40-41 0.090 0.087 0.095 0.088 0.084 0.082 0.076 0.o66 0.054 45-49 0.027 0.028 0.030 O.026 0.023 0.022 0.020 0.014 0.010 Total fer- tility r&t/n.a. n.&. 6.86 6.60 6.46 6.33 6.00 5.33 4.68 Crude birtb rate (o.oo) 45.0 9.6.5 47.09 45.04 44.5 43.56 42.14 39.o6 35.42 0-mas Re- I/ production n.a. n.e. 3.35 3.22 3.08 2.93 2.60 2.28 rate let repro- / duction rate n.&. n.& 2.38 2.38 2.40 2.36 2.21 2.02 / See footnote 1 on pe 28 for definitions. Soure.: Appendix Table 1.10 and 1.11 Table 12: AV3E NUMB 0 CHrLDRE B0R2 ALIVE PER WOAW AT THE MD 0? THE 3R OD3CTO PERIOD. BY EDUCATIONAL EL,. 1972 Number of Pereentage distribution of women ducation Level children in the group aged: 45-h9 Yeears 25-29 years Total ia100.0 100.0 No education 7.32 51.5 30.5 Incomplete PriMa 7.00 23.7 30.0 Conplete priary 5.33 12.8 15.0 IncomPlete secondary 3.94 4.5 8.9 Couplete secondary 3.42 4.T 8.0 University education 3.03 2.2 6.3 Otber bigher education 2.40 - 0.1 Not disclosed 4.69 0.8 1.4 1/ Ed of reprodtctive period, 45-49 years old. Source: Appendix Table 1.13 - 65 - where socioeconomic conditions were rapidly improving, caused a rise of the country's birth rate up to the mid-sixties. However, as the importance of the Sierra population in the total declined and the majority was-living in rapidly growing cities, the country's aggregate birth rate started to decline. If this change in behavioral patterns continues, fertility rates can be expected to continue falling sharply in the large cities, start to decline sometime in the future in other areas of the Costa but continue rising in the Sierra. The data also suggest that two important factors affecting the changes in fertility rates are education and female labor participation. Table 12 shows how, according to the 1972 Census, fertility varies inversely with education and how the better education profile of the younger groups will continue to contribute to a fall in fertility. The 1961 Census showed that in both urban and rural areas, fertility rates were substantially lower for women who were economically active. iii. Components of Growth: ]Hortality Rates 34. The rapid fall in mortality rates during this century has been the principal cause for the acceleration of population growth. (See Graph 1 and Table 13). It took over 60 years to reduce the crude death rate per thousand from 33 in 1976 to 27 in 1940. In the next 30 years the death rate was cut in half to an estimated 14 per thousand for 1970-75. 35. One of the important factors explaining the fall in death rates has been the reduction of infant mortality. This is suggested in Table 13 by the much sharper increase since 1876 of life expectancy at birth than at age 5. Various studies have also found that death rates vary considerably with city size and with degree of urbanization of regions. 1/ This may suggest that the fall in the national average death rate has been caused by both a redistribu- tion of population to the larger urban centers and by the fall of death rates in the individual regions; the relative intensity of the two factors is however difficult to estimate. 2/ Mortality rates can be expected to fall further in the future; life expectancy at birth (55 years in 1970-75) is still below the Latin American average of 62 years (Table 9). Much progress has been made in reducing contagious diseases (like tuberculosis) while the incidence of illnesses related with aging (such as cardiovascular diseases and tumors) has increased. However, infectious parasitic and respiratory diseases remain the principal causes of registered deaths, particularly for the under 1/ A recent study estimated the crude death rate per thousand in the 1969-72 period at: 18.7 for the whole country; 10 for Lima; 15.2 for cities with a population of 2,500 and over; and 23.9 for cities with a population of less than 2,500. (See Bagan, C. "Fecundidad, Mortalidad y Crecimiento", CEPD, unpublished, quoted by the MIinistry of Labor in "Informe Sobre la Situacion Ocupacional del Peru 1972".) 2/ The Informe Sobre la Situacion Ocupacional del Peru 1972 found a close inverse relation (correlation coeEficient = -0.80) between death rates in 1961 and degree of urbanizatio' of the departments in Peru. - 66 - 5 years of age group. 1/ Children less than one-year-old still accounted for 28% of registered deaths in 1970; their death rate was estimated in 1965 at 100 per thousand born alive. Table 13: PERU: LIPE EXPECTANCY, 1876 - 2000 If e E x p e c t a n c y Gross Death Both sexes Pe:centage difference Rates at: Males at: Females at: betveen: o0(0 Birth Age S Birth Age 5 Birth Age 5 Females ard Birth and males at: age S for: Birtn Age 5 Birth Age 5 Census years 1876 32.5 o.a. n.a. 30.08 43.55 30.00 43.36 -0.3 -0.4 44.8 44.5 1940 27.1 n.a. n.e. 34.62 44.29 36.77 47.44 6.2 T.1 27.9 29.0 1g61 15.4 n.a. n.a. 49.56 55.03 52.50 57.23 5.9 4.o 11.0 9.0 5-vear oeriods 1960-65 17.8 49.38 58.27 48.27 57.67 50.54 58.82 4.7 2.0 19.5 16.4 1965-70 15.7 51.95 59.54 50.75 58.93 53.21 60.22 4.8 2.2 16.1 13.2 1970-75 13.8 54.55 o.a. 53.25 59.81 55.91 61.17 5.0 2.3 12.3 9.4 1975-80 2.a.3 56.53 n.a. 55.15 60.56 57.98 61.93 5 .1 2.3 S . 6.o 1985-90 10.2 59.60 n.a. 58.05 61.77 61.23 63.38 5.5 2.6 6.4 3.5 1995-2000 o.5 62.46 m.a. 60.71 62.96 64.30 64.84 5.9 3.0 3.7 C.8 So=ze: Appendix Table 1.17 c. Magnitudes and Consequences of Future Population Growth. 35. Peru has been going through important changes in demographic trends in the last 30 years. The rapid fall in death rates will probably continue, since they still are relatively high. On the other hand, as has been occurring in many Latin American countries, birth rates seem to have passed their 1/ According to Ministry of Health data (quoted by ONEC), registered deaths had the following percentage structure of causes: 1950 1960 '0 76 Total ubercuaosis 18.2 9.3 6.6 Fneumoli4a and bronchitis 14.3 15.0 21.3 Digestive diseases 8.3 10.5 14.0 Camdi-msel'l.r a.8 6.8 10.9 T=ars 5. . ,7 7.TI Accidents 4.4 5.5 ii Other L0: 5.2 -3.0 - 67 - 5rsph Xfo. 1 Pl-RU RATE Of NATURAL INCREASE OF POPULATION, 1876 - ;970 AND ALTERNATIVE PROJECTIONS 1970 - 2000 FATES BIRTM RATE . .... ..... ALTER. If .346.11 49 -- -1* '~ ALTER. III .31.S)_I RATE OF NATUniAL E i INt I mmY . t __ Z 3 t ) \ ~~~~~~~~~~~~ALTER. I i3! 30 _ _I_ _ b_ C _ _-. _-L ALTER. IV 121 J)II EATH, RATE _ ~ ~~~~ -E -lXA - - I~~MATO2 - --.-FLcTI oNs- O1 __ ___ l l_________ 1870 1880 1880 1900 1910 1920 1930 1940 980 1960 1970 1980 18M0 200 31 Jw Tabl. 14 rnd Wer tor -I.,anan. ol .1 ate-rierIontanon.. Nnbmu in 111080 ICI 18tt to1 nd* r 0u ot e~ f18d inuprl8l81 lo 1118 1891 _ 2000 nmiodt unoer 1the diiferent aItnr..i,n. 12 In th. 1970 - 2000 p,.ro1oon Polied. eth. tittded to th. m. of elu.MI inm,e.I n,dw Allerniie I. So-.e: 01m N.o-d. tE euwrsr. y Cwo- IONECI nd APPndi. T.bh 1.7 ';dd &nhe-17133 inflection point during last decade. I/ The important question is whether, without direct Government intervention, the fall in fertility rates will be rapid enough to significantly slow downL population growth in the next few years. 36. Table 14 and Graph 1 summarize the key variables associated with the most recent population projections proposed by ONEC for the 1975-2000 period. 2/ Concerning mortality rates, ONEC assumes that average life expectancy in Peru will increase from 55 years in 1970-75 to 63 years in 1/ A recent sfudy found that since 1960-65, 21 out of 25 Latin American and Caribbean countries have shown a decline in birth rates, 2 have had small increases (less than I poini:) and 2 had no change. See F.W. Oechsli and D. Kirk, "Modernization and the Demographic Transition in Latin America and the Caribbean," "Economic Development and Cultural Change," April 1975. 2/ See ONEC, "Perspectivas de CrecimLento de la Poblacion del Peru, 1960-2000," Boletin de Analisis Demografico No. 16, December, 1975, pp. 35-119. - 68 - 2000, similar to the level attained by Lima in 1968. 1/ These range between a high extreme (Alternative II), which assumes no change in global fertility rates from their level in 1970-75, and a low extreme which assumes a 41% drop by 1995-2000 (Alternative IV). 2/ 38. The most interesting comparison is between Alternative I and Alternative IV. If the decline in fertility rates continues at the same slow pace as in recent years (Alternative I), the growth rate of population will decline substantially only after 1990. 3/ Over the 1976-2000 period, populatioi will remain growing at 2.9% annually and double in size to 31.9 million. 4/ With a faster fall in fertility rates (Alternative.IV), the average annual growth rate would start declining rapidly in 1980 instead of 1990; over the 1976-2000 period the average annual growth of population would be. a lower 2.6% and by the year 2000 total population (29.7 million) would be 2.2 million smaller than in Alternative I. 5/ However, the targets on fertility decline of Alternative IV would be difficult to achieve without a deliberate population policy by the Government. Moreover, taking a longer term perspective shows that even Alternative IV implies a high growth of population. If its trends on changes in fertility rates are extrapolated further in the future, 1/ Since future increases in the average life expectancies will mainly come from falls in mortality in rural areas, which are more difficult to achieve, the projections imply a gradual slowdown in the fall in death rates, but a 50% drop in infant mortality rates. 2/ See Table 14 and Graph 1. The total fertility rate was estimated in 1970-75 at 6.32 (and 4.1 for Lima in 1974). Alternative II that assumes no fall in fertility rates, may be unduly pessimistic, since they have already started falling. On the other extreme, Alternative IV assumes an acceleration of the drop in total fertility rates to 3.75 by 1.995-2000, which may be unrealistically optimistic without a deliberate population policy. Alternative III assumes a slowdown in the fall of fertility rates. ONEC recommends as the most probable one of the intermediary alternatives (Alternative I) that assumes a 26% fall in the total fertil- ity rates in the next 25 years -- broadly in line with what haLs been occurring in the last decade, but above the 4.1 attained by Lima in 1974. 3/ The long lag is caused mainly by the very young age structure of the present population. See Appendix Table 1.9 for projections of age structure of the population under Alternative I. Table 1.9 shows that growth rate of females in reproductive ages (15-49 age group) will be 3.2% a year in the 1975-2000 period. Thus, the fall in fertility rates will be offset by a larger number of women in fertile ages. (See Table 11 for projections of age specific fertility rates under the Alternative I). 4/ The growth rate of population which is estimated at 3.0% in 1975 would still be 2.9% in 1990 but gradually fall to 2.5% by 2000. 5/ 2.2 million is equivalent to over two-thirds of the present pcpulation of Lima. In Alternative IV population would be growing by 1.9% annually in the year 2000 (compared to 2.5% in Alternative I). - 69 - Alternative IV implies that replacement :Level fertility (approximately two children per woman) would be reached around 2020. At this rate, population would become stationary (no growth or decline) at 65 million in the year 2075. However, if replacement level fertility was reached by the year 2000 instead of 2020, population would become stationary at 43 million. 1/ This illustrates that a few years' delay in fertility decline will make a large difference in the ultimate stationary size of the population. Table 14: MRU: ALTERi4ATIVE POPUIATlON PROJECTIONiS ASStlitliG DTFFFREltT FrTILTTY RATES, 1970-2000 1/ L970-T5 1975-80 1985-90 1995-2000 Alternative I - Moderate fall in fertilitr rates. Total fertility rate 6.33 6.o0 5.33 4.68 Crude birth rate (0/00) 43.56 42.14 39.06 35.42 Crude death rate (0/ao) 13.73 12.27 10.18 8.53 Natural growth rate (o/oo) 29.84 29.87 28.88 26.90 Population (in millions) 2/ 15615 18135 23615 31938 Average growth rate 4/ 3.0 3.0 2.9 2.5 143.56 Age dependency ratio 3/ 93.19 91.49 87.70 82.92 Alternative II - la farl in fertility rates Total fertility rate 6.33 6.33 6.33 6.33 Crude birth rate Wa/oo) 44.11 44.21 43.141 Crad" natural growth rate (o/oo) 29.84 31.61 33.63 34.59 Population (in millions) 2/ 15615 18295 25530 36023 Average growth rate 41 3.0 3.3 3.5 3.8 Age dependency ratio 3/ 93.19 91.49 92.55 96.24 Alternative III - Slov fall in fertility rates Total fertility rate 6.33 6.19 5.91 5.62 Crude birth rate (W/oo) 43.56 43.25 42.12 40. L5 Natural growth rate a/aoo) 29.84 30.86 31.70 31.45 PopuLation (in miLlions) 2/ 15615 18226 25012 34309 Average growth rate 4/ 3.0 3.2 3.2 3.3 Age dependency ratio ]f 93.19 91.49 90.49 90.70 Alternative IV - Rapid fall in fertility rates Total fertility rate 6.33 5.82 4.79 3.75 Crude birth rate (a/aol 43.56 40.9T 35.9T 30.15 Natural growth rate (oloo) 29.84 28.84 26.03 21.82 Population (ia millions) 2/ 15615 18042 23591 29685 Average growth rate 4/ 3.0 2.9 2.6 1.9 Age dependency ratio 3/ 93.19 91.49 84.86 7T.19 1. Prepared in October 1975 by ONEC. All Alternatives assume the sere chsnCe in death rates (shown in Alternative 1). 3/ Refers to last year of five-year period. 3/ Refers to initial year of five-year period. 1/ Orovth rate durinc the last ye.r ef t'! five-y-%r -rioA. Source: ONEC 1/ Projections beyond 2000 are from: Peru: Perspectivas por Paises, Population Council, New York, N.Y., 1975. - 70 - 39. The rate of population growth will have significant influence on critical parameters of Peru's socioeconomic development. A rapid population growth will: (a) magnify the constraints on future development imposed by the natural resource endowment that were discussed in Chapter I (food and energy gaps, population pressure in cultivable areas, and water shortages in the Costa); (b) hamper the efforts to raise incomes in rural areas and increase the pressure on job creation and on demand for infrastructure in urban areas. A sustained increase of per capita income in rural areas may be achievable only at zero population growth. As confirmed by recent trends on rural-urban migration, a fast population growth will be mainly reflected in the urban areas. There is a 15-25-year lag before changes in population expansion have an impact on the growth of wording age population. Thus, because of the quickening in population expansion of the last decades, the growth of the age groups in working ages will inevitably accelerate between now and 1990. Under Alternative I, the pressure of population growth in propelling the expansion of the labor force will fall very slowly and only after the year 2000. In these circumstances, it will be an exceedingly difficult task for the Government to avoid an increase in the already high underemployment rates. Moreover, the rate of growth of cities implicit in the above population projections may outstrip the financial and organi- zational capacity to expand urban infrastructure; and (c) increase requirements for public services. Peru already dedicates a high 5% of GDP to education. Yet, one-third of the population is still illiterate 1/ and the quality of the formal education system leaves much to be desired. 2/ Increasing coverage and improving quality will be feasible only if the growth rate of children in schooling ages drastically declines, since it is doubtful that a higher share of the country's resources may be allocated to education. Yet, under Alternative I children in the 6-14 age group will almost double in the 1976-2000 period. 3/ Population growth will have similar impacts on health and other Government services. 1/ See Appendix Table 1.20. 2/ For an extensive criticism of the educational system, see Ministerio de la Educacion, Reforma de la Educacion Peruana: Informe General. 1970. 3/ From 3.7 million in 1975 to 7.3 million in 2000 (Appendix Table 1.8). - 71. - 40. The principal conclusions from the discussion in the last few paragraphs, are that Peru's development prospects will be affected by the size and growth rate of population and that if present trends-in-changes of fertil- ity rates continue, population will keep growing very rapidly. Attaining a lower demographic growth rate will require that a specific population policy be more fully integrated into Peru's development strategy. Raising socio- economic conditions of the poor--and more particularly of women--through better jobs, more education and more health services are important lines of action; as was discussed before all these factors combined with a growing urbanization, are important to change behavior of families towards lower fertility levels. However, to accelerate the fall in fertility to at least the levels of Alternative IV, the effcrts for socioeconomic improvement should be complemented by family planning programs. The effectiveness of these pro- grams should be enhanced by the fact that fertility rates have passed their inflection point. Recent studies have found that a decline in fertility is easier to accelerate than to initiate with family planning programs. Accord- ing to the Berelson Report, these programs should be successful in lowering fertility even in depressed rural areas. 1/ 41. In 1976, the Government expressed its awareness of the importance of population policy but little concrete action had been taken by early 1979. Up till now, family planning programs in Peru have been limited. 2/ In 1964, the Center for Studies of Population aLnd Development (CEPD) was established to research population-related issues ande promote family planning programs with support from the Ministry of Health; the Government, however, stopped its support in 1969. In 1967, the Peruvian Association for Family Planning (APPF) was founded and opened clinics throughout the country; in 1975 the Government ordered APPF to cease all operations. 3/ Up to now, the most significant positive Government action has been the creation in 1972 of the National Institute of Neonatology and Maternal-Child Protection (INPROMI). In 1974, INPROMI started a responsible parenthcood program restricted, however, to mothers whose health would be endangered with more births. 4/ The program, 1/ The Berelson report argues that there is "... no convincing basis for asserting that a (family planning) program would inevitably be a failure in a poorly educated rural populaLtion..." Indonesia is an example of a successful family planning program in a poor rural country. 2/ The past reluctance of the Government towards family planning was a reflection of its belief that genteral economic and social development would eventually take care of the nation's population problem. This attitude was presented by the Peruvian Government at the 1974 World Population Conference. 3/ From 1969 through 1973, APPF's clinics (two of them in rural areas) provided a total of 113,000 consultations for planned parenthood. In addition, APPF placed much emphasis on information, activities, training and research. 4/ This responsible parenthood program is being undertaken largely in the interest of maternal and child welfare. - 72 - which at completion is expected to reach 3% of fertile age women, had enrolled nearly 33,000 active users of birth control methods by the end of the third quarter of 1975. 1/ B. LABOR FORCE TRENDS AND PROJECTIONS a. Past Trends and Characteristics 2/ 42. Table 15 summarizes the most important features of the changes of the economically active population (EAP) in the recent decades according to the census. However, only tentative conclusions on the overall trends can be ventured, because of the definitional differences between the three census. The growth of the EAP has been accelerating rapidly, 3/ but seems to have remained well below the growth of the population in the working ages, because of a fall in the total labor participation rate. 4/ The data suggest that the growth of the EAP has been more heavily influenced by changes in the age- specific labor participation rates (caused by non-demographic factors such as migration, expansion of education and social security), than by changes in demographic factors (such as changes in the age structure of the population). 5/ 1/ In 1975, nearly 12,000 new acceptors enrolled in the program. The most commonly used method in the program is the pill (44% of active users). 2/ The analysis in this section is based on the data from the three Census (1940, 1961 and 1972). There are, however, some problems of comparison because of differences of definition. The 1940 Census included as eco- nomically active only those that had a gainful employment at the moment of the Census; it therefore, included the unemployed as not economically active. In the 1961 and 1972 Census the persons that were unemployed (because they had lost their jobs or were looking for work for the first time) were included as economically active. However, in the 1961 Census the reference period was the day of the Census, while in 1972 it was the week of the Census. These definitional problems introduce biases in the data. The most obvious one is that it overestimates the growth rate of the labor force in the 1940-61 period and the changes in labor participation rates. 3/ In number of persons, the annual expansion rose from about 40,000 in 1961 to well over 76,000 in 1972 (Table 15). 4/ Labor participation rates (the population share in an age group that is economically active) for the population 15 years and older fell from 54.31% in 1961 to 49.81% in 1972 (Table 15). 5/ Table 16 shows that the variations in the age structure of the population that have occurred in the 1961-72 period (caused by the changes in the birth and death rates) explain a very small part of the changes in labor participation rates. Table 15; TERIGD OF ECONOMICALLYACTIVE POPULATION AND LABOR PARTICIPATION RATES, 1940-72 , In thousands for population and EAPr in % for labor percentage ratea Average Annual Growth Rhates Total Male Female Total Male Female 1940- 1961- 1940- 1961- 1940- 1961- 1940 1961 1972 1940 1961 1972 1940 1961 1972 1961 1972 -1961 1972 1961 1972 …--…---------------------------------------- A. Population ----------------------------- -… ----- Total 6 and older n.e. 7926.4 10909.0 n.a. 3929.0 5458.1 n.a. 3997.4 5450.9 n.a. 2.9 n.a. 3.0 n.a. 2.9 Total 15 and older 3600.6 5607.2 7601.0 1732.7 2242.4 3769.? 1867.9 2864.8 3831.3 2.1 2.8 9.2 2.9 2.1 2.7 ^total 15 - 64 3327.4 5236.7 7078.5 1621.0 2540.3 3531.3 1706.4 2696.4 3547.2 2.2 2.8 2.2 3.0 2.2 2.5 6- 4 Dn.m. 2319.2 3309.0 n.a. 1186.6 1688.4 n.a. 1132.6 1620.6 n.a. 3.3 n.a. 3.3 n.a 3.3 15-24 1117.4 1822.9 2569.9 558.7 921.3 1287.1 558.7 901.6 1282.8 2.4 3.2 2.4 3.1 2.3 3.3 25-49 1694.8 2674.7 3544.5 851.2 1307.7 1766.a 843.6 1367.0 1778.3 2.2 2.6 2.1 2.8 2.3 2.4 50 and older 788.4 1109.6 1485.6 322.8 513.4 716.4 465.6 .596.2 769.2 1.6 2.7 Z.2 3.1 1.2 2.3 Total 6 and older n.e. 7926.4 10909.0 n.a. 3929.0 5458.1 n.a. 3997.4 5450.9 n.a. 2.9 n.a. 3.0 n.a. 2.9. Urban n.a. 6602.5 n.a. 3295.7 n.a. 3307.2 n.a. n.a. n.&. .^Sirel n.a. 43o6.5 n.a. 2162.4 n.a. 2143.7 n.0. n.a. n.a. -------------------------------------------- B. Economically Active Population (LAP) … ------ - - Total 6 a.nd older n.a. 3124.6 3871.6 n.a. 2445.4 3071.4 n.a. 679.2 800.2 n.&. 2.0 n.a. 2.1 n.a. 1.5 ct-I 15' -n^d le-'Ar p 5 3o4s5i 3786.1 1912.6 2404.0 5023.0 8oo.9 640.7 763.1 1.3 2.0 2.2 2.1 -1.1 1.6 Total 15-64 n.a. 2896.2 3608.8 n.a. 2279.6 2871.2 n.a. 616.6 737.6 n.a. 2.0 n.a. 2.1 n.a. 1.6 6-14 n.a. 79.5 85.4 n.a. 41.0 45.4 n.a. 38.5 37.1 n.a. 0.7 n.a. .0.9 -.a. -0.3 15-24 n.a. 907.1 1016.0 n.a. 662.0 743.6 n.a. 245.1 272.4 n.a. 1.0 n.a. 1.1 n.a. 1.0 25-49 n.a. 1581.? 2084.8 a.a. 1286.6 1690.5 n.a. 295.1 394.3 n.a. 2.5 n.e. 2.5 n.a. 2.7 °Oand older n.a. 556.3 685.4 n.a. 455.8 588.9 n.a. 100.5 70.9 n.a. 1.9 n.a. 2.4 n.a. - 3.1 Total 6 and older n.a. 3124.6 3871.6 2445.4 3071.4 679.2 800.2 n.a. 2.0 n.a. 2.1 nA.. 1.5 Urban n.a. 1554.9 2388.8 1160.1 1793.5 394.8 595.3 n.a. 4.0 4. 3.8 Rural n.a. 1569.7 1482.8 1285.3 1277.9 284.4 204.9 n.a. -0.5 n.a _ n.e -2.9 -------------------------------…------------ C. Labor Participation Rates --/… ……-------- - --------- - ---- - ---------------------. Tota1 6 aLd older n.a. 39.42 35.49 n.a. 62.24 56.27 n.a. 16.99 14.68 n.a. -1.0 n.a. -0.9 n.a. -1.3 iotal 15 and older 64.25 54.31 49.81 87.30 87.66 80.19 42.88 22.36 19.92 -o.8 -o.8 - -0.8 -3.1 -1.0 Total 15-64 n.a. 55.31 50.98 n.a. 89.74 81.31 n.a. 22.87 20.79 n.a. -0.7 n.a. -0.9 n.a. -0.9 6-14 n.R. 3.43 2.58 n.a. 3.46 2.69 n.a. 3.40 2.29 n.a. -2.6 r.a. -2.3 n.a. -3.5 15-24 n.a. 49.76 39.53 n.a. 71.85 57.77 n.a. 27.19 21.23 n.a. -2.1 n.a. -2.0 n.U. -2.2 25-49 n.a. 59.14 58.82 n.a. 98.39 95.71 n.a. 21.59 22.17 n.a. - n.a. -0.3 n.a. 0.2 50 and older n.a. 50.14 46.14 n.a. 88.78 82.20 n.a. 16.86 9.22 n.a. -0.8 n.a. -0.7 n.a. -5.3 Total 6 and older n.a. 39.42 35.49 n.a. 62.24 56.27 n.e. 16.99 14.68 n.a. -1.0 n.a. -0.9 n.a. -1.3 :.b n.a. 3i6.18 . a. 54.42 n.a. 18.00 n.a. n.a. n.a. R.ral n.a. 34.43 n.a. 59.03 n.a. 9.56 n.a. n.a. n.a. SL/ see para 41, on the problems of Coaparability of data. E/ Ecancacaliy active population as a percent of population in the same age group. Ve: b'EC and Appendix Tables 1.1, 1.5, 1.6, 1.22 and 1.26-1.28. - 74 - 43. The census data suggests that the rural female labor force has been declining. The selectivity of the massive rural-urban migration process with respect to age may have been an important cause for'this: well over three-fourths of the migrants are under 25 and these are also the ages in which the peak of female labor participation rates is reached in rural areas. 1/ Older women have a lower propensity to migrate and do not tend to return to the labor market until after the child-bearing period. Men, on the other hand, reach their peak of participation in the labor market later in their lifetime. Thus, although migrant men also tend to be the younger ones, migration may have had a smaller impact in reducing overall male labor participation rates. 44. In urban areas, census data suggest that the migratory waves propellec the total EAP to a 4% average annual growth between 1961 and 1972 2/, a high rate although below the 5.6% yearly growth rate of urban population. The apparent fall in urban labor participation rates was probably for different reasons than in rural areas and was coucentrated in the under 24 years and over 50 years age groups (see Table 17). 3/ The rapid fall in labor partici- pation rates of the younger age groups was principally caused by the expansion of the educational system. 4/ Table 18 shows how in the case of LLma, for example, the fall in labor participation of persons under 24 years was closely linked to the rise in enrollment ratios. In the older age groups, and parti- cularly those over 60 years, the lower participation in the labor market was probably caused by the exDansion of social security, that allows a younger retirement for a larger number. This declining trend of labor participation for the youngest and older people is not surprising and may be expected to continue with sustained development in Peru. Conversely, participation in the urban labor market of the central age group (25-49 years old) appears to have increased in recent years. 5/ Since practically all the men in these age 1/ See Table 29 for the age structure of migrants. Appendix Table 1.27 shows the age-specific labor participation rates in 1972 for males and females. / The annual increase of the urban labor force was over 92,000 persons. This is probably much more than double the increase achieved in 1961 (which was probably around 38-40,000 assuming an annual growth rate of 2.5Z at that time). 3/ See Table 21 for the rate of growth of urban population. Table 17 shows that (for both sexes) urban labor participation rates fell drastically in the 6-24 age groups, moderately for the 50-74 age groups, but Lncreased for the 25-49 age groups. 4/ The 1972 Household Survey by the Ministry of Labor shows that :in urban areas an average of 11.3% of the EAP is at the same time studenlt. How- ever, in the 14-19 age group, the same share is a high 47.6%, suggesting that with higher family incomes the labor participation of this age group could still fall considerably even with small changes in enrollment ratios. 5/ See Table 17. - 7'5- Table 16: ESThIMATE OF SArtE OF CHANGE IN LABOR PARTICIPATION RATES EXPLAIE BY CHANGES IN AGE STRUCTURE OF POPULATION, 1961-72 ' (in percenta4mcs) Labor participation rates (LPR) Chanlges in LPR explained by changes in: Actuals TheLor. Absolute changes Shares 5/ 1961 1972 1972 Total 4/ 1/ 2~truct.31 Other a. b. c d-b-a- e=c-a-fd-e efd f/d ZAP 6 years and o:.der - Total 39.42 35.49 39.18 -3.93 -0.24 -3.69 0.06 0.94 Male 62.24 56.27 61.50 -5.9T -0.74 -5.23 0.12 o.83 Female 16.99 14.68 16.83 -2.31 -0.16 -2.15 0.07 0.93 - - EAP 15 years and o:.der -- Total 54.31 49.81 54-74 -4.50 0.43 -4.93 -0.10 1.10 Male 8T.66 80.19 87.50 -T.47 -0.16 -7.31 0.02 0.98 Fesale 22.36 19.92 22.51 -2.44 0.15 -2.59 -O.o6 1.06 1/ This table estimates how much of the change in labor participation rates (LPR) was caused by changes in age structure. It rirst calculates what the total LPR would have been in 1972, had the age spei:ific LPRs not changed since 1961 (variable "c"). The dijference between this theoretical LPR and the 1972 actual (variable e=c-a) is the share explained by clianges in age structure. &, Total change in LPR between 1961 and 1972. ]f Changes in LPR between 1961 snd 1972 explaincd by changes in age structure. In other words, how isuch the total LPR would have changed if the age specific LPRs had not changed. 4I Changes in LPR betwcen 1961 and 1972 explained by factors other than age structure (also equivalent to b-c or the difference beaWeen the theoretical 1972 total LPR and the actual 1972 total LPR. 1/ Expresses e + f as sAares of the total change in LPR. Source: ONEC (for methodology), Table 15 and Mission estimates Table 17: PEZRCETAGE CFANGE OF LABOR PART CIPATI ;N RATZS OF URBAN. MAL-EAD FM$A=L POPULATION BEMNE 1.961 ANZD 1972 (in percentages) A G E G R 0 U P S From year 6- lo- 15- 20- 25- 30- 35- 40- 45- 50- 55- 60- 65- 'O- 75 To year 9 14 19 24 29 34 39 44 , 49 54 59 64 69 .74 and over Total -40.0 -44.1 -32.8 -13.3 -0.6 0.3 0.7 1.3 0.7 -2.5 -5.2 -10.2 -15.8 -13.0 1.2 Vale -40.0 -36.5 -31.1. -;3.5 -3. -1.3- -0.8 -0.9 -0.3 -2.5 -4.3 -U'.1 -16.5 -14.0 -.5.4 Female -40.0 -50.0 -35.7 -11.4 s6.8 5.3 4.3 6.9 0.8 -6.6 -11.7 -18.9 -23.1 -11.1 :S.' Sou-ce: Appexdix Table 1.28 - 76 - Table 18: PERU: LABOR PADITCIPATION RATES AND EDUCATIONAL ENROLUMNT RATIOS FOR LIMI-CALLAO MTE- POLITAR ARFA BY SEX AND AGE GROUPS, 1961-72 (in percent) Enrollnent ratios Labor parti:ipation rates Age 6roups Total 4kale Female Total Male Female 1961 Total 26.0 2T.5 24.4 44.8 64.4 25.2 10-14 90.0 92.1 88.1 5.5 14.5 6.4 15-19 49.6 55.6 43.1 42.6 48.3 37.2 20-211 16.5 21.2 u.6 65.6 85.2 44.7 1972 Total 38.7 41.3 36.2 39.2 56! 22.2 10-14 9X.9 96.1 93.T 2.8 2.6 3.2 15-19 63.3 69.8 5T.2 27.2 31.0 23.5 2o-24 26.7 32.9 19.5 57.1 74.o 39.6 Change 1961-72 Total 12. 13.8 11.8 5.6 8.0 3.0 10-1 1 4.9 4.o 5.6 -2.T -1.9 -3.2 15-19 13.7 14.2 14.0 --15.4 -17.3 -13.7 20-24 10.2 11.7 7.9 -8.5 -11.2 -5.1 Source: OEC groups are already economically active, the increase in labor participation rate is explained by the changing behavior of women. 1/ The supply of female labor has been encouraged and facilitated by the change in traditional values that militated against female employment, and the higher educational attain- ments 2/, while the demand for female labor in urban areas has been expanding with rapid development of the service sectors. 1/ Table 15 and Appendix Table 1.26 show that in 1972, the share of men in the 24-49 age group that were already economically active was vver 95% (in the country as a whole and in urban areas). On the other hand, in the same age group, less than 23% of all women and less than 30% of the urban women were economically active, despite the in-migration of young women. 2/ The fall in fertility rates is a factor closely related with female LPRs. However, as discussed before, the lower fertility rates are probably more a consequence than a cause of higher LPRs. - 77 - Table 19: PERU: EDUCATIONAL PROFILE O' T.E ECONO4.VCALLY ACTI;iE POPULAT'Ol (E¶P ), 1961-72 1 (i.. percentages) 1 9 7 2 Location of EAP Level of 1 __1_ _, . __ _ _ _ _ _ __ _ _ _ _ _ Education'/ Totl Total U r b L n Rural Urban Total Lina Other Tctal Total Lima Other Rural Total 100.0 100.0 100.0 100.0 100.0 1.10.0 100.0 27.8 33.9 38.3 No Education 32.8 18.8 8.6 3.4 12.9 :16.7 100.0 4.9 22.5 72.6 Primary 52.2 54.6 52.0 44.6 58.1' 57.5 100.0 22.9 36.4 4o.T Secondary 11.1 19.4 30.1 39.9 22.1 4.5 100.0 54.7 36.9 8.4 Eigher 3.0 5.8 T.5 11.9 3.9 0.4 100.0 69.1 2T.3 3.6 Otber 0.9 1.4 1.8 0.2 3.0 0.9 i/ All daza from the Census of 1963. and 1972 exceplu Lima and "other urban". The Census dat. oh educational profile for the Lima-Callao metropolitan area *nLs not available, thwu the educational profile ectirated by the 1972 Househaold Survey of the Hinistry of Labor was applied to the total SAP figure from the Census. "Other urban" va obtained as a residual. 2/ Refers to one or more grades in each level. Sources: Table 15, Appendix Table 1. 29 , O1IO and Ministry of Labor 45. The expansion of the educatioual system has postponed the entrance into the labor market of a growing share of young people while rapidly improv- ing the educational profile of the labor force (Table 19). For the country as a whole, the main improvements in the 1961-72 period have been a drop in the share of the EAP with no education, and a doubling of the shares with some secondary and some higher education. Rowever, the geographical distribution of the more educated remains very unbalamced. The rural areas still have well over one-third of its EAP with no education and account for less than one-tenth of the country's EAP with some secondary education. On the other extreme, Lima's EAP has a negligible share with no education, and over half with some secondary or higher education. Moreover, almost 70% of the country's EAP with some higher education and 55% of those with some secondary education reside in Lima (although it accounts for less than 30% of the country's EAP). This geographical imbalance and the priiacy of Lima is mainly caused by the concentration of educational opportunities in the larger urban centers, and by the tendency of migrants from rural to Lirban areas to be the better educated. 1/ b. Determinants of Future Growth of the Labor Force and Alternative Projections. 1972-2000. 46. The size of Peru's labor force in recent decades has been determined by changes of both demographic factors, affecting the size of the population in working ages, and non-demographic factors, affecting labor participation 1/ See para. 56 for a discussion of the educational selectivity of migrants. - 78 - rates. Although demographic factors have been accelerating the growth of population in working ages, the impact of non-demographic factors that reduced the labor participation rate, had been more important in determining the size and growth of the EAP. This section analyzes if and how the relative importanc of these two sets of factors will change in the future and what are the orders of magnitude of size of the EAP that may result. The implications on Government policies are discussed in the concluding paragraphs. 47. The three alternative projections of the EAP shown in Table 20 com- bine two of the population projections discussed in para. 36 with two different trends in the change of labor participation rates. Alternative "A" basically postulates a continuation of past trends, combining the annual rate of change of the labor participation rates during 1961-72 with the population projections of Alternative I. 1/ Alternative "B" assumes the same population trends as in Alternative "A" but no change in labor participation rates. Alternative "C" combines a continued fall in labor participation rates with a faster decline in birth rates and population growth rate than suggested by recent trends. 2/ 48. The most important conclusion that can be derived from the projec- tions is that the growth of EAP will probably accelerate between now and 1990, how much depending on the changes in labor participation rates. Since there is a 15-25-year lag between the change in birth rates and their impact on the growth of population in working ages, the acceleration of population growth during the last decades will inevitably maintain a strong upward pressure on the growth of the labor force between now and 1990. After that, the impact of future demographic trends will increase, but will become important only after the year 2000. 3/ If labor participation rates continue to fall as rapidly as in the recent past (Alternative "A"), overall EAP will grow annually by at 1/ See Appendix Table 1.24 for a projection of the age specific 'Labor par- ticipation rates. Alternative I of the population projection assumes that fertility rates will continue to decline at the same speed as in recent years. 2/ This corresponds to Alternative IV of the population projecticins that assumes an aceleration in the fall of fertility rates. 3/ Comparing Alternatives "A" and "C" shows how in 1973-90, different birth rates would have little impact on the size of the population in working ages, except for the 6-14 age group. A lower birth rate would lower the growth of the 15-24 age group only in the 1990-2000 period. Table 20: PEHU: ALTE;RATIVES PROJECTIONS OF ECONOMICALLY ACTIVE POPULATION, 1972-2000 In thousands for population and ELPi in % for labor particip. rates Average annual growth rates Alter. A Alter. B Alter. C 27 Est. Alter. A Alter. B Alter. C Eat. 1972 199q 1972 1972- 1990- 1972- 1972 19-0 1972- 1972 1990 300 1990 2000 1990 2000 1961-72 -90 .2000 2000 -90 2000 2000 -90 2000 2D00 …---------------------------------------------------------…Population---,---------------------------------…-------------------------- Tota; 6 and older 10909 19613 26212 19613 26212 19357 25039 2.9 3.3 2.9 3.2 3.3 2.9 3.2 3.2 2.6 ' D .otal 15 and older 7601 13911 18920 920 13911 18612 2.8 3.4 3.1 3.3 3.4 3.1 3.3 3.4 3.0 3.3 Tota1 15-64 7079 13091 17798 13091 17798 13091 17490 2.8 3-5 3.1 3-3 3.5 3.1 3.3 3-5 2-9 3.3 6-i4 3309 5732 7292 5702 7292 5446 6427 3.3 3.1 2.5 2.9 3.1 2.5 2.9 2.8 1.7 2.4 15_:!, 2570. 47?5 6331 4775 6331 4773 6023 3.2 3.5 2.9 3.3 3.5 2.9 3.3 3.5 2.4 3.1 25-ij 3545 6610 9138 6610 9188 6613 9188 2.6 3.5 3.3 3.5 3.5 3.3 3.5 3.5 3.3 3.5 5C and older 1436 2525 3401 2525 3401 2525 3401 2.? 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 …--------…-…-…- -------------------------------…Economically Active Population (EAP)…------------------_ --------------__________- TonaO 6 and older 3872 6343 8365 7136 9680 6342 8293 2.0 2.8 2.8 2.8 3.5 3-1 3-3 2.8 2.7 2.8 Total 15 and older 3786 6256 8277 6988 9488 6258 8216 2.0 2.8 2.8 2.8 3.5 3.1 3.3 2.8 2.8 c.8 1o;a-1 15-64 -3t09 6034 8008 6712 9165 6036 7947 2.0 2.9 2.9 2.9 3.5 3.2 3:.4 2.9 2.8 2.9 c- . °5 87 88 148 192 84 77 0.7 0.1 0.1 0.1 3.1 2.6 3.0 -0.1 -0.9 -o.4 52 1. 10:6 1373 1543 1907 2534 1373 1482 1.0 1.7 1.2 1.5 3.6 2.9 3.3 1.7 0.8 1.4 2'_!,9 2085 3820 5386 3891 5407 3820 5386 2.5 3.4 3.5 3.4 3-5 3-3 3.5 3-4 3.5 3.4 50 and older 685 1063 1348 1190 1547 1063 1348 1.9 2.5 2.4 2.4 3.1 2.7 3.0 2.5 2.4 2.4 Yae.rly net addition to AP 6 aas older 76 173 229 238 291 172 218 5.9 4.7 2.8 4.0 6.6 2.0 4.9 4.6 2.4 3.8 15 - 64 71 170 226 227 284 170 216 6.1 5.0 2.9 4.2 6.7 2.3 5.1 5.0 2.4 4.0 …--------------------------------------------------…Labor Participation Rates------------------------------------- :otul 6 and older 35.49 32.34 31.91 16.44 36.93 32.76 33.12 hotal 15 and older 49.81 44.97 43.75 50-31 50.15 44.98 44.14 -otal ;5-64 50.98 46.09 44.99 51.35 51.419 46.11 45.44 2.58 1.53 1.21 2.60 2.63 1.54 1.20 5-2. 39.53 23.75 24.37 39.96 40.03 28.75 24.61 25-45 58.82 57.79 58.62 58.87 58.85 57.79 58.62 5, ara older 46.14 42.10 39.64 66.14 45.49 42.10 39.64 i/ Asumes a continuation of past trends of both population growth and change in labor participation rates. It therefore applies a projection of the age .-acific labor participation rates (assuming that the average yearly change recorded during 1961-72 in each age group will continue in the future; see 4pFendix Table 1.24 for details) to the projected age structure resulting from the Alternative I of the population projection (see Appendix Table 1.9 .or catalls). / A-sames a continuation of past trends in population growth but no change in labor participation rates. It therefore applies the age-specific labor Farticipation rates recorded in 1972 (see Appendix Table 1.26) to the same projection of population age structure as in Alternative A. Ahsumes an accelerated fall in population growth and a continuation of past trends of changes in labor participation rates. It therefore applies the .saz.e age specific labor participation rates of Alternative A to the age structure resulting from the lowest population growth alternative (Alternative IV) sLown on page ;/ ;..csber of net new entrants to EAP at average annual growth rate of previous period. ocirce: Table 14-15, Appendix Tables 1.9, 1.24, 1.26 and Mission estimates. - 80 - least 2.8Z between 1973 and 2000, compared with 2% in the 1961-72 period. 1/ This implies that the number of new job-seekers in the 15-64 age group will more than triple from around 70,000 in 1972 to over 225,000 in 2000. However, this projection assumes that through higher family incomes and a continued rapid expansion of educational opportunities and social security coverage, the share of economically active people will fall 50% in the 6-14 age group, 40% in the 15-24 age group and 30% in the over 64 age group. 2/ If this does not occur, then the annual growth of overall EAP could reach 3.3% and the number of, new job-seekers in the 15-64 age group 285,000 in 1990 (Alternative "B"). Since most (if not all) of the increase in EAP will take place in urban areas, the behavior of female participation (particularly in the central 25-49 age groups) will also be an important variable determining the size of the labor force. The projections assume a very small change in female labor participa- tion rates, which may be an unrealistically conservative assumption. 4/ 49. The two main operational and policy implications from the projections are the following: first, the Government will be faced by an accelerating growth of the labor force for at least 15 years to come. In fact, in the coming years the labor force will be growing at its highest rates in Peru's history; the growth in the last 15 years is relatively small in camparison. The second implication of the projections is that the extent to which the growth rate of the labor force will fall beyond 1990 will depend crucially on 1/ Although the total dependency ratio (population under 15 and over 64) is projected to start falling in future years because of the drop in birth rates (see Table 14), the projections in Alternatives "A" and "C" imply that the economic dependency ratio (number of non-economically active per 100 economically active) will keep increasing up to 1990 because of the fall in labor participation rates. The latter is projected to rise from 250 per 100 in 1972 (see Appendix Table 1.25) to 283 per 100 in 1990. However, after 1990, it is projected to start falling because of the greater impact of demographic factors. 2/ Alternative "A", by assuming a continuation of past trends, projects a fal in the labor participation rates between 1972 and 2000 from 2.58 to 1.21% for the 6-14 age group (which is about 30% of the population 6 years and over), from 39.53 to 24.37% for the 15-24 age group (which is about 24% of the population 6 years and over), and from 33.91 to 23.98% for the 65 and over age group (which is about 5Z of the population 6 years and over (see Table 20). 3/ See para. 66 for a discussion of patterns of urbanization. Moreover, as shown in Table 15, already in the 196L-72 period, all the increment of the labor force was located in urban areas. 4/ An increase of one point in the labor participation rates of females 25-49 is equivalent to a 0.5% increase in the total labor force. The labor participation rate of females in this age group in Lima is about 5-6 points higher than in other urban areas and about 20 point:s higher than in rural areas. (See Appendix Tables 1.26 and 1.27.) - 81 the speed with which birth rates decline in the next decade. The slower fertility falls, the more into the long--term future will the time be postponed when labor force growth starts declinini. - C. SPATIAL DISTRIBUTION AND MIGRATION 50. As mentioned before, Peru has been going through a process of massive migration and spatial redistribution of its population. This section reviews this process in more detail highlightin, its main features, causes, consequences on development and the role of Government policies. First, the main charac- teristics and dimensions of urbanizatioa and migration are discussed. Then, the regional disparities are reviewed to identify the causes and private returns of the migration process. This is followed by a search for the con- sequences for the regions involved and for the country as a whole. Finally, the role of Government policies is reviewed within the context of the main issues raised by the possible future trends in spatial distribution of the population. a. General Characteristics of Urbanization and Migration 51. Urban population growth has been accelerating rapidly and at the present yearly rate of 5.6%, urban population will double in 13 years (Table 21) 1/. In general, the urbanization process has had three principal features: movement to the coastal areas, increased concentration in larger cities and overwhelming primacy of Lima. However, in recent years the urbanization patterns have been changing gradually, with the medium-sized cities growing faster than Lima and absorbing an increasingly large share of the incremental population. 2/ 52. The Costa has been the main target in the urbanization process. Its share of total population rose from 282 in 1940 to 46% in 1972, 3/ and in the 1961-72 period almost 70% of the pcopulation population were residents of the Costa. 1/ In this report urban population is defined as those living in towns of more than 2,000 inhabitants. The Census definition differs slightly since it includes also the capitaLs of districts regardless of their size. The adjacent map shows that the Coastal departments in 1961-72 were in general growing faster than those in the Sierra and had attained a higher degree of urbanization. However, many departments in the Sierra continue to have population densities equal or higher than above those in the Costa. 3/ See Table 8. PERU: DENSITY, GROWTH RATE AND DEGREE OF URBANIZATION BY DEPARTMENT, 1972. POPULATION DENSITY AVERAGE ANNUAL GROWTH RATE DEGREE OF URBANIZATION INHABITANTS PER SQUARE KILOMETER 1961-72 IN PERCENT PERCENT OF POPULATION IN TOWNSIJ (Nuionol Average: 10.9 pesom per lan") (Nalional Average: 2.9%) (Natlonal Averug.: 53.2%) I ti b .' 4 1 3 * 9S 4 I 0 93. 9 |0 1 3 .6 0I( .S j 1E3i 2.0 ... ........'j>ES 30.1 45'.0 i0.3 56 -0.1 01. .__ 31.t I 11 Urban populolion d er9ned as lining in loans of AM AS NUIIIBERlS CORflESPDND TO NAt4ES OF D EPA PITMNENTSUSTA Au. d n*u 0h... o5i, 2..p drAlI_ Urban .p.lat-- deined a livig in twns o I AMAONAS CAJAMARCA 11 JUNIN 18 MADRE Of DiOS 21 SAN MARTIN ArI nr,nrgn,h h 2000 inhobilonls or ore. 2 ANCASH 7 CUZCo 12 LA LIlERYAO 17 MOOUEGUA 22 TACNA World n;; n U n S.o,,.. App.nndrn abba.s I I In 1 3 3 AP'URIMAC 8 HUANCAVELICA 13 LAMOAYEr3UE IS P~ASCO )3 ISIMRES 4 AREGUIPA 9 HUANUCO 14 LIMA 19 POURA S AYACUCIO 10 ICA IS LOSETO 20 POUNO - 83 - Table 21 GROWTH AND SHARE IN TOTAL UI3BAN POI ULATION OF CITIES, 1940 - 72 L-- 0 C Number of years A ShLre of urban Averaee annual grovth to double T In thousands population rates (in S) population I _ _ _ _ __ _ _ _ _(in, % a e i )p p 1 t o City size - N 2/ 1940 1961 1972 1940 1961 1972 1940-61 1961-72 Bates-. 1940-61 1961-72 Total population 6208.0 9906.7 13572.1 2.3 2.9 1.26 30 24 RUIal population 4536.8 5933.0 6373.3 1.3 0.7 0.54 51. 99 Urban population " 1671.2 3973.7 7198.8 100.0 100.0 100.0 4.2 5.6 1.33 17 13 Lim-Callao CC 601.8 1692.7 318.4 36.0 42.6 43.9 5.0 5.o 1.16 14 12 100 - 350000 280.7 691.4 1392.7 16.8 17.4 19.3 4.4 6.6 1.50 16 11 Costa 100.6 330.8 T29.0 6.o 8.3 10.1 5.8 7.4 1.28 12 9 Trujillo NC (37.0) (103.0) (253.6) (5.0) (8.5) (1.70) (22) (12) Chiclayo NC (31.5) (95.7) (189.7) (5.4) (6.4) (1.19) (13) - (3i) Chimbote NC ( 4.2) (60.0) (159.0) (13.4) (9.3) (0.69) ( 6) ( 8) Piura NC (27.9) (72.1) (126.7) (4.6) (5.3) (1.15) (15) (13) Sierra 148.3 302.8 552.4 8.9 7.6 7.7 3.5 5.6 1.60 20 12.7 Arequipa SS (80.9) (158.7) (304.7) (3.2) (6.2) (1.94) (22) (12) Cuzco SS (40.7) (79.9) (120.9) (3.3) (3.9) (1.18) (21) (18) Huancayo CS (26.7) (64.2) (126.8) (4.2) (6.4) (1.52) (17) (11) Selva 31.8 57.8 111.3 1.9 1.5 1.5 2.9 6:2 2.14 24 12 Iquitos IS (31.8) (57.8) (111.3) (2.9) (6.2) (2.14) (24) (12) 50 - 99999 55.5 158.2 301.4 3.3 4.0 4.2 5.1 6.o 6.18 14 12 Costa 53.1 131.8 238.1 3.2 3.3 3.3 4.4 5.5 1.25 16 13 Ica CC (20.9) (54.2) (95.3) (4.6) (5.3) (1.15) (15) (13) Sullans 2C (21.2) (50.1) (83.6) (4.2) (4.8) (1.14) (17) (15) Tacna SC (11.0) (27.5) (59.2) (4.4) (7.2) (1.64) 4(16) (10) Selva 2.4 26.4 63.3 0.1 0.7 0.9 12.1 8.3 0.69 6 9 Puvallpa LS (2.4) (26.4) (63.3) (12.1) (8.3) (0.69) (6) (9) 20 - 49999 Si 212.1 403.7 n19.6 12.7 102 10.0 3.5 54 1.54 20 13 2 - 19999 521.1 1027.7 1626.7 31.2 25.9 22.6 3.3 4.3 1.30 21 16 / Ranking by city size based on 1972 population. The Table shows how cities with a given size in 1972 grev in 1940-72. It therefore differs from Table 22 thet shovs the concentration of population by city size. g/ NC = North Costa CC = Central Costa SC = Southern Costa CS = Central Sierra SS = South Sierra B = tr4 Selva. 3/ Urban population defined as all cities with 2000 inhabLtants or more. J A close look at the Census data showed that the metropDlitan areas of variou: cities had absorbed contiguous districts. The latter were added to the population of the city: City Districtu added Trujillo El Porveiir, La Esperanza, Victor Larco flerrera and Laredo -uancayo Chilca and El TanDo lca IlL linguiia, r.rco.2a, Salat, San Juan Bautista and Subtanjalla Sullana Bellavi:ta Iluce.1lpa Callaria Lnd Yari iacocht Tacna lPocollAy 5/ Excludes Bellavista. incllded with Sulluna. Sourf e: ON?lC nmid Mis!;ion astirates. - 84 _ TIble 22 : DISTRIBUTION OF POPULATION IN CITIES WIT4 20,000 INEANITAJTS OR MORE 3Y REGION, 1940-72 In thousands Percentage Average annual Number of Structure Growth rates cities 1940 1961 1972 1940 1961 1972 1940-61 1961-72 1940 1961 1972 Total Urban Population 16T1.2 3973.7 7198.8 4.2 5.6 187 268 347 20,000 and more 91. 2682.7 5518.9 100.0 100.0 100.6 5.2 6.8 10 26 .36 Costa 739.8 2136.4 4429.1 8o.4 79.6 80.3 5.2 6.9 6 13 20 Lima-callao (601.8) (1692.7) (3158.4) (65.4) t63.1) (5T.2) (5.0) (5.8) (1) (1) (1) Other (138.0) (443.7) (1270.7) 415.0) t16.5) (23.0) (5.7) (10.0) (5) (12) (19) Sierra 148.3 462.1 900.0 16.1 17.2 16.3 5.6 6.2 3 11 13 Selva 31.8 84.2 189.8 3.5 3.1 3.4 4.7 7.7 1 2 3 2000 - 19,999 T51.3 1291.0 1679.9 2.6 2.4 167 242 311 1/ P.azked by city size in each year. The table shows the degree of concentration of urban population in cities. It therefore differs from Table 21 that ranks cities according to size in 1972. Source: O:$C Table 23: PERU - INDICATORS OF URBA. CONCENTRATION, 1940-72 l/ Index Average Annual _____ ____ ____ ____ Growth Rate 1940 1961 1972 1940-61 1961-72 Gini coerficient of concentration 0.632 0.730 0.734 0.7 0.1 Index of dissimilarity V 50.00 61.10 66.48 1.0 0.8 Pereentage share of population in cities: - 2,000 - 19,999 within urban popul. 45 32 23 -1.6 -3.0 - over 20,000 within urban popul. 55 68 77 1.0 1.2 - 20,000 - 99,999 within urban popul. 19 19 14 - -2.7 - over 100,000 within urban popul. 36 49 63 2.2 2.3 - over 100,000 vithin cities 65 70 87 1.7 2.0 over 20,000 17 To measure the degree of concentration of population in number of cities of different sizes. The size groups used were: 2000-2499, 2500-4999, 5000-19999, 20000-99999, 100000-349000, and over 350000. 2/ Varies between 0 (meaning equal size of all cities) to 1 (meaning extreme concentra- tion). Calculated as o C r (Xi_t * T;) - 3c tX - yj ) 3 10,000 -.i i- where Xi = population in size grdup as % of total urban population. Y. = number of cities in size group as % of total number of cities. / Compares the growth of the share of population in one size group with the growth of the share of cities in the size group. Varies between 0 (mearning the change in popula- tion share is equal to the change in the share of number o1f cities) and 100 (meaning the population share grew without any change in the share of number of cities). Calculated as 1 - 2 CcCX. - Y.) ] where Xi and Y. are the same as in footnote 2. Source: OX=,"Contribuci6n al Estudio de la Concentraci6n Urbana en el Peru 1940-72, Boletin de Analisis Demogrifico No. 14, 1974 and Mission estimates. - 85 - 53. The concentration of urban population in the larger urban centers at an accelerating rate can be seen front the indicators on Tables 21 to 23. Between 1940 and 1972 the share of urban. population living-in cities over 100,000 almost doubled to 63%, while the share living in towns under 20,000 persons was halved to 23%. 1/ The overwhelming primacy of Lima is also obvious from the Tables. 2/ 54. However, the pattern of urbanization has been changing in recent years and within the group of medium and large cities the degree of concen- tration of population is diminishing. 31' In the 1940-61 period the population growth rate of Lima was 20% higher than that of total urban population and higher than that of most other medium-sized cities. 4/ In the 1961-72 period, the growth rate pattern has been reversed: Lima is growing only 4% faster than total urban population and slower than most other medium-sized cities. 5/ In other words, the urbanization process is starting to gradually decentralize to a greater number of emerging large ciLties with population over 100,000. The North Costa now has the most dynamic urban growth; three cities located there -- Trujillo, Chiclayo and Chimbote -- have a combined annual growth rate of almost 8% and now account for 15% of the urban population outside of Lima. However, even within the Sierra the growth rate of two cities -- Arequipa and Huancayo -- rose by over 60% between the two intercensal periods and surpassed the growth rate of Lima. 6/ 1/ See Table 23. 2/ See para. 59 below for a more detailed discussion of regional concen- trations of economic activity and of the primacy of Lima. 3/ See Tables 21-23. The fall in concentration in the group of larger cities explains the fall in the yearly growth rate of the Gini co- efficient of concentration from 0.7 to 0.1% and of the index of dissimilarity from I to 0.8% (See rable 23). 4/ Medium-sized cities defined as those with a population in 1972 of 50,000 or more and large cities those with a population in 1972 of 100,000 or more. The only medium and large cities growing faster than Lima in the 1940-61 period were Chimbote (which was growing at over 13% a year because of the fishmeal boom), Chiclayo and Pucallpa. See Table 21. 5/ The only exceptions were Piura, Ica and Sullana in the Costa and Cuzco in the Sierra. See Table 21. 6/ The annual growth rate of cities in the 50-350,000 range rose from 4.5% in 1940-61 to 6.5% in 1961-72. (See Tables 21 and 22.) - 86 - 55. The urbanization process has been fed by an ever-growing wave of permanent migrants. 1/ Inter-provincial migration in recent years has been growing well in excess of 6% a year, and by 1972, about- one-fifth of the country's population had moved to a different department than that of their birth (Table 24). 2/ Tables 24-28 give a synoptic view of the patterns and final destination of the migratory process in the 1940-72 period. There is substantial evidence to highlight the following probable features of the migratory process in Peru: (a) Although in aggregate terms the direction of migration has been from rural to urban areas, there seem to be two parallel movements: rural to small urban centers and from there to larger urban areas following a "fill-in" pattern between generations. 3/ Thus, in the case of Lima, over half of the migrants come from towns of more than 5,000 inhabitants (where only 25% of the population lives); over 70% come from the Sierra (travelling relatively longer distances); and for most, it was the only migratory movement in their lives. 4/ 1/ The review here refers to permanent migration. In addition, a large share of the rural population migrates seasonally in search of employ- ment and income mainly to the larger agricultural farms, but also to mining centers. The main sources for the analysis are the 1940, 1961 and 1972 Population Census, a National Household Survey by the Ministry of 'Labor in 1970 (referred here as the 1970 Household Survey), and a 1965 Survey of Immigrants into Lima. "Migrant Population" is defined in the census as those who were born in a different location than where they resided in the Census year. 2/ How rapidly migration has been growing can be inferred from i:he yearly growth rate of urban population, which rose from 4.2% in 1940-61 to 5.6% in 1961-72 (Table 8). Since, as discussed before, the rate of natural growth of the urban population is probably not higher than that of the rural population, then total migration must have been growing well in excess of 6% yearly in recent years. (See also Table 24). 3/ Emigrants from rural areas move to small urban settlements in nearby provinces or departments. Then, in subsequent generations, they move to larger and more distant cities replaced by a new generation cf migrants from rural or less urbanized provinces. However, this generalization does not imply a movement "by steps" for specific migrants. What it implies is that subsequent generations will migrate to larger cities. 4/ The 1965 survey of immigrants to Lima showed that 33% came from cities with 20,000 inhabitants or more and 17% came from cities between 5,000- 20,000 inhabitants (Appendix Table 1.51.) The survey also showed that 80% of those that arrived in the 1956-65 period moved only once in their lifetime. Table 26 and Appendix Table 1.48 show that 70% of the immigrants before 1972 came from the Sierra. - 87 - Table 24 : PE: nf=A.L MIGRATION TREDS, 1940-72 Pe-centage of: Average Anw-al Iz thousands Total polston Total ni carts Orovtn naze 1940 1961 19T2 1940 l96l 1972 19L40 1961 1972 9 61 1961-T2 Total° oDulation 612.4 9840.0 13416.5 100.0 100.0 100.0 100.0 100.0 100.0 2.3 2.9 Total Migrants Between. / Provinces (&.b+c.i) 677.0 2279.8 n-. 11.1 23.2 n.a. 6.0 M.&. .3egions 3/ 385.0 1162.2 1981.8 6.3 11.8 14.8 56.9 51.0 5.4 5.0 b.Departments withi.rsgion! Y 179.9 406.2 627.0 2.9 4.1 4.T 26.6 17.8 4.0- 4.0 c-Provincs within depart 2/ 112.1 711.4 n.e. 1.9 T.2 n.r. 16.6 31.2 9.2 n.L. Provinces vithin region (b+e) fi 292.0 111T.6 n.a. 4.8 11.4 n.e. 43.1 49.0 6.6 n.e. Departmnts (&+b) V 564.9 1568.4 2608.8 9.2 15.9 19.4 83.4 68.8 5.0 4.T Pet Migration Fros: 8/ Sierra to other regiso 191.2 565.T 10 15.?. 5- _5 to Cost& (total) laKV 53 .10 5.1 5.t to Lima (133.1) (48T.9) (895.4) (6.4) (5.7) to other Costa (55.3) (T5.4) (1114.4 (1.5) (3.9) to Salva 2.8 2.4 5.9 -0.7 8.5 Other Costs to Lim 57.7 195.8 299.2 6.0 3jj 1/ Refers to persams that in the census year were resident elueghere than where ther were born. It therefore excludes Sig.snts vho died between ceneua years. T otal number of persons who in the censa year were resid±rLg in a province other than the one where they vere bor. It i-ludcee terefore, all the zigrantsa except thoag who migrted within the provice. 3/ Migrats to aother region tha the one where they were born. Y/ Migrants to a province in different department but in the siam region as the one whare they were born. 5/ Migrants to another province within the ame department whvre they were born. 6/ igrants to naother province witbin the sam region. xi/ Mgra=ts to another depaetont than the one where they verak born. / drants minus immigrants betysen regions. Sonree: ONC Table 25: PERU - MIGRANT POPUIJkTION TO EIST CITIm BSt YSA OF AIUtIVAL, 1970 I' Total Before 1941- 19,1- 1961- City Location Migrants 1940 50 60 70 1970 A. Percentage Distribution of Migrants by Decade Lime Costa 100.0 17.I4 19.9 27.9 34.8 (8.6) Areguipa, Sierra 100.0 14.6 15.0 28.6 41.8 (10.1) Trujillo Costa 100.0 11.9 14.9 26.5 46.7 (6.6) Cuzco Sierra 100.0 14.6 17.1 26.3 42.0 (12.7) Chiclayo Costa 100.0 12.2 13.8 28.2 45.8 (9.7) Piur Costa 100.0 10.3 13.0 29.6 47.1 (11.9) Puno Sierra 100.0 10.8 10.5 25.0 53.7 (8.4) Iquitos .. Selve 100.0 22.6 16.3 27.0 34.1 (16.3) B. Relativz Grovth of Migrant Porulation- Lima Costs 100.0 114.4 714.8 53.5 (9.44) Arequipa Sierra 100.0 102.8 97.0 71.9 (11.2) Trujillo Costa 100.0 125.3 99.2 87.7 (7.1) Cuzco Sierra 100.0 117.2 82.5 72.5 (13.5) Chiclayo Coste 100.0 113.1 108.5 84.6 (14.6) Piura Cost- 100.0 126.5 127.0 89.2 (10.7) Puno Sierra 100.0 96.8 115.8 115.0 (19.5) Iquitos Selve 100.0 72.0 69.5 51.8 (9.2) 1/ The eight cities accotnted in 1970 for about 35% of total population and 65% of urban pooulation. 2/ Migrants in decade divided by accumulated nigrants ia previous decades times 100. Source: Appendix Table 1.49 - 88 - On the other hand, the Arequipa department -- where the second largest Peruvian city is located -- sent over 70% of its immigrants to the Lima-Callao area and received over 65% of its immigrants from Cuzco and Puno, which are less urbanized and poorer departments in the Sierra. In turn, the cities of Puno and Cuzco received over 70% of their immigrants from other provinces within the same department. 1/ (b) The largest flow of migrants has been between regions (from the Sierra to the Costa and from areas within the Costa to Lima) and over two-thirds of them have gone to Lima. 2/ However, in recent years the relative importance of net migration within regions 3/ and to other cities in the Costa has been growing. 4/ The latter is consistent with the 1/ See Table 26 and Appendix Tables 1.47, 1.48 and 1.50. 2/ Table 24 shows that about half of the interprovincial migration is between regions and about one-fifth is between departments. About one-fourth of the interdepartmental migration takes place within regions. Table 26 shows that almost 65% of the migrants between regions came from the Sierra and over 70% went to Lima. However, Table 28 shows that in relative terms, the net outflow of people (measured as the ratio of net immigrants to total migrants) has continued to increase in the Sierra but has started to decline in the Costa (the latter due to relatively more people coming from the Sierra and less going to Lima). 3/ Between provinces and between departments of the same region. 4/ Table 24 shows that the average annual growth rate of net migration to Lima declined between 1940-61 and 1961-72 from both the rest of the Costa (falling from 6 to 3.9%) and the Sierra (falling from 6.4 to 5.7%). On the other hand, the average annual growth of net migration from the Sierra to the Costa excluding Lima, which was -0.7% in,1940-61 rose to 3.9% in 1961-72. Table 25, on time of arrival of migrants according to the 1970 Household Survey, also suggests that the intensity oE migration into Lima is declining while it is rapidly rising in other cities. At Lima the migrant population was doubling almost every two years in the 1940-60 period, while it grew only 50% in the 1960s. In other cities of the Costa and the Sierra, the migrant population increased between 70 and 115% in the 1960s. - 89 - XgJS 26 *tXU AD DEZSTIM;AMZ0 CY MTCR?MRAS Ul!W= DCPUTMSS AND 1. REG':I;S, 1972 (ia zrault) 3OaGD 0 IIIOAN VEDTIU.T7CN OF EWWAXo3; Total Liin Costa Sierra S.lya Total Lisa Casma Siqrra SoIj'.- * atween Peions Total Migrants 100.0 , 24.6 63.6 4.1 100.0 . 15.1 12.4 Lina-Callo T_1_. - 2 7 t _ - 6- r. Costa 100.0 17.1 - 80.9 2.1 l00.C 72.3 - 26.4 1.3 Sierra 100.0 39.6 52.2 - 8.2 100.0 78.7 19.2 - 2.0 Selm 100.0 17.4 16.6 66.1 - 100.0 11.T 7.6 26.3 - Jetwen Deaartuets Total Nignts 100.0 10.7 21.5 6] 6.1 100.0 5 14.4 2k.8 2. Ll,-Callao 100.0 8.1 23.0 65.5 13i 100.0 4i.3 24. Costa 100.0 13.6 20.2 66.5 1.6 100.3 62.5 ;3.5 22.9 1.1 Gierra 100.0 15.0 19.8 62.C 3.1 100.0 59.7 14.6 24.1 1.5 Jalv 100.0 10.4 19.8 39.3 k0.6 1oo.0 51.1 5.8 la.7 24.4 Bmxoe: Appd --Shl 1.46 to 1.48 Tbl@e nR a1rS 0? DaGsc3.o0 AND XIMaaATION BErIYER DPAR$KTS AfD BE5IEEm R=IONS, 1940-72 (in percent) Pt±iration Rate Iigz'tion RNte 2/ 1960 196i 1972 1960 1961 1972 Total between Denartments 9.2 1S 19.4 2.2 I15. 12-A LlMA-CaLll o qt, 11.2 ll.2 30.8 39-9 40.8 Costs 9.8 17.6 19.7 9.6 12.6 16.1 Sierrn 8.8 17.0 22.8 4.2 7.9 10.3 Salve 9.0 1U.2 13.9 9.2 7.7 8.8 atRW betveen aiens 6.3 11.8 14.8 6. 1l.8 14.8 Llm-Callao 5.7 6.5 6.2 26.9 36.5 37.5 Costa 8.0 14.8 1T.0 T.8 9.6 11.3 Sierra 6.1 12.5 17.3 1.4 2.9 3.9 Balm 3. T.6 10.5 3.6 4.0 5.2 / iLr=-ts x 100, vbere native population " total population - net migration. Native population j IlaijDnts as a percent of total populatioi. Source: Tables 1.36 and 1.3T Table 26: MET NIGRAATIO tWDICES B£TWEFU NEGIONS, 1940-72 'In percentages) 1960 1961 1972 Total LI= Costa Sierra Sclva Total Lia Costs Sierra Selve Total Lima Cost& Sierra Selva Total - 72.7 -2.1 -64.4 2.5 - 7E.5 -26.3 -65.1 -32.6 - 80.0 -23.5 -67.5 -35.8 Lima-Cll o 7T.7 - -72.6 -72.7 -73.6 78.5 . -75.9 -79.7 -76.5 80.0 - -T7.5 -82.1 -TS.6 Csta -2.1 T2.6 - -53.2, 4674 -2..V3 7!;.9 - -32.5 -0.3 -23.5 74.5 - -30.9 - 2.4 siesra r 72.7 53;2 - 29.0 .1 7TS.7 32.5 - 9.5 -67.5 82.1 .30.9 - 13.1 SaLia 2.5 73.6 -47.h -29.0 - -6 7T6.5 0.3 -9.5 - -35.8 78.6 2.4 -13.1 - Iigrants - Uhigrants / igration iadez - X 100. E-wlkes migrants witbin region. Iigrants + dgrants Source: Appendix Table 1.35 - 9o - recent rapid acceleration of the growth of medium-sized cities discussed above. 1/ (c) By 1972, the people that had migrated from the Sierra were equivalent to almost one-fifth of the native population 2/ while almost two-fifths of Lima's population had been born in other regions (Table 27). 56. The migration process is highly selective. Within the community of departure the average migrant tends to be the young, dynamic risk-taking indi- vidual with an above-average education, and not of the poorest income groups. Those with relatives or friends in the target city and those who suffer ethnic or social discrimination in the native community are also more apt to migrate. 3/ The evidence on age and education is particularly conclusive. According to the 1970 Household Survey, 60-80% of the migrants over 15 years old were in the 15-24 age group; this share is twice the national average for the total population. 4/ However, half of the migrants were 14 yesLrs old or 1/ With immigration from rural areas, mining centers have at times become burgeoning towns and in some cases small cities, such as La Oroya, Morococha and Cerro de Pasco in the Central Sierra, San Juan in the Central Costa (iron ore mines) and Tacna in the Southern Coast (Toquepala Copper Mines). However, the migration flows last only as long as the mines expand, and the growth of these towns later stagnates. Moreover, Martinez ("Las Migraciones Internas en el Peru, 1968") states that a large share of the migration to mining centers is transitory either to return to their original towns seasonally or until better economic opportunities appear. Other studies by Figueroa have found that there is a complementarity between the agricultural harvesting cycle and the mining production cycle, giving support to the assertion that a part of the migration to mining towns is seasonal and transitory. 2/ Native population defined as total population minus net migration (Table 27). 3/ See Alers, J. and Appelbaum, R. La Migracion en el Peru. un Inventario de Preposiciones, CEPD 1968. This study quotes various surveys thiat show that a high share of migrants chose their target city because of the presence there of relatives or friends. In a study now being prepared of the urban poor in Lima, A. Scott found that kinship was an Lmportant determinant in economic relationships and that frequently small businesses filled their labor requirements with relatives and friends in their place of origin. 4/ Those over 15 are probably the decision-makers in the migration process and are in the economically active ages. The 1961 Census showed the same young bias in the age pyramid of inter-provincial migrants, alt:hough with a somewhat smaller share in the under 15 age group and a larger share in the 15-29 age group. - 91 younger, suggesting the movement of whole families. 1/ Table 30 suggests that the educational profile of migrants is better than average for the total population and for the non-migrant population, although the educational profile of the migrants is worse than that of the natives of-the city to which they migrate. 2/ There is some evidence to also suggest that the more educated migrants tend to go to the larger cities. 3/ The evidence on how poor the migrant is before migrating is more fragmentary. However, the following points can be made: (a) the above-average education probably also means higher incomes, given the close correlation between the two (particularly in rural areas); (b) various studies of migrants to Chimbote and Lima suggest that in the community of origin a lower share of emigrants than non-emigrants was dedicated to agriculture; 4/ and (c) for most, a minimum income level is necessary to cover the costs of moving before migration can be considered. 5/ 1/ The evidence on sex composition of migrants is not fully consistent among different studies, although it seems to also vary with the size of the cities of final destination: the larger and more urbanized the recipient area, the higher the share of women. The 1961 Census found that for the country as a whole, the male-female ratio was 113, but ranged from 125 for those who migrated to the more rural Sierra to 105 for those who migrated to Lima (Appendix Table 1.54). ThLe 1970 Household Survey also found that particularly in the 15-24 age group the sex ratio of immigrants was below 100 in most of the larger ci'ties -- Trujillo, Chiclayo, Piura and Arequipa -- but over 100 in the less urbanized Cuzco and Puno (Appendix Table 1.55). However, the survey did find a sex ratio somewhat over 100 for Lima. One important aspect determining the pattern of lower sex ratios in urban areas is the large number of female migrants who obtain employment in the cities in domestic services. 2/ Appendix Table 1.58 shows that, according to the 1970 Household Survey, the migrants into the eight of the main cities consistently have an educational profile below that of the natives. Appendix Table 1.54, however, shows that the educational gap with the natives of the recipient city rapidly diminishes. 3/ Appendix Table 1.57 shows that a 1better educational profile of the migrants to any of the eight citi,es studied is directly related to the size of the place of origin. Since, as was discussed above, the migrants tend to move from rural to small lirban centers and then to larger cities, then it is probable that the most educated migrate to the larger cities. 4/ Nonetheless, within the migrants, the largest share were employed in agriculture before migrating. (See Alers, J. Op. cit. and the 1965 Lima Migration Survey). 5/ Alers, J., also argues that the motives for migration are probably the search for opportunities rather than the absolute poverty in the place of origin. Bradfield in a study of migration from Huaylas in 1967 found that migrants ranked very low (eighth place) the lack of land as a reason for migrating. (See Bradfield, S., "Selectivity in Rural-Urban Migration: The Case of Huaylas, Peru," in Bruner E. and Southall, A. (eds.) Urban Anthropology, Chicago 1968.) _ 92 - Table 29: AGZ STmRUCT 0Y )lG1ARI lo siml CcTnIX COMPlARMD TO TOTAL POPUL4TION, 1970 Axe GroulDs 15-214 0-14 1524 24 an over Total 15 and over 15 mnd over x 100 TOW poPulation 44.7 18.8 36.5 100.0 55.3 34.0 Iiginits to: LIM 43.9 42.5 13.6 100.0 56.1 75.8 Trujillo 45.1 35.0 19.9 100.0 54.9 63.8 Chiclao 57.6 16.3 26.1 100.0 42.4 38.4 'piura 48.2 33.0 18.8 100.0 51.8 63.T Arequipa 51.3 36.5 12.2 100.0 48.7 75.4 Cuzeo 54.3 35.'5 10.2 100.0 145.7 77.7 PuRo 47.6 34.9 17.5 1oo.o 52.4 66.6 Iquitos 56.T 35.0 8.3 100.0 43.3 80.8 Soureo: AnpendifTable 1.56 Tab' 30: WIChTMoAL PROFfLE OF NATM AND )CIWtAW POPULATION 14 Yu= MND OLDEt (in Percentage) Nbacational Lvel Total Itive lwaat Total 100 100 100 go advA tion Nr 1 7 -Pri education } 18 45 52 SeoCdary educoatin / 13 22 25 igtor eduetion / 5 5 6 I Iudos those wo have not couplaeta level. sowua: ertfunes, 31. O hodo Rwual al Per*", CEPD, June, 19T3. p. 1149. - 93 - b. Regional Disparities and Causes of Urbanization and Migration 57. In very general terms, this process of rapid migration and urbani- zation has been the inevitable result oi- the modernization of the economy. The process can be seen as a chain reaction engendered by economic and political factors. As reviewed in Chapter III, up to the Second World War, the prime movers of the Peruvian economy were primary export products whose production was located in the Costa, with the exception of some of the minerals. Even in the case of the Sierra's mineral product:ion, Sierra, the main beneficiaries were the ports (through which the minerals had to be exported and where the owners and managers of the mining companies lived) and the Government. The towns that grew around the Sierra's minLng centers soon stagnated, since mining had few linkages with the rest ore the local economy and employed relatively few people. 1/ Agricultural investment concentrated on the river valleys of the Costa; these offered not only better soils and a more predictable climate than in the Sierra, but also much easier communications to the export markets and to the larger domestic urban markets. 2/ Pari passu with this shift in economic activity, the structure of the labor force changed propelling the migrations to the Costa. The gradual concentration of people and economic activity in the coastal urban areas also reinforced the concen- tration of political power there. 3/ This in turn biased government policies in favor of urban areas. The latter not only absorbed the bulk of government investments and services, but benefited from favorable terms of trade with rural areas. 4/ This closed the circle and stimulated the acceleration of the spatial shift by increasing further the disparities between the Sierra and the Costa and between urban and rural areas. In the post-Second World War period, the process of rapid industrialization wihich continues today began, financed mainly with the earnings of the coastal exports and supported by the buoyant incomes of the rapidly growing coastal uirban areas. The accelerating indus- trialization, the rapid rise in economic and social expectations of the Sierra rural population, and improved communications within the country all contri- buted to turn the migration process into an ever-growing wave. 1/ The towns that developed around the mining centers grow very rapidly at the beginning (when a new mine is completed) but then stagnate and do not surpass the 20,000-50,000 range. An example is La Oroya, the most im- portant mining refinery center in the country. Its annual population growth fell from almost 3% during 1940-61 to less than 0.5% during 1961-72 and has a population of 25,000. 2/ The shift of agricultural emphasis from the Sierra to the Costa has been occurring since the Colonial era, but accelerated with the beginning of large-scale sugar and cotton production in the last century. 3/ Until recently, voters in national elections had to be able to read and write. The center of political power has been in the Costa, particularly in Lima since the beginnings of the Colonial period. (See Chapter III); the modernization and urbanization process has reinforced it. 4/ Only in the last fifteen years has the Government tried to reverse this bias. (See Chapters III and IV.) - 94- 58. Within this general process, the cities that grew the fastest were those that combined more economic and topographic advantages, although historical and political factors also influenced the extreme concentration of growth and economic activity in Lima. In the Costa two fundamentally different kinds of urban areas have emerged. A series of dynamic cities developed in "natural locations" where water, land with agricultural potential and port facilities were available, allowing for economic growth and urbanization. For example, the site for Lima was chosen because of its river, its location at the foot of a pass to the Sierra and the natural harbor formed by the headland at Callao. Trujillo and Piura developed into two of the best agricultural areas of the country (producing sugar and cotton), made possible by the flow of rivers. 1/ More recently Chimbote has flourished with the fishmeal boom based on its excellent natural harbor and adequate water supply nearby. 2/ In all these cities, growth emerged out of export orientation of their economies (in which ports are crucial), and was later stimulated by the industrialization process financed out of the export proceeds and later on by the demand for services generated by the city's own development. In con- trast, a series of other cities developed around one economic activity, but with poor complementary factors (these locations might be termed "artificial"). These cities did not develop as dynamically as those in "natural locations." For example, Talara was built near the coastal petroleum fields as a conven- ient place to have a refinery. However, short in water and with very limited agricultural land, the city has not branched out into other economic activities and has hardly grown in recent years. 3/ 59. There are several reasons why Lima has attained such a primacy, although the most important one probably has been the extreme concentration there of all the Government's administrative apparatus since the Colonial period. After its foundation, Lima became the main trading and administration channel between Spain and the most important of its viceroyalties in America. All trade went through Lima, all taxes and export earnings that were not sent abroad went to Lima, and all Government and business decisions were! made at Lima. In the post-Colonial period this changed little. Local governments that have emerged since then are weak and have little or no independent tax revenues. Economic activities that developed in other areas of the Costa and in the Sierra are dependent on Lima for finance, inputs and decision-making. Government investment focussed on developing Lima and connecting it with the main mining and agricultural centers. Only in the last 15 years have important 1/ The Moche River and the Chicama River in Trujillo; the Piura River in Piura. 2/ The establishment of the steel industry at Chimbote has also stimulated its development. 3/ Another example of "artificial location" is Mollendo, the port to Arequipa, which probably for similar reasons has not grown much in recene years. In the 1961-72 period, Lima, Trujillo and Piura had average yearly growth rates in excess of 5.5%, while Talara and Mollendo grew by only 0.6 and 3% respectively. - 95 - efforts been made by the Government to spread out investment into other areas of the country, although not decision-making. The large and complex controls which the Government has increasingly i.mposed on economic activity require approval from Lima, even for simple procedures. 1/ Within this context, what is surprising is that the massive exodus to the capital did not start earlier, and in larger proportions. 2/ 60. The economic and social disparities between regions and between rural and urban areas that have resulted fromi this process are staggering and are illustrated by the following facts: (a) The main coastal cities account for well over 80% of indus- trial activity and commercia. bank credit the bulk of it concentrated at Lima (Table 31). 3/ (b) In the Costa and in the urban areas output per worker is much higher and the share of the persons earning less than the minimum wage much lower than in other areas of the country. Agricultural output per worker is about one-fifth of the average of other activities in urban areas 4/ and the share of the labor force earning less than the minimum wage is 2.3 times lower (Tables 32 and 33). 5/ Moreover, the overaLl average for agriculture hides the large differences between the export-oriented farms of the Costa -- 1/ See Chapter VI for a discussion oil the characteristics and efficiency of Government controls. 2/ See Chapter III for more details on the role of Lima in Peruvian economic history. 3/ The only centers with some importance in industry in the Sierra are La Oroya, a metal-refining center,, and Arequipa. Although Arequipa is the second largest city in the country with a population about 1/10 that of Lima, its industrial production is equivalent to only about 1/20 that of Lima. See Chapter VIII for a iaore detailed discussion of the spatial distribution of industry. 4/ Table 32 also shows that the average annual GDP growth rate in agriculture is estimated at one-fifth that of manufacturing. Although the size of the agricultural labor force appears to have stagnated, the output/worker is growing faster in manufacturing. 5/ Underemployment is estimated by the Ministry of Labor as having worked less than 35 hours/week or having an income of less than the minimum wage. Table 23 also suggests that since underemployment in agriculture in urban areas was 38.9%, underemployment Ln rural areas must have been well above 65.4%. Moreover, since agricultural output has been growing slowly but probably faster than the rural laDor force, the underemployment problem there was probably worse before. - 96 - which employ less than one-tenth of the rural families but produce about two-fifths of the sector's output -- and the small farmers of the Sierra which account for almost 60% of the-rural families. 1/ (c) The regional differences in productivity and employment are reflected in sharp income disparities. Even after taking into account regional price differences, the average per capita expenditure of a person in the eighth decile (within the top 30%) of the rural Sierra is in real terms below that of a Limeno in the third decile (within the bottom 30%). 2/ A Limeno family earning the minimum legal wage has to work less than six days to have an income equivalent in real terms to the average expenditure of a rural Sierra family in the eighth decile. 3/ (d) The maps on the following pages show indicators of the large differences in literacy rates and availability of services between urban and rural areas and between Coastal and Sierra departments. Here again, the main contrasts are between the urban Costa and the rural Sierra. The literacy rate is 67% for the total population but only 33% for the rural women. 4/ 11 See Chapter VIII for a discussion of regional differences in agriculture. 2/ These are mission estimates based on the tabulations by the Ministry of Economy and Finance (MEF) of the data of the 1972 National Food Consump- tion Survey (known as ENCA). Based on information of price and volumes of food consumed, the mission estimated an inter-regional price index. Applying the index to the MEF estimates, income and expenditures by deciles gives the estimates in Table 34. Although the differences in incomes are more pronounced, expenditures are considered a more reliable indicator since there seemed to be considerable amount of under-reporting of income (particularly imputed income for auto consumption), particularly in the lower deciles of the rural Sierra. 3/ In other words, the income for persons in the richest third of the rural Sierra will in real terms be below that of someone in the poorest third decile of Lima (Table 34.) 4/ This assumes there are 1.9 workers per family in both Lima and the rural Sierra (the national average for the economically active population). Since the minimum wage in 1971-72 at Lima (deflated by the regional price index) was SI. 99.04/day, and since the estimated average montlly family expenditure in the eighth decile in the rural Sierra (deflated by the regional price index) was S/.581, it would therefore take a Li;neno family 5.87 days of work at the minimum wage to earn an equivalent income. (See Appendix Table 1.69). PERU: RELATIVE INCOME PER CAPITA AND LITERACY RATES BY DEPARTMENT RELATIVE INCOME PER CAPITA LITERACY RATES I 972J I 1 9 6 1 TOTAL PERU RURAL FEMALES (Nalional Averoge: i00.0%) (Nalfnol Avurge: 67.0%) (NoRonol Average: 32.5%) n.y'; ,7-4 i* t s~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~- +~~~~~~~~~~~~~~~~".1 8 5e - V7 8'11 70 .60 " Ibl .0 197.n % , ::: " 7~~~~~~~~~'?.46 -78.90 , 47.86 -61 .20 _, 10 .1 -0 Ibi).0 63.-'7fi 7 2.45 38.21 -4'7.85 ,'5 I_]6.0 li1 (lstW \,E:1S)31 63./15 E 30'01 -3-8.20_ {t ) 4b.0 65.-i! 3 4 34.40 50.30 1[]4-30o 30.00 -'/ NUMOERS CORRESPOND TO NAMES OF DEPARTMENTS U,e,,.,,..,.,n,,,.,., 1Pencenl ol popolalton 5 years ond older not able IAMAZONAS N CAJAMARCA Ol JONIN 1A MADRE DE DIOS 21 SAN MARTIN pb h.,p,o.J . ,;.; h.....,. ;, 8 Io read and -r,le 2ANCASH 7 CtIZCO 12 LA tIBETAD 1 7 MOOUEGUA 22 TACNA 5Jlr7Ida.II.,,k C boa,r-e. Appendro Tables I o3 and I .65 3 APURIMAC 6 HUANCAVELICA 13 LAMBATEOUF 18 PASCO 23 TUMBES 4 A REIJUPA 9: HiJANUCO 74 1 tMA 19 PIUIRA AVACUCNO 0 ICA IS LORETO 20 PUNO ~~~~~~~~~~~~~~~~~~~~~~~~~_L LOET _ _IUNO PERU: PERCENT OF POPULATION WITH ELECTRIC LIGHT BY DEPARTMENT, 1972 PERU TOTAL URBAN RURAL (N.i...nnI Ave-ag- 34,9%) (NoJional A.erage 57 2%) (Nolionol A.ov*.ge 2.9%) 9........ ........... . ........ ....AAM . A It211114 A MII( iE 11.............1 ~~.i... xa~~~~~~~~~~~~~~~~~~ AYCI(I IS VA l (141 )O 25¢¢9 151141)-iZS z . . '. '. '. -. '. -. '. '. '. '. '. '. '. '. '. '. '. '. '. '. '. '.'.'.' . . . . . . . . . . . . . , t;.'. '. '. '. '. '. '. '. '.'.'Ii.'''''''''''''' '~~~~~~. '. '. t S. . , , , , . . . . . .. . . . . . . . . . . . ~~~~~~...... \................... .. ......... r...... ,r..,.... \. . . . . . . . . . . . . . ''' -< . . . . . +_ _ \. 1 1 . . . . . . . . . . fi~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~2 . . . . . .' o1 . . j. ~~~~~~~~~~~~~I MONS6CJMRAI Ut16MNR ) ))S2 A ASN. . . . ,i.. .. . ., . . . . . h.1. .O Source Ap,elidix Trbe I /SS 2A(AS14 7CIIC0 121A IltRIAD 1/ U)OIIItilJA 22 lACtA hs......... .......I ....~~~~ ~ ~ ~ .........: .........A13tABALUt I A5( 2 U Bt .....4 .N(lJP 9.IAICO1..MA1 Lli __ .. .... .....C1( ....ISI(R [ 0l l( PERU: ACCESS TO WATER AND HEALTH FACILITIES BY DEPARTMENT, 1972 ACCESS TO HEALTH FACILITIES ACCESS TO WATER PERU TOTAL URBAN RURAL PERCENT Of POPULATION WITH PERCENAT O, PEPLTO IHPPDWT HOSPITAL BEDS PER 1000 PERSONS PIPED WATER OR PRIVATE WELL PRIVATE WELL OR COMMON STANDPIPE (Nalional Average; 1 94 beds per 1000 persons) (NatIonal Average: 55.5%) (National Average 9.2%) .. .' . .. .- ..... .. .. .. .- 30.1 . NUMr*anS CORRESPOND TO NAMES OF DEP'ARTMENTS _~~~~ a Su~~~~~~~~~~~~~~~~ eApnl .. . rl 16rn165............3APURIMAC S HUANCAvELuC. 13 lAMBAYEGUE ISM0 U PACN 23 h.TUM, ,,, ,BES ~1 D _ - 4 ARHGUIP R 14UANUCO 14 IAMA 19 PIURA ___________________ 5 AVACUC-l0 .10 ICA 1S REO 20 PU2NO__ - 100- Electric light and piped water reaches over one-half of urban households, but a negligible share in rural areas (particularly in the Sierra). 1/ The maps also show that there-are six departments in the Sierra (accounting for 22% of the rural population) that are the poorest and have the lowest number for practically all the social indicators (Cajamarca and Amazonas in the Northern Sierra, and Apurimac, Ayacucho, Huancavelica and Puno in the Southern Sierra). These departments are also among those with the highest net emigration rates and lowest degree of urbanization. 2/ 61. These differences in economic and social conditions are the main determinants of aigration. Of particular importance are the income differen- tials and the opportunities for better employment. Most household surveys give direct evidence of this: migrants rank economic reasons as the principal reason for moving. S/ It is difficult to determine whether urban "pull" factors are more important than rural "push" factors. However, as was said before, the migrants are not of the poorest strata in the place of origin and some surveys found that the emigrants gave a relatively low rank to "push" factors such as the lack of land. This may suggest that "pull" factors of better job and income opportunities are probably more important than the "push" factors of abject poverty. There is also some evidence that long distance movements respond to more long-range planning by the migrant, while shorter distance movements are more sensitive to more immediate or sudden changes in their economic situation. 4/ The "push" factors may be more crucial for the latter group of predominantly rural to smaller town migration. The recent changes in the direction of incremental migration, suggest that migrants are sensitive to the shifts in the labor market. 5/ I/ See adjacent Maps. There are 7 Sierra departments accounting for 40% of the rural population, where less than 30% of the rural women are literate. 2/ See Map after para. 50. 3/ See Maps and Appendix Table 1.52. 4/ See Alers, J., op. cit. and Martinez, H., El Exodo Rural en el Peru, (working draft), CEPD, 1973. 5/ There are three important aspects in this respect: (1) in recent years Government investment has decentralized outside of Lima and a larger amount of industrial investment is going to cities such as Trujillo; (2) at Chimbote the fishmeal boom and the steel plant expansion in the 1960s; (3) the differential of minimum wages between urban areas ias somewhat fallen in recent years; and (4) as suggested by the table below for 1970 the rates of open unemployment in secondary and tertiary activ- ities may be higher in Lima than in other major cities: . lao )°QJ1 .100.0 100.0° 1 . ° ! 1 ,100 1Q° V 6 Pril.ary activities 27.3 6.9 17.1 5.5 70.9 48.9 -3.1 8.6 28.0 13.5 5.9 33.3 Agr.cultore 21.6 -1.0 7. a 5.4 ,4JI 35.4 -.9 11.0 1U.2 5d 6. t) - 4 ricetr:ea 0.7 1.6 2.4 -1.2 -0.4 14.3 -4.6 1.1 4.1 1.9 -0.9 1.1 YKln,cS 5.0 6.3 7.0 1.3 17.0 -0.8 5.4 -3.5 5.7 5.7 0.8 3.B Seconidary acttvitiaa 21.7 19.5 30.1 27.8 74.6 3,3 7.9 64.2 22.6 23.2 31.9 2b.2 Manuf4cturiGg 8.2 14.5 25.1 21.7 44.9 29.B 18.7 45.8 20.1 21.4 24.5 23.7 CoastructiGn 13.6 5.0 5.0 6.0 29.8 -26.5 -10.8 18.4 2.5 1.8 7.4 4.5 Tetriary activitles 51.3 76. 1 52.3 66.7 -45.5 46.2 95.2 _ 27.Ž3 _.49.4 63.3 62.2 54.3 _/ ._:i : . IM.f G(iP ero-jtb wita; o' or over ,' PrcXr.t-Ee ctr,.ctura of increase ir (lP :tioe "T 41.0 15.2 14.8 8.8 9.8 3.5 8.4 3.9 14.9 11.9 5.3 7.- Oefizd A y.arz wtn G rotaw 5% or ov'r. __ CGO eq4ua to CMP plua iAe" s=dLn& frac O c 3hawa Lo tarsa of trad. ' .-ert (craacity to L:ret) calzulatsd by d±yi4±ez e .awta *t curent p.±oeo by tb. £mot pria. iAdez. Z .-'o4 *Sefoct of soimm 4 torp of trade. Zxcluas effects of pins fr- ter" of t4er. iare: APP x lale 2.1 A 1 - 168 - grow by over 4% annually between 1950 and 1959. However, the expansion of the capacity to import still fell short of the increase in imports and a small resource gap resulted. 1/ Most of it was financed with Drivate foreign capital that flowed into the mining sector and to a lesser extent into the agricultural sector. However, excluding the losses from terms of trade, the economy achieved a resource surplus over the decade since gross national savings grew faster than investment. 2/ 142. Starting in the early 1960s the trends were reversed. The growth of export volume slackened and gross national savings (excluding terms of trade gains) started a long decline that still continued in 1975-76. 3/ During the 1960s a rapid growth of export volume combined with a sharp improve- ment of terms of trade, propelled import capacity by almost 9% yearly, somewhat faster than domestic demand. 4/ The crucial role played by the terms of trade gains is demonstrated by the fact that without them gross national savings would have in absolute terms (at constant 1963 prices) stagnated between 1959 and 1969, while investment increased 60%. 5/ Between 1969 and 1976 terms of trade marginally improved, but export volume dropped by 10%, resulting in a 1/ Although in 1959 resource use was practically identical to GDP (measured at current prices), over the decade as a whole the resource gap was around 2.6% of GDP (Table 47). 2/ Measured in constant prices and excluding the effect of terms of trade, in the 1951-59 period, GDI grew by 3.2% annually and GNS grew by 4.2% annually (Table 47). 3/ The average annual growth of export volume was 10% in the 1951-61 period and 3.0% in the 1962-69 period, between 1969 and 1976 export volume declined at an annual rate of 1.5%. 4/ Export volume increased at an average annual rate of around 6% during the sixties mainly because of a jump in mining exports in 1960-61 (result- ing from the completion of the Toquepala mining project) and a tripling of fishmeal exports between 1960 and 1969. The volume of agricultural and petroleum exports, on the other hand declined substantially (see Appendix Table 3.17). In the 1959-69 period the terms of trade index (1963=100) rose from 96 to 128 (Appendix Table 3.35). Between 1959 and 1969 the annual average growth rate of GDY was 6.5% while that of domestic demand (consumption plus investment) was 6.3%. 5/ If the gains from terms of trade are excluded, GNS in real terms (at 1963 prices and exchange rates) were $387 million in 1959 and $383 million in 1969, while GDI rose from $392 million in 1959 to $629 million in 1969 (Appendix Table 2.25). - 169 - 6% contraction of import capacity. l/ Domestic demand, however, grew by over 7% anually, stimulated mainly by the exogenous increase in public investment and the buoyant expansion of consumption demand, partly caused by the redis- tributive measures of the Government in favor of the middle class. 2/ The large resource gap that ensued was financed mainly with massive foreign borrowing by the public sector. 3/ c. The Role of Exports in the Economv 143. Another important macroeconomic feature of the 1950-76 period is that economic activity remained closely linked to the behavior of exports, although the relationship has been changing over time. The link is not direct between exports and output, since exports directly account for a smaller share of GDP in the 1970s than what they did in the 1950s; most of the incremental output has been of domestically oriented activities. 4/ Nor is the relation- ship a result of a larger forward or backward integration of the export sectors with the rest of the economy. 5/ With the possible exception of fishmeal and tourism, 6/ the export sectors import most of their non-labor inputs, and remain concentrated on the production of primary commodities and raw 1/ The terms of trade index (1963=100) rose from 128 in 1969 to 143 in 1974 but declined to 134 by 1976 (Appendix Table 3.35). The main reasons for the fall in export volume were the collapse of fishmeal production after 1971-72 (fishmeal export volume in 1976 was still 1/3 that of 1972) and a 17% contraction in the volume of agricultural exports (mainly because of a sharp fall in cotton production) (Appendix Table 3.17). 2/ See Chapter VI for a review of public sector policies in the 1968-76 period. 3/ The GNS-GDI balance (measured at current prices) changed from +3.1% of GNP in 1970 to -7.8% in 1976 (Table 2.15). 4/ Table 47 shows that the share of exports in GDP averaged 20% dur:.ng the 1950s and 14.5 during the 1970-76 period. 5/ As discussed in Chapter III, the little integration of export production with the rest of production has been a traditional characteristic of the Peruvian economy. By emphasizing primary commodities as the main thrust of exports and orienting the manufacturing sector almost exclusively towards the domestic economy, the Government policies have not been conducive towards a more rapid integration of the productive syscem. 6/ The fishmeal industry has developed important backward linkages with the rest of the economv and has been instrumental in the development of various industries such as boat construction and nylon net producl:ion (Chapter I). - 170 - materials. 1/ The link is financial. The economy remains heavily dependent on the foreign exchange and the savings generated by the export sectors. The agricultural sector imports most of its inputs and investment goods, and out- side of the export sectors generates little savings. 2/ Moreover, the economy has become increasingly dependent on imported foodstuffs. 3/ The manufacturing sector remains heavily dependent on imported goods and since it has been almost totally oriented towards the domestic market, it generates little foreign exchange. 4/ At the same time, since it has developed behind heavy protectionist policies, it remains inefficient and also dependent on government subsidies to finance its investment. At its turn, the Government depends on export activities for between 1/3 to 1/2 of its revenues. 5/ The economy's dependence on imported energy has also been mushrooming in the last decade. While petroleum reserves were declining, heaivy subsidies and price distortions created by price controls have not been conclucive to a rational and efficient use of the scarce energy resources. 6/ Fin&lly, since personal savings and public sector savings have also become negligible, it is the export sectors that provide the bulk of gross national savings. One conclusion is therefore that the fall of exports as a share of output does not mark a decline in export dependence on the economy, but rather that the exports have not been growing fast enough to sustain the growth in. other sectors. d. Cyclical Fluctuations 144. This close link of export earnings with the rest of economic activity and the heavy concentration of exports in primary commodities results in another of the important characteristics of Peru's macroeconomic developments: the sharp cyclical fluctuations. Tables 46 and 47 show that in the last quarter century Peru has gone through four full cycles, with annual GDY growth 1/ Primary commodities and raw materials in 1976 still account for over 90% of merchandise exports (Appendix Table 3.20). 2/ However, after the agrarian reform even the export products (sugar, cotton and coffee) are generating foreign exchange but little savings (Chapter VIII). 3/ Although the exact figure is not known, the mission estimates that imports of food and other inputs for the agricultural sector rose from $150 million in 1970 to $400 million in 1975. Since agricultural exports in the latter year were about $394 million (and in volume terms have been declining since 1960), the agricultural sector is becoming a net user of foreign exchange (Appendix Tables 3.3 a:-d 3.29). 4/ See Chapter X for more details on the imnport dependence of the manufac- turing sector. 5/ See Chapter VII. 6/ See Chapter I. - 171 - averaging over 7% in the high growth periods, 1/ and under 3.7% in the low growth periods. The general pattern of the cycles is roughly the following. Coming out of a recession when the import capacity booms (because of higher export prices or higher export volume or both), the incomes and savings generated stimulate the growth of manufacturing and tertiary activities, accelerate GDY growth and boost Government revenues. Investment reacts forcefully about one year later, delayed principally by the lag in preparation of Government projects and the spare capacity available in the productive sectors. Then, when export earnings collapse, the country's financial situa- tion deteriorates. However, partly because of the bulkiness of many Df the investment projects and the efforts of the Government to avoid the downturn of economic activity through expansionary policies, total domestic demand keeps growing and the resource gap widens financed with foreign savings. 2/ Imports react closely with GDY in the periods of expansion, but the slowdown in their growth follows the same one-year lag as domestic demand. 'However, the efforts to maintain output growth momentum when export earnings fall, only postpone the need for, and increase the magnitude of, strong contract:onary measures to restore financial equilibrium. 3/ I/ High growth was defined as any year in which GDP growth was over 5 percent. High growth was achieved during 16 years and low growth during 1C0 years. 2/ The table below illustrates how GDI maintains a high growth rate in the first year of the low growth period but then falls sharply. (The numbers in the table are annual growth rates): Year GCP CDT Imports 1952 2.7 22.8 11.6 1953 2.2 -0.2 '2.2 1957 0.7 9.2 14.0 1958 3.3 -14.4 -14.0 1967 4.b 2.9 1.3 i968 9.9 -29.0 -12.3 1975 2.' 1&.0 '7.5 1976 2.4 -13.0 -t4.1 Source: Apitendix Table 2.3 3/ The table below shows that by fueling domestic demand, the resource gap continues widening through the first year of the low growth period. The table also shows that there is usually a one-year lag between the fall in growth of GDY and the contraction in the volume of imports;. Percabt of '? (at current prices) 1/ ' 11 '/ I' 1951 19524 953 1956 1957 1958 196 1967, 196S 19,4 :975 1O'S ZS-G,DI 0.1 -3.3 -4.8 -6-A -7.9 -7. -4.6 -5.8 -s -6.3 -a.- -7.S 1/ Fi:st year of lov growch pe:'od. Sourcv. Appendix Table 2.' - 172 - 145. However, within this general framework, the specific characteristics of the cycles have been changing over time aainly because of the larger role of the public sector in the economy. In the 1950s, the timing of the fluctua- tion of import capacity and economic activit:y were more closely linked. The bulk of investment was by the private sector and most of the financing of the resource gaps came from private foreign direct investment. Starting in the mid-1960s, however, public consumption and :investment became important exoge- nous factors that lengthened the periods of expansion. But by the 1960-66 period, although import capacity expanded rapidly, the Government's expansion- ary policies fueled resource use to grow faster than GDY and financed the resource gap with heavy foreign borrowing. Then in 1967, although import capacity stagnated and gross national savings fell by 12%, the Government was able to maintain a GDP growth of 4% with continued deficit financing. However, as the investment-savings gap (at current prices) expanded to almost 6% of GNP, the financial situation collapsed in 1968. In the 1970s, the public sector stepped up the use of foreign borrow:.ng to counteract the fall of export earnings and of national savings. Following the recession of 1967-69, GDP grew vigorously in 1970 mainly because of increases in export volume and prices. In the 1971-75 period both import capacity and gross national savings contracted 25%, but the Government stimulated the economy with massive consumption subsidies, the income redistribution measures and 25% average yearly increase of public investment. The requirements of foreign savings to finance the resource gap that resulted rose from 0.4 to 11.4% of GNP between 1971 and 1975 and the external public debt outstanding and disbursed more than tripled. In the second half of 1976, the Government was finally forced to apply sharp contractionary measures. 1/ 146. A review of how the cycles in resource availability are reflected on the supply side provides some clues on what sectors suffer most in periods of recession. 2/ Through the effect on export earnings, primary activities are important determinants of the cycles. The change in export prices affects incomes of producers in primary activities but much less volume. Their changes in volume seem to reflect more long-term conditions (created by market prospects and investment) rather than short-term fluctuations. The erratic growth rates of the agricultural sector do not move in line with cycles, but are more influenced by natural conditions 3/ and the long-term gradual decapi- talization of the sector. 4/ The behavior of fisheries is almost completely 1/ The widening of the resource gap during the first year of the low growth period has been getting larger since the 1960s, and was particularly large in the 1975 recession. While the growth of output slowed down markedly in 1975, the impact of expansionary policies almost doubled the size of the resource gap (relative to GNP) which in turn now requires a much more draconian stabilization effort. 2/ See Table 46. 3/ Mainly weather. 4/ As is shown on Table 46, the average growth of the agricultural sector since 1960 has been below 3% in both the high and the low growth periods. - 173 - determined by the availability of anchovies and the growth rates follow a random pattern. Similarly, mining growth comes in spurts following the comple- tion of large projects; the adjustment to short-term changes in worlc market conditions comes more through prices and value than through volume. 1/ The manufacturing sector, on the other hand, has been able to maintain remarkedly high growth rates even in periods of recession. 2/ The main adjustment during the cycles are made mainly through violent fluctuations in the growth rates of the construction sector and tertiary activities. 3/ In periods of forced stabilization these are the sectors that absorb a large share of the contraction. Construction fluctuates with investment which, as seen above, follows the cycles closely. The tertiary activities fluctuate mainly because they are the resi- dual sector in the economy. In fact a large share of the output classified in the tertiary sector are artisans, informal small-scale industry and the self- employed. Various studies have shown that the self-employed have close links with the formal industrial sector. Because of the institutional rigidities of the labor, the larger industrial firms satisfy the peaks of demands by sub-contracting production to artisans. When demand collapses these are, of course, the first to suffer. This link between the formal and informal sectors (where most of the urban poor are employed) is an important reason for trying to minimize the violent fluctuations of the economy. 1/ Moderate contractions of volume caused by world demand factors occured only in 1975. However, even then an equally important cuestion was the supply constraint created by labor unrest. The labor force fluctuates little in the mining sector. 2/ Except in the 1967-69 period when the average growth rate fell to 2.9% annually (see table 46). 3/ The table below shows that (except in 1953) the construction sector contracts during the first or second year of the recession. ?ercect Zrowth race during year ll1 l/ 'J 1/ 9S5' i953 :937 19<3 !'67 19? 9 ' ??5 S;?' Tartiary' activities 2/ -1.9 -3.0 -2.7 5.1 0.8 3.1 4.4 2.2 (of which other services ) (-4.6) (-5.1) (-4.8) (5.9) -3.4) (3.2) (4.5) (r c-.) Conatruction 17.6 7.5 2.7 -7.7 -10.6 -16_. 2C.0 2.' 1/ ctrsc year of la- SrOwCh perioL. 2/ Tertiary ac-'vitces minus puol4c utli.ties, ownershi? of dellin~; and 7'Oer-.ozz. Saurce: Appendix Tzbla 2.16 The table also shows that the growth rate of tertiary activities (and particularly non-Government services) also falls or become negative during the recessionary periods. (Also see Table 46.) - 174 - 147. In sum, in spite of the moderately high growth performance, the macroeconomic trends of the past 15 years are not tenable in the long run. The principal structural questions are: (i) how can the resource gap be reduced without leading to stagnation -- through less consumption or through less investment with more efficiency? (ii) how can the vulnerability of the economy to violent cyclical fluctuations be reduced? The remaining of this Chapter discusses the principal issues related to answering these questions. B. INVESTMENT AND SAVINGS 148. Graph III shows the trends during 1950-75 of the investment and savings ratios to GNP. During the 1950s, the savings ratio fluctuated between 18 and 23% of GNP. However, these fluctuations were short-term movements around a relatively high horizontal trend. The structural change occurred after 1961. The savings ratio continued its fluctuations, but this time around a declining trend. Regression analysis suggests that the long-term marginal national propensity to save averaged around 11% of GNP for the 1950-74 period. 1/ The decline in the national savings effort was accompanied by a 1/ The trends shown in Graph III were more rigorously tested through regression analysis relating GNS to GNP. Technical Note I shows that of the various specifications tried, the ones that performed the best were: .~~ . ___---i--- ails - 7i;t60, + o.j.o6 cG!r - )0;iJA. i - (6.317) (28.?39S) (3.1(57) - * 230.0 j 1,6.505 + o.003 (;'f/P - 0.552 (2 Mc - 3.7 DUM- } O02L C O . G900 i (7.(03)81 (3.9Mo) (1.623) (1.932 ) wher tX1ru =rcal CTM pr fcopiFrt | G- = GNP deflaLor (}1D _) = expectcd ra'Le of inflation mea:7o s a si.X-Year TOV'I.f of th Srt of ch axg cf the GT11P dce f'ator In t ctUr:runt a-:1O1 the past f ,iv, years. DUill. dwnuiy r. a'nle c0,1_l' to 1 for 1°. -;.0O audl 0 for 1961-7;- SsE Y. = standard error of cZti .:t&- ofC nW Darbin ',: "non statisLi'C, li"gures in5M etle.s'e t thle I coefficie.^ts are th.c t - _tatistic s. c Soorce; Tec a.nral iT,ot I Tlhe regree Lot.;I:', SUo,gest tl~.e.l; a l'.n?'i.:' prom-p'-Jty to onve of boto,: C ' t' aVern;e for tlte 1950--! U' Cod. ^,e72-v;ce, eI' ..-e 'o sOCOl'nPr,-,; t;e .:o..:. -. th?t a str a-torct-J. cP-'t ge' oclrrod afLo' l;' 0i 'fc *MCOO?(j OCrl:i:t i en m:ol' 'o.;}' the nuat .on:l S.lVil.S clfrt icded t.; '' c .i thc ri.- k 0:01 .' ,,er capita. The role of in? Lat:ollCarr CX;'4:ctal. i v 000 noL 1..tii. .: 3: y : :. -, alt.hou.3Ji the coeffiicient ofi the cxpcceLcd I 11iD;.] warin'3oi . al: t tI:! .V 4 (0 _ 175 - IIIj: .PERU: INVESTMSNT AND SAVINGS, 1950475 A. In Biliions of Soles at Current Prices so so 70 GROSS DOMESTIC INVESTMIENT 60 so 30 GROSS NATIONAL SAVINGS 10- EXTERNAL SAVINGS =io 1 1 I II I I .! I 1 I I I I I ! I I i I I I I I !0' I 1950 '55 '0 70 '75 B. Percent of GNP 26 14 GROSS DOMESTIC NV ESTMENT 24 22 20 16 14 ~~~~~~GROSS NATIONAL SAVINGS 12 10 '' a~~~~~~~~~~~ / ~~~ EXTSRtNAL SAVINGS 711itllst~~~~~~~ol , I !( I\ l I 1800 'Sb liG 'S 6570 '75 ' ource: Amm-h.dls T.bli, 2.11 16x {4k-X7231 - 176 - fall, although less pronounced and with some lag, in the investment effort. Its cyclical phases were coincidental with (but more pronounced than) those of national savings in the 1950s, mostly unrelated in the 1960s, and clearly in opposite phase during the 1970s. RegressiorL analysis also suggests that the marginal propensity to invest over the 1956-75 period was 22% of GNP, double the one estimated for national savings. 1/ Therefore, it appears: (i) that the long-run tendency for the investment ratio to fall was the result of the long-run constraint imposed by the declining national savings effort; and (ii) that autonomous investment booms, such as th.e one in 1972-75 fed by public investment, result primarily in sharp rises in external savings requirements. 2/ 149. Before venturing into generalizations on the causes for the long- term fall in the national savings effort, it is important to review the behavior of its major components. However, the analysis has to remain quali- tative and not go beyond a review of the factors that seem to have affected 1/ The regressions relating GDI to GDP that statistically performed best were: _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - - - - - - - -I. _T__ _ _ I _ _ _ _ i !i2 ;W ilc Of7r7, -ID= 3929.9+0.220 Gr-?124L.5 (d) -81.09 T:T.s'+ 5!5.9° Y7! 0.93; 3 ;71 1.9531 195C-75 I (C.,l:7) (17.1I4) (J..925)(P ) (1.251) (±.1)i 100= 41.01 - 0.0007 ( 1 25 C 5 (6.205) /i.62X) (2.394) wlhere ( __)Il (= exrected rate of ilnfation measuret1 a a si ix year movingl averit..z of the ratc of ch!x.,nc of consura.Tr pri cen in the cuirrent yCear and the past fiveyer. i i - l = (exrortv nrire 10!C in Soles (i or-t r.i ces) j DU'; = Ljfor 1956-63 and 0 foi- 19U'9-'5. = re2.- per capita GoP, where rD 2s p j 120i deflator 2oulr~e: i'cc:hLici'C Note 1 2/ Various factors contribute to the close. link between investment and external savings: (i) investment expenditures have a high imported component; (ii) excess demand resulting from the excessive investment effort (in relation to saving) tend to raise prices which may reduce some forms of national savings and, cou.pled with a fixed exchange rate, reduce the international competitiveness of Peruvian output (less exports and more imports). - 177 - more significantly the trends. A more detailed quantitative analysis of causal relationship is hazardous mainly because the components of savings are interdependent and because the data available are of poor quality. 1' 150. Graph IV shows the trends between 1950 and 1975 of the majcr compo- nents of national savings. The sharp difference in behavior before and after the early 1960s, that was found in aggregate national savings, is reflected in all components except business savings. During the 1950s, all components grew steadily in nominal terms, but as ratios to GNP they fluctuated around a horizontal trend. The personal savings effort was high at around 8-11% of GNP (practically at the same level as business savings), but its volatile variations (in opposite phase to government savings) were the principal source of instability for aggregate national savings. The ratio of business savings to GNP fluctuated much less, with only two relatively small downturns in 1952-53 and 1958 (both were periods of world recession during which Peruvian exports collapsed). Finally, government savings were relatively unimportant, but were consistently positive, even during the 1958 crisis. 151. In the early 1960s started the long-term decline that still continues today of the ratios to GNP of personal and government savings. Two closely interrelated factors seem to have been major contributors to this sharp change in trends: the uncertain political environment and the expansion of the role 1/ There are many pitfalls in attempting to explain the behavior of the components of national savings separately, and many researchers have avoided their individual analyses (for example, see Zeff, N. and Sato, K., "A Simultaneous Equation Model of Savings in Developing Countries," Journal of Political Economy, Vol. 83, No. 6, December 1975). One problem is the interdependence between components. In particular, policies that affect Government savings tend to inversely influence savings of the nongovernment sectors. For instance, increases in personal and business taxes normally tend to raise government savings at the expense of personal and business savings; increases in social security contributions tend to raise government savings of the social security system (classified in the business sector in Peru) at the expense of personal savings. (See Munnell, A., The Effect of Social Security on Personal Savings, Ballinger, Cambridge, Mass., 1974). It might be expected that the personal savings component would be most affected by government policies because of its smaller size. However, since changes in policies affecting government savings through mazay channels might be expected to take place simultaneously, this would tend to conceal the effect of government savings on private savings. Another source of difficulty is statistical. Not only is the qua.Lity data poor, particularly that on personal savings which is a residual of residuals, but transfer of activities from one sector to another (for example, the state enterprises are classified in the business sector) make difficult a long-run analysis of national savings co:aponents. - 17a - GRAPH lI PERU: COMPONENT5^ O1 GI'OSS NATIONAL 'AVI;CGS A. In Billion of Soles at Current Prices 405 - NON-GOVERNMENT SAVINGS 35 _ /E 30 - / 10 1_ / N BUSINESS SAVINGS 5 ~ ~ ~ ~ \ ! . /y P" r^ , GOVERNMENT SAVING. 1950 j53 '60 65 '70 7 B. Percent of E~NP 22 2 0L / NON-GGVEM'4E'-T SAVINGS 16L~ O, / " '/__ yPCRISONAL SAVINGS 9V,0 VG '' . /4-utu-l ,:;, ,* .:, ,:-lt Z.,!; wn,lel .'œ,,;. :r,7'1 - 179 - of the State in the economy. The emergence of the workers in the coastal agri- cultural complexes and in the modern urban industry as strong political groups, the turmoils in rural areas demanding a land reform, and a general desire towards basic changes in the economy resulted in both uncertainty in the private sector and a pressure on the Government towards a larger role in the economy and a wider provision of public services. 1/ Reacting to these pressures, the Government's expenditures and transfers to the public sector rose sharply. However, the Government was unsuccessful in raising sufficient resources and its savings collapsed arnd were negative during 1964-67. 2/ 152. Although the data on personal savings are particularly weak and there is little conclusive research in Peru or elsewere on the determinants of personal savings, various important hypotheses can be brought forward on how these two factors affected personal savings. 3/ The first nypothesis is that the rapid fall in real interest rates--caused by the fixed nominal rates at a time when the Government's expansionary policies were fueling inflation--in an uncertain environment may have induced either a shift from personal savings to consumption or (more certainly) a shift from financial assets to "inflation hedges" (real estate, commodities, foreign exchange, etc.). in both cases, the resources available to satisfy the investment demands would have shrunk. More- over, in the latter case, the financial repression reduced the efficiency of 1/ See Chapter III for details. 2/ See Chapter VI for details. 3/ The quality of the personal savings estimate is poor. In tne first place, it is calculated as a residual difference between personal disposable income and consumption, which itself is a residual in the GDP account. At the same time, Peruvian accounts do not impute any personal savings- investment on housing and durables. There is some evidence to strongly suggest that personal savings are underestimated. The ENCA 1971-72 household survey estimated that personal savings was about 9% of Fersonal income, a ffgure which seems reasonable. Moreover, ENCA's estima:e of personal income is less than half the estimate of national income accounts for 1971-72. There is no explanation for this discrepancy, but it is unlikely that such a large gap can be explained by differences in sectoral coverage. Besides statistical problems in Peru, the fact is that little is known anywhere on the determinants of personal savings, particularly household savings. Although it is possible to determine some basic relationships of personal savings to other economic aggregates (such as per capita income), there are examples of rising and declining trends throughout the world for which there are no established hypotheses (see Burkhard, S., "Saving Behavior in Western Germany and the United States," American Economic Review, Papers and Proceedings, Vol. 45, May 1975, pp. 210-216). - 180 - the financial intermediation system in the allocation of investable resources. 1/ The second hypothesis is that the Government s policy of subsidizing the prices of some consumer goods stimulated private consumption. The third hypothesis is that the movement of political thinking and of Government policies towards income and wealth redistribution affected personal savings in two ways. On the one hand, the "rentier" class of traditional landowners viewing the land reform as inevitable, not onily stopped investment, but decapitalized their haciendas and consumed the proceeds. 2/ On the other hand, the few who benefited from actual redistributed income flows, belonging practically all to the upper-middle income group, had a high propensity to consume. The relative importance of the vari.ous hypotheses formulated is not known, what is certain is that the Government: took no measures to stimulate personal savings. 153. During the 1970s, the deterioratiorn of government and personal savings accentuated. The series of causal faLctors was probably the same. Shari growth of government expenditures with no concomitant increase in revenues, resulted in negative government savings and stronger inflationary pressures. Uncertainty became larger and thE real interest rate on financial assets more negative. Consumption subsidies expanded and the share of the population who benefited from the redistributive measures was much larger-- although practially all were of the upper miedle income groups. 154. The pattern of business savings throughout the 1960s was different and remained principally determined by the cyclical fluctuations of general economic activity. During 1960-66, the boom in export volume of minerals and fishmeal and a fast recovery of terms of trac.e fueled business profits to expand vigorously (from 10-11% of GNP in 196C-61 to a peak of 16% in 1966, see Graph IV). The downturn of economic activity during 1967-69 caused a severe contraction in nominal terms of business savings. 3/ 1/ McKinnon in his book Money and Capital in Economic Development (The Brookings Institution, Washington, D.C., 1973) presents the theoretical basis for the hypothesis on the relation between negative real interest rates and savings. The disequilibrium situation, which effectively prevents the financial markets from satisfying the demand for investment is called the "conduit effect" by McKinnon. However, most empirical studies trying to relate savings or investment to inflation or financial repression ran into failure. There is much more evidence on the relation between financial markets and inflation (see for example Galbis, V., "Financial Intermediation and Economic Growth in LDCs: A Theoretical Approach," Journal of Development Studies, January 1977). 2/ There is little direct quantitative evidence in Peru proving or disprov- ing this hypothesis. Robertson (in "The Mobilization of Domestic Resources: Measurements and Analyses of Saving Performance and Behavior" mimeo, UNCTAD, New York, 1974) found indirect evidence in other countries on this pattern of behavior of "rentier" classes. 3/ See Tables 46 and 47. Industry, mining, and construction suffered particularly severely during 1967-69. - 181. - 155. The determinants of business savings, however, changed in the 1970s. Two interrelated factors were superimposed on the forces of the economic cycles: the State's expansion into commodity-producing activities and price controls. The proliferation of State-owned enterprises--through take-over of private companies and the creation of new State corporations--strained the Government's management capacity affecting their performance. However, the impact of price controls was probably more important. The State enterprises were forced to absorb heavy losses 1/ and the profits of private companies subject to price controls fell. Therefore, the business savings/GNP ratio rose somewhat in 1970-72 but since then has been declining, in spite ol the economic expansion that continued until 1974. Here again, therefore, a major cause of the deterioration of savings performance can be directly linked to government policies. 156. The analysis of the components converges on three complementary hypotheses on determinants of national savings in Peru during the last 15 years. The first is that the ultimate major determinants of the deterioration in the national savings effort were the transition towards a mixed economy and the strains created by change in ownership structures and in the socio- political constellation of power. The second hypothesis, corollary to che first one, is that in its zeal to do much in a short period of time anc to satisfy the demands of the new upper-middle-class, the Government did rot attach sufficient attention to the problem of resource mobilization; the negative effects of its policies were pervasive throughout all sectors. The third hypothesis is that in their zeal to help Peru achieve the socio- economic transformations, external lenders were willing, in the mid-1960s and since 1972, to finance a large share of Peru's growth, but in doing so, allowed the internal disequilibrium to become larger each year, climaxing in the present crisis. In turn, however, this financing of the inadequate national savings effort with external savings, increased enormously Peru's dependence on foreign lenders, contrary to the Government's desire to reduce foreign dependence. 157. Was this fall in national savings effort a necessary cost of naking important changes in the socio-economic and political structure of the country? An economist's answer to this question would be no. Chapters VI and VII discuss why many of the Government expenditures were not necessary to a:tain the objectives pursued, and why a larger effort at revenue generation, :ax reform and better pricing policies by state enterprises was not only feasible but would have helped promote personal and business savings and would have been more conducive than present policies to attain the development and redis- tributive objectives of the Government. The next section of this Chapter discusses why and how more resources could have been gathered by the financial system and more efficiency in the resource intermediation process would have been attained, if more realistic interest rate policies had been followed. On the other hand, the political realities may have made this larger savings effort impossible. However, in this case, the problem with the policies followed was to make a high level of investment an objective itself, while 1/ See Chapter VI for details on performance of public enterprises. - 182 - the realistic approach would have been to reduce investment to the national savings realities, and focus the attention on increasing massively the efficiency and productivity of the resources invested. This is probably one of the largest lessons for future policies to be extracted from macroeconomic developments in the last two decades. There is little chance that Peru will overcome its present financial difficulties and resume a path of accelerated development less dependent on foreign phenomena, until efficiency of resource use and the savings effort are drastically improved. The fundamental issue of efficiency and the type of measures that would help achieve it are analyzed sectorially in other Chapters of this report. C. FINANCIAL INTERMEDIATION, INFLATION ANEI INTEREST RATES a. The Financial Structure of the Economy 158. Appendix Tables 6.26 and 6.27 summiarize the results of a flow-of- funds study done in 1969-71 which is the best available to sketch the finan- cial structure of Peru's economy. 1/ The study suggests that the financial and business sectors are the major pillars of the financial structure, 2/ and that there is a large credit market functiorLing outside of the formal financial sector. 3/ The public sector is not well integrated in the national financial markets since it finances most of its deficits with foreign borrowing and with forced placement of bonds in the commercial banking system. 159. Total financial assets grew rapidly in the 1965-74 period, 4/ but Tables 48 and 49 suggest that there is an inverse correlation between rate of total real financial asset growth and the rate of inflation. Financial real growth averaged a low 2.3% during 1966-68, cL period of high inflation, but accelerated to over 7% in 1969, when strong stabilization measures lowered 1/ Ministerio de Economia y Finanzas, Prinmer Estudio de Fuentes y Usos de Fondos de la Economia Peruana, Periodo 1969-1971, Direccion General de Asuntos Ffnancieros, Lima, 1975. 2/ Enterprises and financial intermediaries originated 68% of the total sources and uses of funds according to the 1969-71 study (Appendix Tables 6.26 and 6.27). 3/ This is suggested by the fact that the size of the credit market is twice that of deposits. 4/ The average annual growth of real financial assets was about 8% between 1965 and 1974. - 183 - drastically the growth rate of prices. 1/ In 1970-72, the real growth of financial assets gathered more momentum, in response to the combined effects of less inflation and the 1970 measures forcing capital repatriation. 2/ Finally, since 1972 financial growth decelerated in response to renewed inflationary pressures. Table 49 also suggests that variations in financial assets of the nonfinancial sectors tend to follow a pattern similar to those of the financial sector. 3/ The close link between real financial growth and inflation was particularly marked because of the fixed interest rates policy that was followed. 160. The fast growth of total financial assets in real terms has been more a reflection of macroeconomic disequilibria than of financial progress. The main problem was that in its zeal to regulate the private sector, the Government's interventionist policies tended to increase the fragmentation of financial markets and thereby to reduce their efficiency. 4/ There are three concrete aspects of policy since 1970 that affect this retrogression of financial markets: 1/ The pattern of growth up to 1970 is broadly corroborated by the results of a study of flow of funds for the period 1965-70 by the National Supervisory Commission on Enterprises and Securities: ?inancial grovth Rate of change in in real terms 1Wr deflator 1966 2.7 12.5 1967 h.0 12.2 1968 0.5 17.9 1969 7.2 7.8 1970 17.5 7.4 lkte of cbazge of total financial assets deflated by thc GNP deflator Souree: Comision Nacional Sup>crisora de Zrpresns y Ytalores, ?1ujo de Fondos Yinancieros cn el Pcru.. Per.odo 19C5-1970, Liz&, 1973, Tomo I. p. 18. 2/ The capital repatriation measures of 1970 banned the holding financial assets abroad or in foreign currency at home. 3/ This evidence tends to contradict the hypothesis brought forward by Leff that financial flows in the nonfinancial sectors tend to offset varia- tions in those of the financial sector, thereby neutralizing the effects of some monetary policy actions (see Leff, N.H., "Capital Markets in Less Developed Countries: The Group Principle," Chapter 5 in Money and Finance in Economic Growth and Development: Essavs in Honor of Edward S. Shaw, edited by R.I. McKinnon, Mercel Dekker, Inc., 1976). 4/ For a discussion of these problems in developing countries in general, see McKinnon, op. cit., and Shaw, E.S., Financial Deepening in Economic Development, Oxford University Press, New York, 1973. Tubli1 48: Guonli oIF iTrAL YISA24CIAL A17E0M. IN TIIE AINUthY, 19;bb-74 Finanicial grNth Rate of chang,e in in real terms GDP deflator 1966 5.1 11.3 1967 1.9 12.3 1968 -0.2 18.2 1969 T,3 8.0 1970 13.6 7.3 1971 16.0 4.5 1972 13.9 5.3 1973 6.2 14.6 197k 9.1 18.2 1/ Rate of chanee of financial assets deflAted by the CDP deflator. Source: Appendix Table 6.13 Table I 9: FINAMCIAL ASSETS OF -lHE ECCNCMY, i966-74 Average Annual Real Ceowth December Rotes 1965 1i968 1972 1974 1966-66 1969-72 1973-74 A. TOTAL FINAN4CIAL ASSETS OF ECO1OMY' z of GDP Total 71 69 90 93 3.0 12.8 9.8 Private Sector 45 47 65 65 5.0 14,8 8.1 Public Sectov 7 6 11 iS 0.7 20.1 23.0 Foreip Sectcr 19 16 14 14 -1.3 2.2 6.8 Fiaancr l Sector (indirect finance) 35 33 46 50 1.8 14.2 12.6 Other tdirect finance) 36 36 44 43 3.9 11.4 6.8 B. STRUCtURE OF FIXtYCIAL ASSETS OF PRIV'ATE SFCTOR S of Total Total 100 100 100 100 100 5.0 14.S 8.1 Ir Type of Asset Honey 25 22. 23 23 0.2 10.3 8.7 Time and savings deposits 18 14 13 11 -4.2 8.4 -0.8 Hort;age bacds 2 3 6 6 19.9 29.2 8.1 roreigm currency deposits 6 5 - - -0.2 -44.9 -2.1 Shares 45 48 47 49 7.1 10.0 10.9 Bonds 3 5 3 7 25.7 27.2 3.3 Other 1 3 3 4 44.5 14.8 8.9 3y Institution OriginatinR Asset Financial System 56 49 48 46 0.5 14.0 5.5 Zanking Systen (51) (42) (39) (37) (-1.4) (12.7) (5.6) Other (5) (7) (9) (9) (17.2) (20.8) (5.2) Public Sector 1 3 6 6 40.3 41.8 5.4 of which. InveCtment bonds (-) (-) (1) (2) ( - ) ( - ) (44.5) Agrarian debt bonds (1) (4) (3) ( 2 ) (82.9) (-7.6) Private Sector 42 48 46 48 9.2 13.6 11.2 of which: Shares (40) (45) (44) (47) (9.7) (1'.I) (1.) Forefrn Sector 2 - - - -]. t Source: Appendix Table 6.13. - 185 - (a) The financing of the massive public sector deficits of the 1970s has been an important element in the fast real financial growth. These deficits offered an opportunity for development of a domestic financial market for public debt instruments. For this, the following conditions ought to have been met: adequately high interest rates, development of market dealers and maintenance of public confidence. Unfortunately, no measures were undertaken to meet any of them, thereby missing the opportunity for savings absorption, and possibly encourage- ment of private savings. This left the public sector with only two financial avenues: to pursue an aggressive foreign borrow- ing policy, 1/ and to force the commercial banking system to acquire public debt. 2/ As regards the latter point, the substantial expansion of Public Investment Bonds, that did take place (Table 49) was not a sign of development of financial markets. On the contrary, it was both a reflection of the disequilibrium of the public sector and a contributing factor to foster such disequilibrium. Furthermore, forced placement of bonds contributed to lowering the efficiency of commercial bank credit markets. (b) In regard to the external sector, although the capital repatriation measures of 1970 achieved their immediate objective of increasing significantly the level of international reserves, their long-term impact was negative. 3/ The initial abundance of reserves encouraged the authorities to embark on a program of excessive investment and foreign borrowing, while the measures 1/ Paradoxically, in many cases a higher interest was paid on foreign debt than what the Government was ready to pay on the domestic debt. 2/ See Appendix Table 6.22 for details on the selective credit system imposed on commercial banks. 3/ It should be noted that the capital repatriation measures did not auto- matically alter the volume of external savings, because the net capital inflow resulting from those measures was offset by the increase in inter- national reserves. For this reason, no increase in external savings was observed in the initial years. Thus, the only direct effect of those measures was of accelerating the rate of financial assets growth. How- ever, what is contended here is that their effects were transmitted to the real side by permitting the expansion of external savings in the medium term through excessive public borrowing from abroad. To see how absence of those measures might have altered the course of economiz development, note than an economy starting with a low level of interna- tional reserves would have had more difficulty in attracting the massive volume of credit from outside that propelled it into unsustainable expan- sion in the long run. - 186 - themselves contributed to segmenting the domestic and foreign markets. As in the case of the public sector finances, the success of a policy of liberalization in this area would also hinge crucially on raising domestic interest rates to appropriate levels and on maintaining public confidence. (c) Several aspects of structural reforms of private enterprises worked directly against the development of financial markets. A portion of the market for corporate stock became blocked as a result of the requirement that 15 percent of annual profits be deposited in a fund that could not be sold by the community of workers nor used by individual workers as personal property. 1/ The same conclusion applied with greater fZorce to the agricultural cooperatives, whose shares were effectively pulled out of finan- cial markets by the requirement that members might not sell shares to non-members or among themselves. This has introduced serious rigidities to the process of "direct" financial intermediation. 161. The financial sector--whose volume of financial assets equals that of the rest of the economy and has edged marginally up in the last decade-- plays a crucial role in channeling financial resources in Peru (Table 49). Within the financial sector, banks are by far the most predominant institu- tions, with around 80 percent of total financial assets, but the rest of the financial sector has tended to expand its share steadily since 1965. Major institutions are commercial banks, state development banks, the National Bank and COFIDE. Savings and loan associations and cooperatives are the most important quantitatively in the private non-bank field. 2/ Chart I sketches the structure of the financial system at the end of 1976. 162. Since 1968, several institutional changes have been made in the finan- cial system in pursuit of securing greater national control over credit sources. 1/ These regulations have recently been liberalized to permit individual workers to dispose of their shares. 2/ The three main objectives of the financial sector reform were: (i) to promote savings and investment; (ii) to create selective credit channels in order to increase production in "strategic sectors" to improve the "democratization" and regional spread of credit, and to support the beneficiaries of the agrarian reform; and (iii) to secure national control of credit sources and institutions. For more details on the objectives of the financial reform, see de la Melena, G., La Reforma Financiera, Ministerio de Economia y Finanzas, Lima, 1973. - '87 - CART I. PFRU: STRUCTURF OF Ta FINANCIAL SECTOR (as of March 1976) fISIVA I iM-.*S ,,- a s ..ca_ c a ... I PWNAU COC'43ACIAL tASSCIA1lO IAaC. d 4. 3#.EIC. C. 'td ME ?a. .j S 8L4 k. Cp .1l, l.a. E 4 Lad.. V t' t4 . Lr MtIOiAL COARCIL P 44 Lam S.. __- M.. 1A..I D* k I Ta'v. I00A COWAC e C 'Lr eS .i'1 | &~US W..1-4 |.m V SU Fk..} | C* l.i AA... e, .-..a.* S.o, SamCn.. 5ourcef Constr. -fro -c -ano C'l d e .r def Pe.r Cuentas Fin cieras de C_ l . Pe 965-97-.de L I$8~~~~~~~~~~~~~~~~~~4*- 04 ht _ |u eX*r. b ^ s IA= U I~~~~~~~~~~~~~~~~~~~~~~~~~~~0,0 i.i..5- P~ cz^¢..w v-$ . ic 0c 7~ ~ ~~~~A Cu-.,-iacersdl eu 96-96 - 188 - Policy-making bodies were modified, 1/ COFIDE was created in 1972, several commercial banks were taken over by the State, and other public financial institutions (such as the National Bank and the Housing Bank) were reorganized. 2/ Most of these changes achieved their stated objectives of strengthening public control and credit selectivity. The creation of COFIDE was probably the major institutional development, but the diversity of functions assigned to it--a combination of multipurpose development bank and holding company (mainly for the public enterprises but also for some private ones)--make it unlikely that it will be able to carry them out with the required efficiency. Consideration should therefore be given to separate its role as holding company from its other role as general coordinator and major supplier of medium- and long-term credit. This would require creating two separate institutions, only the latter of which would remain within the financial sector. 1/ Most changes in administrative control and responsibilities originated in a general Decree-Law of Reorganization of Ministries of December 1968. Then, the Organic Law of the Finance Sector of March 1969 reorganized the Ministry of Economy and Finance and defined the Central Reserve Bank of Peru, the Inspection of Banks and Insurance Companies, the National Bank and the Central Mortgage Bank as public decentralized agencies. In 1971, the National Supervisory Commission of Enterprises and Securities (formerly the National Securities Commission, created in 1970), the four state development banks and COFIDE were also declared public decentral- ized agencies. Also, a number of permanent advisory, supervisory and counseling bodies were created. Most relevant for the financial sector are the Monetary Policy Council, the Council of State Credit Policy, the Council for Coordination of Associated Banks, the Council of State tanks, and the Committee for Coordination of State Banks and COFIDE. The Monetary Policy Council acts as the supreme authority in monetary policy matters and is composed of the Minister (Chairman) and Vice Minister of Economy and Finance, the Inspector General of Banks and Insurance Companies, and the presidents of the Central Reserve Bank and COFIDE. The Council of State Credit Policy coordinates the country's credit policy, both public and private, and is composed of the heads of some other ministries in addition to the members named above (see Chart I). 2/ The most significant changes were made at the National Bank, which bene- fited from: (i) decrees ordering all public sector entities to keep their deposits there and obliging all commercial banks with principal office in Lima to deposit there 30% of their increase in deposits after 1969; (ii) external credits; and (iii) beginning in 1972, refinancing by the Central Bank. The National Bank also controls a large share of Peru's foreign exchange. - 189 - b. Monetary Policy Instruments 163. Virtually no change was made in the last decade in the use of monetary policy instruments. The only effective instrument consisted of Central Bank refinancing of the banking system. Between 1965 and 1974, the State development banks and, within them, principally the Industrial Bank, were the channels preferred by the Central Bank to inject new money (Table 50). In the last two years, Central Bank credit has soared and has been oriented through the banks (mainly the National Bank) to finance the Government's deficit. The Central 3ank has not used legal reserve policy only for purposes of money supply control. 1/ Instead, it has been imbued with an elemenat of credit selectivity by the requirement that a fraction of legal reserves be deposited with the National Bank. This tends to ccnfuse the use of legal reserves as an instrument of money supply control with portfolio guide:Lines for credit selectivity. It would be better to separate out the portion of legal reserves that are held in cash and/or deposits with the Central Bank from those which are not. 164. The use of selective credit policies has been dramatically intens- ified since 1968. The aim has been to channel a larger share of credit: through the State-owned banks and to the Government and priority sectors. A variety of devices has been used to achieve this, including institutional development of public financial institutions, rediscount policy, portfolio requirements and interest rate policy. 165. The first objective of channeling a larger share of credit through state-owned banks has been achieved (Table 51), the result of policy directives from monetary authorities that tended to put private commercial banks at a disadvantage with respect to the rest of the institutions. They had tc derive their resources mostly from the private domestic sectors at interest rates 1/ Appendix Table 6.15 summarizes the complex legal reserve system for banks and other financial institutions in effect since 1969. Central Bank measures were aimed at two main objectives: to discriminate among insti- tutions in line with other selective credit policies, and to ensure that almost all credit institutions which derive substantial resources from the domestic money and capital markets voluntarily deposit part of their legal reserves in Public Investment Bonds. The latter objective had a two-fold purpose. One was to avoid direct financing of the GovernMent by the Central Bank, and the other to reduce the foregone interest income to the financial system. Whether the adopted measures were the most appro- priate for meeting the stated objectives remains questionable. The use of legal reserves as a selective credit instrument may produce disi:ur- bances in the money supply if switches occur in deposit holdings among institutions that are subject to different legal reserve ratios on the same or similar financial instruments. This may be aggravated when insti- tutions are free to choose whether to hold legal reserves in the form of cash or deposits with the Central Bank (which do sterilize Mloney supply) or in the form of deposits with other financial institutions or bonds (which do not). Such policy confuses unnecessarily the use of legal reserves as an instrument of monetary control with the use of portfolio regulations which is an instrument of credit allocation. - 190 - Table 50: Components of Net Domestic Credit by Central Bank, 1965 (in billion Soles) j1965 196t 1970 11972 1975 197o Total 6.7 10.5 12.4 18.5 39.9 14 .3 Net Public Sector 4.6 9.1 9.4 9.8 9.4 9.9 Commercial banks 0.1 0.2 1.2 1.7 2.1 14.1 State Development Bank 2.0 1.4 2.2 4.9 14.4 25.1 Nationial Bank -.Z - 2.1 5.9 38.2 COFIDE and FONAPS _0. _. - 0.6 3.9 -0.2 -0o.4 - 7.1 3.0 source: Appendix Table 6.2 Table 51: Banking Systenm ad COFIDE. Percentarc Distribution of Credit in Grcup's Total 0 1965 | 1968 1 1975 1 9 I Co=mercial banks 63.6 56.6 44.6 State developncnt banks 26.7 37.5 37.0 National Bank 9.7 5.9 9.9 COFIDE _ -- 8.5 Tocal 00.0 100.0 !00.0 Source: Apprc1ix Tables 6.2 to 6.8 Table 52: State Develop5.cnt Banks: Percenta;c Distribution of Credit in Group's ToLal (in percent) i 1,965 19C8 1975 i'roducLiol-oriented banls 82.2 83.1 60.6 Agricultural Bank 42.3 42.3 28.2 Industrial Bank 34.8 33.6 28.9 )Rinine Dank 5.1 7.2 3.5 Housing banks 17.8 16.8 39.4I Central 'ortgnazg Bank 17.0 1G.5 30.1 Housing Bank 0.8 0.3 9.3 Tocal J t0).0 I Or). 1 0 0). 0 _ Appondix > 6.2' 'lo 6.8 - 191 - which were both fixed and unrealistically low, paticularly after the outbreak of significant inflation in 1973. Commercial banks were not refinanced by the Central Bank in any significant amount except after mid-1975 when they suffered from a severe liquidity squeeze owing to the rapid loss of international reserves. The 9 percent rediscount rate that applied to them on normal rediscounts was higher than rates fixed for the remainder of the institutions. Finally, the marginal reserve requirements that all commercial banks were required to deposit with the National Bank acted de facto as a mechanism of retransfer of resources from the private to the state-owned banks, because the National Bank provided considerable credit only to the latter ones. Indeed, the volume of credit provided by the National Bank to the associated banks exceeded the total volume of reserves deposited with it by the entire commer- cial banking system. 166. By contrast, the pubLic financial institutions received credit from the Central Bank and capital transfers from the Government. However, within the group of State development banks, the ones that grew the most were not those that had more privileged assistance from the Central Bank and the Government--the Industrial, Mining and Agricultural Banks--but those that paid highest interest rates in their deposits--the Central Mortgage Bank and the Housing Bank (Table 52). The rapid expansion of the Central Mortgage Bank was almost exclusively based on resources obtained from the domestic private sectors, an expansion made possible by the relatively high interest rate offered on Savings Mortgage Certificates. The Housing Bank derived resources from both the Central Bank and private sectors; its importance increased after it was authorized to issue the same certificates as the Central Mortgage Bank. 167. ThR relatively higher success of the Central Mortgage and Housing Banks raises an important policy issue on selective credit policies. As discussed in various of the sectoral chapters, the problem of instituticnal credit in Peru is not high interest rates; the problem is access: the finan- cial sector has too little resources (institutional and financial) to lend more to a larger number of borrowers. In this case, the housing sector banks have been *.he most successful and with relatively little help from the Govern- ment, except to allow them to have less negative real interest rates than other financial intermediaries. Therefore, a simpler policy of more realistic interest rates and higher rediscount rates addressed to priority uses would have resulted in more resources flowing to the financial system, a higher share of the resources flowing to the priority sectors and a larger number of borrowers served by institutional credit. - 192 - 168. To channel credit to priority sectors the monetary authorities imposed dilferent variances of cumbersome portfolio regulations. 1/ These have a series of deficiencies. Firstly, in practice the portfolio policies have been less a policy to redirect commerciaL bank credit and more a policy of forcing the purchase of public investment 'onds. Indeed, the banks have to some extent little choice but to buy these bonds. Secondly, since income from the public investment bonds is tax-exempt, the commercial banks may be making a profit out of holding public investment bonds, because the gross tax- free income from these bonds can become so large as to eliminate any tax obligation from other profits. 2/ Therefore, the banks may be allocating less resources to priority sectors than they would have in a market-oriented environment. Thirdly, the State's loss of tax revenues may be larger than 1/ The first portfolio measures on commercial banks--apart from those implied in the marginal legal reserve requirements--were taken in April 1969. They required all commercial banks with principal office in Lima/Callao to direct 60 percent of the increase in their credit after May 1, 1969, to specific priority uses, described in detail: within agriculture (at least 10 percent); fishing (at least 7 percent); industry (at least 40 percent); and construction (at least 3 percent). These detailed intrasectoral allocative guidelines were revised subsequently to facilitate bank's compliance and to fit the evolving goals of Government policies. In March 1971, commercial banks were authorized to cover deficiencies in selective credit to agrizulture and fishing by purchasing instead Agrarian Development Bonds; in the case of deficiencies in selective credit for fishing for domestic consumption, the requirement could theretofore also be met through credit to the Fishing Development Fund, managed by the Industrial Bank. Beginning April 1971, banks were allowed to comply with the 40 percent credit requirement toward industry in a global manner, which meant they couLd lend to first- and second- priority industries. from October 1971 on, banks not able to comply with any number of existing portfolio norms were requested to purchase Public Investment Bonds by the amount of the total deficiency. Those bonds also became the only means of covering those deficiencies; all previous alternatives were thereby abolished. In addition, banks were requested to hold at least 20 percent o' the increase in credit after October 29, 1971 also in Public Investment Bonds. Finally, in March 1972 all intra- sectoral guidelines were suspended; the only requirement left was to dedicate at least 60 percent of the total increase in credit that had taken place after May 1, 1969 to any of the designated sectors and uses or, alternatively, to purchase Public Investment Bonds in the amount of the total deficiency (Appendix Table 6.22). 2/ This is so because net profits for tax purposes (i.e., excluding non- taxable interest income from those bonds) became negative. This is particularly probable since the operating costs of holding Government bonds are much lower than those incurred by providing loans. In this case, banks may prefer to place more funds in bonds and less in priority sectors where risks and costs are higher. - 193 - the gain it makes from placing the bonds at low interest rates. Finally, this policy has nourished the Government's disequilibrating expansion, wnile it diminished incentives to rationalize public investment. 169. As mentioned before, interest rate policy has been characterized by almost complete inactivity, despite substantial variations in price changes. 1/ From the beginning of the century up to May 1976, a usury law fixed maximum interest rates of 14%. The only source of flexibility resulted from some diversification in the structure of savings instruments, and the increased reliance on selective credit policies. In May 1976 the maximum limit of the usury law was raised to 19% and two small adjustments took place in the normal rates in June 1976 and March 1977. In light of the historical record, these measures were significant but with the higher inflation during L978, raising the limit from 14 to 19% has had little practical significance. c. Monetary Behavior and Inflation i. Money Supply 170. Graphs V to VIII present a synopsis of the principal components affecting money supply. Graph V shows that at least since 1968, annual varia- tions in the rate of change of the monetary aggregates have been the result of much wider variations in the rate of change of the monetary base; this implies that short-term changes in money multipliers have tended to counteract: and smooth out variations in the rate of change in the monetary base. Graphs VI and VII show that all money multipliers exhibited significant variations, but they were mainly dominated by the banks' reserve ratio which acted as a built-in stabilizing force tending to offset sharp variations in the mionetary base. Since interest rates were fixed, the explanation for the offsetting behavior of multipliers probably lies in the existence of lags in the adjust- ment by banks and the public to the unpredicta-ble gyrations in the monetary base and thereby in their liquidity. For instance, a sudden expansion (con- traction) of liquidity held by banks, as a result, say, of a temporary rise (decline) in international reserves would make it difficult for banks to utilize (replenish) quickly their excess liquidity and would therefore tend to reduce (increase) the money multipliers. The only autonomous long-term variation appears to have been the moderate upward tendency of the multiplier for Ml as economic agents succeeded in economizing in their use of casha. However, the variations were very small. It follows, therefore, that most of the rise in the monetary aggregates was attributable to the increase in the monetary base. 171. Graph VIII shows the behavior of the principal components of the monetary base: net foreign assets, net domestic credit and other miscellaneous _/ See Appendix Tables 6.16 to 6.19 for a detailed presentation of interest rates of the various financial system institutions (including the Central Bank). GRAPH V PERU RATES OF CHANGE OF MONETARY AGGREGATES 1 : 45 }9 Monernry base 403M AmW2 30- A 25 - 1 15- 5-- 1~~6 1967 ~19638 1 069 1970 1971 1972' 1973 1974 1975 1976 GRAPII VI PERU MONEY MULTIPLIERS 286 _ \ fv1~~~~~~~~~~~~~ultiplier of M* 2.4 2.0 - < ,.s*^*,, /' 2.0 MuAIplier of M X/ 1.6 - 1.4 - Mulirolier of M, 1.2r I I I I , :933 193(3 1337 1233 1232 1370 1971 127~~~~I N,2 19'73 1S7 4 1975 1976 - 196 - GRAPW Vil PE,RU FACTORS AFFECTING THE MONEY MULTIPLIERS 1.8 1.6 - I2fOey mvu&lier for Mt 1.4- 1.2 1.0 - 4t ^ t w ..............~~~~~~~~~.........._ 1 .0 Rato of bank reserncs to :'ema,ld deposxis 0.8 0. Ratio of currency to demand depo sits Z8~ 2-4 Money mlulprJier for M04 2.2 1. 0.4 _Rato of curmncy to other lquid essefs held by the publc 0.2 Raab of b3nk reserves to other lquid assets held by rhe pub/.c 0 1 1 1965 1 %66 1967 19G8 19069 1970 1971 1972 1973 1974 1975 197's (A) 0 0. 0 Li - I I I C-) I 0 -U I - 0 (4 2 (3 rn - 2 -I U) (3 0 -fl (0 rii1 0 mc:: -1 (4 H -. H H (0 0 2 r71 -1 Li - 7, -< K::- w - . (I) C) (Ti (3 U)'.1 3C) ,P..IP ______________________ I I p - 198- items. The latter--largely accounted for by the Central Bank's own capital and reserves--behaved in a stable manner, but there were pronounced changes in the other two components linked to public sector and balance of payments performance. Up to mid-1973, net domestic credit had been growing at a steady pace, the movements in the monetary base reflected mainly changes in net foreign assets caused by balance-of-payments developments. However, since then, public sector deficit financing has unstabilized money supply growth in two different phases. Up to mid-1975, the exploding public sector deficits were financed with massive foreign borrowing which fueled net foreign assets to grow very rapidly, increasing the monetary base. However, by mid-1975 the public sector deficit was so large that foreign inflows were not sufficient and net international reserves started falling. To finance the public deficit and compensate for the squeeze in private sector liquidity, the Central Bank was forced to expand net domestic credit. rhis highlights two crucial aspects; (i) unless public finances tend towards equilibrium, monetary stability cannot be achieved; and (ii) the monetary authorities were handicapped in their response to the crisis for lack of a flexible interest rate policy. ii. General model of the demand for real monetary balances and the effect of inflation 172. A general model to analyze the relationship between monetary aggregates and prices is provided by a demand for money function of the Cagan type. It is postulated that the desire to lhold monetary assets depends posi- tively on real income (as in the simple quantity theory) and on the interest rate on these assets, and negatively on infLation. While the interest rate is assumed to be known to asset holders in advance, the rate of inflation that will prevail over the relevant horizon for decision making is variable and can only be guessed. It is then assumed that inflationary expectations are formed on the basis of past observed inflation rates over a finite .number of periods (n). Each period is assigned a weight ac-cording to a weighting scheme determined by a geometrically declinlng series with a parameter h as the base (O < h < 1) and integer exponents which increase as the period moves away into the past (i = 0, 1, ...n -1). 1/ The parameters n and h are chosen -2 empirically so as to maximize the R of the regressions. 2/ 173. Table 53 contains the results of the demand for money regressions estimated in accordance with the method just described. 3/ (See Technical Note B for details on the demand for money functions.) Interest rates do 1/ All weights are normalized so that they add up to one. 2/ For h = 1 all the values of the actual and past rates of inflation that enter into the determination of the expected rate of inflation are given equal weight; for values of h approaching zero, the weights decline very rapidly and become insignificant for earlier periods. 3/ For an earlier study on the demand for noney in Peru along the same lines see J.R. Patron-Basurco, et. al. , "La Demanda de Dinero en el Peru," Symposium on Capital Mlarkets in Peru (Comision Nacional de Valores, Lima, 1972). - 199 - not appear in any of the regressions because they did not change significantly in the sample period. This implies, of course, that the expected inflation variable was almost perfectly correlated with expected real interest rates and was thereby a correct proxy to be used. 174. As expected, the equations for all four monetary aggregates showed a very high correlation between monetary aggregates in real terms and real income (Y) and a significant negative correlation with the expected rate of inflation (dp/p)e. It turned out on the basis of these estimates that the relevant period of formation of inflationary expectations was of five or six years including the current one, and that all these years were assigned equal weight. A direct consequence of these results is that past and presernt infla- tionary experience was detrimental to real monetary growth. 175. The inclusion of a dummy variable in the regressions to correct for structural shifts detected in the demand for monetary aggregates as a result of the 1970 capital repatriation measures showed, as expected, that there was indeed a shift in all variables except M3 in that year, and that the demand for money Ml was the most significantly affected. In the case of M3 the reduc- tion in foreign currency deposits appears to have offset the increasec demand felt throughout the spectrum of monetary assets. iii. Requirements for Balanced Financial Growth - Implications for Money Supply Policy 176. A simple monetary rule consistent with the above results on the real demand for monetary balances would consist of expanding the supply of money aggregates in a certain proportion of the planned rate of economic growth 1/ plus the projected increase in prices. In addition, some allowance ought to be made for the rise in velocity (above the average observed in the past) during periods of visible and persistent inflation; this means that the target for expansion in the money supply should be somewhat less in relation to income and the tolerated rise in prices during those periods than under stable prices. 177. Annual monetary programs essentially consistent with this framework were followed by the Central Bank in the 1970s. Although it was able to meet its annual targets fairly adequately through the use of redis- count policy, those targets had to recognize the de facto constraints imposed by the need to accommodate, although indirectly, the large public sector borrowing needs, as well as the changing circumstances regarding the outcome of the balance of payments. These factors explain the limited success in controlling inflation during some periods, e.g., in the last three years. The highly expansionary plans of the public sector during those years--despite 1/ This proportion depends on the real income elasticity of demand for each of the four monetary aggregates. - 200 - Table 53: l'cru: Dctmalnci or Mloney 1zcrcssic,ns (1959-1975>-- ln. (li/P) a + b In Y I c (dP /P) (n = 5, h 1.0) in (m1/P) = -1.05614 + 0.86347 ln Y - 0.0200G(dP/Pj)e - 0.28071 DULTI (3.36989) (11. 37582) (2.42751) (8. 27.755) - 2 0.98951 ; SEE = 0.03/i31 ; DW; = 1.29047 (n /4, h- 1.0) 1.1 (112/IP) = -0.91879 + 0.88638 ln Y - 0.0215,(dP/P) - 0.12193 ]ITh (5.07650) (21 .61S533)' (6.10419) (5.97035) -2 R = 0.991,73 ; SEE - 0.02011 ; DW = 1.92290 * , 4 PI fh 1.0) 1i. (1v1i/) -) -1.101854, -- 0.421272 Iin Y - 0.012563(csP/p'e (9.7539465) (34.462:55) (4.596418) - R = 0.987568; SEE .0,026133 ; DW = 1.6042739 (n = 5, h - I. 0) in (m4I,/r) -1.68104 + 1.09891 ln Y -- .02772(d1'/P) - 0.07656 Dl)ri (6.70255) (18.52027) (1,.19'9 3) (2.82185) R =0.99105; SEE = 0,027/46 ; D!4 1.53069 1/ The varial1c DUM is dcfincd s I for 3.959--69 C' - dJ 0 for- 19o70--5; (141/P) is thc rl-OJl valm.c o£ movxe.y sLtoc:!' -j (i - 1, 2, 3, 4); Y is C,G2T) in billionis of const'.nt solos; (dl]'/l') P is the pctecl ratc of ci ;lmjck j. L>1)0. consul.-.cer prce imdcx . 1Olcu1win8_ uzui.1-p.1C tS , tlU pi,gu e' p1 ;l'(n-- tlh-e.s's bc1low. tlhe cm.ffci.nl n - r' tlhe corre As nl ; t-.t ci c . . s explained in thc text,s n ls t:hc numl.cr of p,: ol; uzcd in cons- vucetindI thle (l)c-tcCd i.UL:rcn seriee innd h tS .t 9 Cl -;!e bc L):ccn 0 !-6.- 09 - 2.7:.7!:' 2.43i -6.4 ''374^2 2.75- - . .7-- 0.0C -G. 7 2' -3 2.:; -i *.;i 2.7.9 -3.:i :95i Q.G -;^.:cc 5 ;.9?i'7 -?.571CC.7 2 .i4 -9 2.543 -3;-I ;,_39 C0.^ -).0.47E:<.S 7.?^ -S.C 3S^C S 2. C; I .9 ).i 2.63Z -S.;:---?' _;,0 . 0.030 -,9. _2' 3 "' . . - 7._4,1252.' 1 '57 _7.j;$27; 2,52. -7. <,.3_ o?zl 0.^oco -a. 3c: l ' -6.95tr: c 2 -6.9 027,9 2.?;2 -S.-'-2 ;.03 -£7.3; . -5.5245 2 1S :-.i64, 3_. O I. :.3 . *3 -8.2*2cI 5 2.,: -4.52v .69 . -4. :336Z4 3.077 - -- _;74 O.Cc; -,7. .3 -3.93C.. 2 1. 69 2_25 _3.5 _ Z L. _v1;75 O.C1103 -!C. :16-i74 1.423 -)1J.C:5215/ 1 6.7 Z0 '2977 A 2.i7B - (o~I ec, 2 a.1 :O~tznc !0=41a i ,2,3 ) .e.:1 .. e~; ~ .~~~~~~~~03 ~~~~~~~~~~~. . .............. .EC: ths vzl:Zs of(0 /77 V"c t-'.~-d c t-3gtattiic Of th-C i==d f'Oz ==27::~si ..o(/V + 'D la~ (J/.) Z19 1/ This is one of the two weighting schemes (the most conservative) that performed best in the demand for money regressions. The adoption of the alternative weighting scheme would produce only slightly diflferent results. Either of these two weighting schemes would also be approxi- mately equivalent to the one that would have been derived from the earlier results on the demand for MI and M2 by Patron-Basurco et. al., OD. cit. - 204 - 182. In contrast with the above rule, the figures in Table 55 imply that the expected real yields on savings and time deposits were consistently nega- tive throughout the whole 1959-75 period because the expected rate of infla- tion was always above 7 percent, or the effective yield on time deposits. 1/ Thus, for instance, in accordance with that rule, a rise of just under seven percentage points in the interest rate structure would have been required in 1975 in order to bring the real rate of growth of monetary aggregates--which was negative since 1973--into the path of equLilibrium financial growth. 2/ 183. The pursuit of a realistic and flexible interest rate poLicy as described would be decisive for the efficient. mobilization of resources, although it may not effectively raise the total level of private savings. 3/ By having preferential interest rates on the lending side of financial insti- tutions or Central Bank rediscount rates, a conflict with the design of credit selectivity can be avoided. Furthermore, the policy could have advantageous long-term repercussions in other areas. For example, it- could engender some type of a free domestic market for government bonds--which would undoubtedly develop if yields were appropriately attractive in line with those on financial instruments--and it could initiate small-scale open market operations once such a market had developed. The higher cost of money for the public sector that this policy would imply could help check overambitious public investment plans and help improve the rationality of public investment decisions, without eliminating the opportunity for leadership of the public sector. But public investment plans would then have to be financed to a larger extent by raising an appropriate volume of taxes or by cutting unnecessary expenditures. 1/ This study does not directly address the issue of whether interest should be paid on demand deposits, as some recent theoretical work has suggested. However, it should be clear from the theoretical premises established earlier that the 2 percent payment allowed bv the June 1976 measures is a step in the appropriate direction. 2/ For a discussion of the underlying theoretical framework and its apDlica- tion to Peru, see Vicente Galbis, "Interest Rate Policy Framework for a Developing Economy: The Case of Peru," IMF, DM/76/98, November 16, 1976. The theory itself is developed in several recent works amongst which mention might be made of the following: Vicente Galbis, "Financial Intermediation and Economic Growth in LDCs," IMF, DM/74/103, October 1974 (in the Journal of Development Studies, special number, January 1977); Basant K. Kapur, "Two approaches to Ending Inflation," Chapter 9, in Money and Finance in Economic Growth and Development, Essays in Honor of Edward S. Shaw, edited by Ronald I. McKinnon (New York, N.Y., Marcel Dekker, Inc. 1976); Ronald I. McKinnon, Aoney and Capital in Economic Development (Washington, D.C.: The Brookings Institution, 1973) and Edward S. Shaw, Financial Deepening in Economic DeveloDment (New York: Oxford University Press, 1973). 3/ See Vicente Galbis, "Financial Intermediation and Economic Growth in LDCs: A Theoretical Approach," Journal :f DeveloDment Studies, special issue, January 1977. - 205 - 184. Among other advantages, a realistic and flexible interest rate policy may help to avoid the adverse effects of readjusting back to stability and, in particular, to reduce to a minimum the waste derived from having to stop or curtail already started investment projects for lack of finance; it may help avoid the development of an oligopolistic financial structure; and it may prevent deterioration in the distribution of income. It should be empha- sized in this latter regard that the lack of opportunity for a safe, positive yield on financial savings tends to militate against the interests of the lower-class saving units because they are less likely to have access to any of the three escape valves (luxury consumption, inflation hedges and external placement). This aspect of the policy would therefore also be relevant toward the implementation of the development goal of increasing the participation of all citizens in the socio-economic benefits. 185. Interest rate policy should not, of course, be expected to offset all the disturbances caused by inflation, but only those afflicting the financial system and the efficient functioning of the saving-investment process. For this reason, activation of interest rate policy should not diminish the principal emphasis on demand management policies to control inflation. A type of inefficiency imposed on the economy by inflation results when foreign trade controls are intensified to check the deterioration in the current account of the balance of payments. Inflation may also produce undesirable consequences on the labor and factor markets through the disruption of the wage bargaining process, strikes and other attending events; or through inefficient controls on domestic prices and rents and wages, which may have to be escalated and used as counter-cyclical weapons. Again, the atmosphere of crisis created by the rise in foreign indebtedness, eventual loss of international reserves, lack of confidence in the domestic currency and devaluation, tends to compound inefficiency by increasing several kinds of economic risks. - 206 - CHAPTER VI: THE ROLE OF THE PUBLIC SECTOR 186. The role of the State in the economy has been changing rapidly since the early 1960s. In the 1962-66 period, the Belaunde Administration expanded the General Government operations rapidly (Table 57) and invested heavily in basic infrastructure and social services (mainly transport, educa- tion and health.) The institutional structuire of the General Government was somewhat rationalized and efforts were made towards development planning, tax reform and land reform, but the principal role of the State remained in the provision of infrastructure with relatively little direct intervention in the rest of the economic activity. This changed drastically after 1968. Chapter IV showed how the new development st:rategy called for a pre- dominant role of the State in shaping and controlling the allocation of resources in the economy. This Chapter first gives an overview of the institutional organization of the State as :t relates to the economy and of the magnitude of penetration into key economic sectors. Then the dis- cussion focuses on the most important aspect:s of public sector operations influencing the past and future developments of the economy: (1) public investment and efficiency of resource allocation; (ii) public savings and resource mobilization; and (iii) management capacity. Tabie 57: 13rDICATCrUS OF 7FLYT:vE SIZE O' T2 71BfLIC SEC2CH, 1960-76 19t6 1 l962 1970 1' 74 7975 197S A. Percent of CuP General Gover=ment Current revenues 15.3 17.9 18.9 19.8 20.4 19.9 16.8 Current eypenditures 12.3 18.3 17.5 16.3 17.1 18.8 17f! Public Scctcr Invest=ent 2.2 4.5 3.t 4.0 8.1 8.2 7.5 General S3vern=ent 1.5 3.2 2.4 2,5 3.4 3.1 2.1 Puo l.e Enterprises 0.7 1.3 1.3 1.5 4.7 5.1 4.7 B. Percent of Total Fs ed Cazital Formation Pu'blic Sector Snvestment 13.0 27.] 28.5 '2.2 51.o 4 1 51.3 General Gcvernment 9.2 n.e. 183.1 20.3 21.7 17.8 19.4 P?l.c Entertrises 3.8 n.a. 10.4 11.9 29.9 .9.3 32.4 Sources: Appendix Tablz 2.1, 2.7, 5.7, 5.1O, Ccntr!l Bank led Fltnnnin:g Of'fice (IINP). A. IMPORTANCE AND INSTITUTIONAL ORGANIZATIOlI OF THE ECONOMIC PUBLIC SECTOR 187. Since 1968 the State has become the principal agent in the economy by massively increasing its ownership in key econoic sectors and by setting up a voluminous body of legislation to steer economic activity outside the public sector. - 207 - 188. "his expansion of State activities increased the power of the General Government, but did not resulted in a large increase in its relative size. 1/ Table 57 shows that the shares in GDP of its current expenditures and its fixed investment have not grown significantly since 1968, in contrast with the sharp rise of the earlier 1960s. 2/ Moreover, Table 58 suggests that by international standards the size of the General Government in Peru is not unusually large. At around 17.5% of GDP, current expenditures are lower than in various other Latin American countries and similar to the average for all developing countries. 189. The expansion of the size of the State has come from the accumula- tion of state enterprises. In 1968 the State owned less than 20 enterprises, while it now fully owns around 110 companies and partly owns at least 60-70 other enterprises. 3/ Of the 110 fully-owned state companies, 42 are legally 1/ The General Government in Peru revolves around a highly centralized Central Government. Local Governments are weak, have few indepenc.ent sources of revenues and depend on the Central Government for a la-rge share of their financing. A large number of decentralized agencies perform health, welfare and other government services. With some ex- ceptions (such as some welfare institutions and the national technology institution, ITINTEC) these agencies also depend on Central Government transfers. The Social Security System completes the structure of the General Government. Although it has been growing rapidly, the Social Security System still covers only a small share of the urban labor forcE. As discussed below, in recent years it has become the largest supplier of General Government savings used predominantly to finance deficits in other parts of the public sector. 2/ Data on public sector finances in Peru are poor and incomplete. Before the mid-1960s, even information on the Central Government was unreliable because of the disaggregation of accounts into special funds with earmarked revenues. Since 1968, the information on Central Government operations has vastly improved and is now produced accurately and with little delay. Data on the rest of the public sector remains poor, particularly for public enterprises and the Social Security System. The only fairly accurate time series are on domestic borrowing from the banking system and foreign medium and long-term borrowing. 3/ The total number of wholly- or partly-owned state enterprises is d:.fficult to estimate because in nationalizing some companies the State acqu:red a long chain of subsidiaries. For example, PETROPERU inherited a large shareholding in a company manufacturing gas stoves and HIERROPERU inherited a customs agency. By origin, the state enterprises can be classified into three groups: (i) those that were foreign or domestic-owned a-Ld were partly or totally nationalized (the majority, some 110 companies); (ii) those that existed before 1968 as wholly- or partly-state-owned corpora- tions (not more than 20); and (iii) those new enterprises created since 1968 with State participation in the original capital (some 50 enter- prises). - 208 - TAble 58: CETEPAL GOVER(TIN0T EXPMD;TU.ZES AS PZ.'VT OF CDP I'J VARIOUS COUtI2RIES, 1960-73 196o 1973 Industrialized '. untries 25.1 -n.2 neveloi05 countries 13.3 17.4 Western Itemisp?bere /1.9 16.8 Peru 12.3 17.4 Chile X. .0 37.7 Uru.-uay 22.0 37.7 Brazil 15.T 25.4 Fcua:dor 19.1 22.5 panama 13.4 18.2 Costa Rica 12.9 17.8 Boliria 13.6 17,6 Venezuela 15.3 14.6 Argentina 8.5 11.2 Colombia 7.5 10.14 :4ex.ic3 5.1 7.2 I/ Includes developing countries unly. So0urce: 'Table 1, .Ap;endix Tpbles 2.1 and 5.7, and I WRD, World toble5 1976, pp. 440o444. considered public enterprises, the rest are run as private enterprises. 1/ There is little comprehensive data on the aggregate importance of state-owned enterprises in the economy; however, some rough indicators can be provided. Total assets of state enterprises are consi.erably larger than those of the entire (private and public) manufacturing sector, making the State by far the largest entrepreneur of the economy. 2/ The current expenditures of a sample of public enterprises in 1976 were equivalent to 16% of GDP and their fixed investment equivalent to one-third of the tctal fixed capital formation in 1/ For example, CENTROMIN, the largest mining company in Peru, is fully state-owned but is legally run like a private company. Other state enterprises in the same status include agro-industrial companies, cement and paper mills and three commercial banks. 2/ In 1973 the Planning Office (INP) estinated the total assets of the non-financial state enterprises at around SI. 65 billion compared to S/. 50 billion estimated (by the Ministry of Industry) to have been the total assets in the manufacturing sector. However, the 1973 total of state enterprises does not include three of the most important ones -- PESCAPERU, CENTROMIN and HIERROPERU -- which were nationalized duting 1973 and in 1975. (See INP, Analisis Economico de las Empresas Publicas, 1973 Lima.) - 209 - the country. 1/ Fitzgerald estimates that with the addition of state enter- prises, the State now controls at least one-eighth of output and one-tenth of employment in the country; if small-scale and artisanal activities are excluded, the shares rise to about one-third and two-fifths respectively. 2/ 190. More significant than the sheer size is the selective penetration of State control into the key sectors of the economy, illustrated in the functional classification of the principal state enterprises summarized in Table 59. In the first group are those enterprises through which the State has gained a predominant position in the traditional export sectors. Foremost is PETROPERU, which is in charge of all petroleum refining and marketing operations and is the State's agent for petroleum exploration and production. It is also responsible for negotiating with foreign oil companies the conditions of their operations in the petroleum field. 3/ In mining, a wholly new enterprise, MINEROPERU, was created as a holding company for all other state-owned mining ventures; it is to develop and manage most of the new large mining projects in the country (alone or in joint ventures with private capital.) It also holds the monopoly for new metal refining activities. MINEROPERU is the shareholder of two other state-owned mining corporations -- HIERROPERU (iron) and CENTROMIN (other minerals) -- formed after two of the three largest mining companies in the country were national- ized. 4/ In fisheries, PESCAPERU holds the monopoly of fishmeal and fish 1/ These estimates pertain only to those companies referred to as "public enterprises" and not all state enterprises. The figure on current expenditures (Source: Appendix Table 5.1) includes expenditures fo. purchase of goods traded by the marketing enterprises. On investmelnt by public enterprises see Table 57. 2/ The esrimates on the size of output and employment controlled by the State are from a detailed study by Fitzgerald. In studying the patterns of ownership in the economy, he breaks down output and employment first into what he calls a "modern sector" and a "traditional sector" and then breaks down the "modern sector" into state-owned and the rest. The "modern sector" corresponds roughly to enterprises of more than 5 employees. However, as Fitzgerald points out, the esti- mates from his analysis are subject to wide margins of error and should be taken as orders of magnitude only (See Fitzgerald, E.V.K., The State and Economic Develooment: Peru Since 1968. Occasional Paper 49, Department of Applied Economics, University of Cambridge, Cambridge University Press, 1976, Chapters 3 and 4). 3/ For a broader review of PETROPERU's operations see Chapter I. Petroleum was an important export in the past. Although at present PETROPERU does not export much, the Government had high hopes that it would become a major exporter. However, as discussed in Chapter I, the chances of this occurring at present are low. 4/ For a broader review of the operations of. the state-owned mining companies see Chapter IX. - 210 - oil production, and the three main food fish processing plants 1/ are partly owned by the State. The fishing operations of anchovies to produce fishmeal were also monopolized by PESCAPERU from 1973 to mid-1976, when they were returned to the private sector. 191. In the industrial sector, the Government's strategy was to increase State participation mainly in the basic heavy industries. It now has the monopoly of production of steel (SIDERPERU), fertilizers (FERTIPERU and the Cachimayo Company), cement (with four plants) and telephone equipment (FETSA). It also has an important share of the shipbuilding industry with SIMAPERU and through INIDUPERU has majority control in companies that produce tractors, diesel engines, air compressors and machine tools. 2/ Finally, the State controls a large number of small industries -- ranging from gas stove production to tobacco and coca processing -- which it inherited from previous administrations or which were subsidiaries to other nationalized companies. 3/ 192. The Government has also placed a growing share of the trade and marketing functions in the hands of state enterprises. MINPECO, s subsi- diary of MINEROPERU, has the monopoly for export sales of minerals; EPCHAP has the export monopoly for fishmeal; and EPSA has the monopoly for coffee exports, most fresh food imports, and for the domestic marketing of an increasing number of agricultural products. EPSA operates on both sides of the market, supporting or controlling prices. ENCI controls the imports of some important industrial and agricultural inputs and of some manufactured foodstuffs (such as candy) and alcoholic beverages. Finally, the Central Committee of Sugar Cooperatives (CECOAP) is the only export seller of sugar. 193. The State has also expanded and strengthened its position in the financial intermediation system. In addition to the previously existing specialized public development banks (Industrial Bank, Mining Bank, Agri- cultural Bank and Central Mortgage Bank), a, Housing Bank and a large Development Finance Company (COFIDE) were set up. COFIDE combines various functions. It acts as holding company for some of the public enterprises and also acts as financial agent for all public enterprises in the domestic and foreign capital markets. At the same time it promotes industrial and tourism development by participating in equity capital of and extending long-term loans to private corporations for debt financing. The State has also gained an important position in the ccmmercial banking system by taking over three of the five largest commercial banks, while the National Bank's size and influence have been considerably expanded. 4/ 1/ Mainly freezing and canning. 2/ See Chapter X for a more detailed review of the operations of SIDERPERU and INDUPERU. 3/ See Table 59. 4/ See Chapter V. - 211 - Tikble SS iT ;a or PBLIC E£NlrRIS2 AND OTHP21 STATE-OWN3I CO'AYIES (EXCUJ0INC TIACIAL lTFSWIMM :AS L BM - Created Capital stock after in 19t3 (in Description and Cements 1968 soles million) PNtrol eum PMoTrEu * 3686 petroleum production Lnd sale CZ=OThN 3'/ m mining production )aNERPERU N * 1544 mining production HIEltROPERU 6- n * aU. 8 iron production Industry fluWERu a' 263 industrial holding and mwagemeot company SIDEBMU eJ 6768 iron and steel production TERTlPESU 95 711 fertilizer production and sales hnp Pub. Ind. Cachiayo 491 fertilizer production 1NDSLPMU * a.a. wm=ition and uniforms Sactorr ZWIIA 120 tobacco and cigarette product EPPAPZRU 31 bandicrafts promotion 2WAL 40 salt production and sale Z2iACOCA 5 coca rarketing DfCD 1f/ 2 coca processing C0RSAN 3' * 14149 various industries SDIAPERU * 105 ship building and repair IAI 20 production azd sale of industrial alcohol nJTTA 2J * telepbone equipment maufacturing Fisheries ?ESCAPEU 11½ 7038 fises.1 and fish oil processing zp5 12/ * 349 marketing of food fisab C:iPE* 10 fisb quality control Tourism EXYURPIU 504 tourism promotion and hotel operation Trade !PSA 4/ 3 422 agricultural marketing 1NCI 1 srkeSting of industrial inpo- PCEAP * 34 marketing of fish products 9DIPECO * a. a foreign sales of minerals Transportation cPY 2500 comercial shipping line CMUPAC 985 airports authority MEAFE?XU T 721 railvays KAFUV 4136 ports authority AEROPrU 66 cormercial airline Power ELZCIrROPERtr * 8a653 electricity generation and sales ZCrOLTMA a.a. ideo for Lima Water !SAL 2498 Lima vater end sewerage ESAR 260 Arequipa wter end sewerage Communic atiOnS EIELPERU j/ * 853 telecomunications ENRADPnU n * a.a. official radio station 3SI PU a. a. official new agency CNEPEU .* . a. film production PULPE * a.a. public sector sdvertising. Other Seetors E4ADIPERU * Ii81 real estate sonagement company EDImoRAPERU 9 publishes official gazette ('M Peruano") EPPODE * a.a football loteries NnAnP9T 5 au military airplane repairs IIAI * Si aro-duAtrs, _ ] * -tntu lso a4mA 7 financiai interr.dlari., v ia turn nre p4rt Of full owners eof coc-ercia. bak3s, 2 insuranc* aa>inL * at 131 iadtrisl cigties. V Two sul;sidiazics prnducing houtbold apli.mcv% nd a smal t=nker fleet. ftpLaxd A sller state corpany (p). I State-owned but lcgaly a private com*W. Three subsidis.iss: two small oicing compnnies wM a flotation reaeCnts proXcsr. H{olding Co*Pany of all state-awed aiming enterpryis. Owns a euntoe agency. IJ j4nagea four joint-ventur vita foreign paretnrs. 3/ sisitu'i bcfore 1968 snd -az called SMl. / Created in 1964 an a Stato/priwt. Tscture, fully mlAal.is*d in 1969. ;o/ hprena !tacional de La Coca y DerirnAos (NtCI). D 33 ::obsidinries in induntrie-s ar services rel.ated to fishiveal proauctino And 1 real ectate cnapeny. Ej Three aubaidisie2 in fiAib cnoing sad fr"Isil3m. LY ,l-pep.n,L UOUIisO (Touris. Carp,r*413a) creotd -fmor, l968. Lh/ Two saubidinrieu in liseatonek fminr_ >1 rwlm oil productin, I. .uererr^ret chnin, 1 chain otf food r.t.'dl atoru:4; ;,ad 1. Mn(orcxlL c.MJAAY. 1J , illhol anries operatinor t-e 4 adteLe ntri n-t intern,ti-i. nteLlitr. t. i.;. part owner -)V '?)r ru-lio and et*levj!tio-s iitntionn. - 212 - 194. Finally, the State has consolidated its control and strengthened the institutional framework in the public utilities transportation and communi- cations sectors. In the power sector, the series of private companies that existed prior to 1968 were gradually nationalized and regrouped with the existing state-owned companies to form one large public enterprise, ELECTROPERU. Ultimately, this enterprise is to own and manage the whole electric power system. ENTEL was created on the basis of the nationalized telephone and telecommunications companies and will be in charge of all future developments. In transportation, public enterprises were created to operate the railway system (ENAFER) and the ports network (ENAPU) and to run a commercial airline (AEROPERU). B. PUBLIC INVESTMENT AND EFFICIENCY OF RESO'URCE USE a. Magnitudes and Characteristics of Public Investment in 1971-76 1/ 195. During the last six years, the size of public investment expenditures have been rising rapidly and their importance within the economy has reached levels unprecedented in Peruvian history. Between 1968-70 and 1974-76, public investment expenditures rose in real terms at an annual rate of almost 21% and doubled their share in GDP to almost 8% (Table 60). 2/ Of perhaps more significance, the share of public investment in total fixed capital formation more than doubled between the two three-year periods from less than 23% to 50%. The predominance attained by public investment was the result of its 1/ As noted before, data on public sector olerations are weak and incomplete, mainly because of the lack of comprehensive data on public enterprises. The estimates on public sector finances used in this report (and showed in detail in Appendix Tables 5) are based on a consolidation of the accounts of the following institutions (niet of inter-institutional transfers): General Government (including the Central Government, the Social Security System, local governments and other decentralized agencies and welfare institutions) and a sample of the 15 largest public enterprises. The sources of data are primarily the Central Bank and the Ministry of Economy and Finance, although additional estimates from individual agencies and from the IMF have been used. The esti- mates reached for agencies other than thea Central Government are in many ways subject to wide margins of error (although efforts were made to reconcile them as much as possible wi:h the balance of payments, national accounts and monetary data) and should be interpreted with caution. 2/ Three-year averages have been used to smooth-out the large year-to-year fluctuations. It should be pointed out that a level of public fixed investment of 8% of GDP is not high by international standards. The same ratio was in 1976 well over 8% :Ln Mexico, Ecuador, Dominican Republic and Brazil, and around 7% in Costa Rica. - 213 - 1:,I f,-I: 33C 1I! AI 1? (0. 1: l i') Cli w lrli; 11 1' 1 4:;T':,'J(T'N AtM2 oEI) n lv0-7H -~~~~~~~~~~~~~c,c,:K, __ijUw .1 '-rt-i nf Realt gron ,, .,1 / 2'r.*lt. VeraR,¢ k.tru.2 Gr'),t:h 11.zturt, Perlcl gt'. ..i;uiLurs -' Drit,?,,)= in Hteal l tCrr! _i/ - - 71-?0/ 9 979 10 A;/ 1 9-01 -7 6 1968-70 1'2 *'T3 15( .-16 97168 -70 973--'73 1968-70 Tota.L YiJxerd C(0it, 1T :iiti ' t 330.1) 100.! 00To.3I 66 1 .O 35.8 3.0.1 . I u. 1 }'0t'v'!sl;e S uter 819 7(1 8oA ''}03 9w nt7.9 le.7 3 I 7 .3 9.3 ?rivstv ].Llts rrriseK. - 717.4 61,1. 49.9 a.r 7.9 7.9 2.0 5.0 3.9 Tiubic `--terpriL:s 1.5 15.0 3'.7 1.6 2.0 4.3 13.2 4,., 28, 6 Goner!%l r,:vv,rlnr%! 18.1 23.6: .9. 4 2.' 3.1 jSi 18.6 7., 12.9 hizblic Srctorc 22. 6 38.6 50.3 3.9 5.1 7.9 16.4 25.:3 23.6 / Calc .ated es tk g- ;,cth rntes bet-we-n the mid-points of '0he 4vhree-yer excragec. 3/ Residual of enterrise sector minus public enterprises. 4/ Mnio iteuo: esual to Ccmieral Governs±ent pul pul;lic inturprises. Scmzrces: A,nuondi:4 ta'vies 2.1, -.6, sole &nd j.40. rapid growth and of the sluggish growth of private investment. 1/ The shift in relative importance was not from the enterprise sector to the general government. In 1976 each held roughly the same share of gross fixed capital formation as they did in 1968. The main shift was from private to public enterprises (Table 60). The latter accounted for 3/4 of the increase in public investment expenditures between 1968-70 and 1974-76 (Table 61). 196. With growth also came a sharp change in the direction of public investment. Priorities shifted away from physical infrastructure and social services to the commodity-producing sectors. In the 1960-66 period, one-half of public investment expenditures went to infrastructure, while only one-fourth went to the commodity-producing sectors (Table 61). Conversely, since 1968 almost all of the increase in public investment expenditures has gone to the commodity-producing sectors. The lion's share was absorbed by the petroleum sector although agriculture and mining also showed big gains. 2/ The low priority attached to infrastructure is reflected not only in the sharp fall of its share in the total, but also in the decline in real terms of annual 1/ Table 4 shows that between 1968-70 and 1974-76 fixed investment by private enterprises rose at an annual rate of 3.9% in real terms, well below the growth rate of GDP. 2/ Between 1968-70 and 1974-76 the share of public investment going to petroleum rose from less than 6% to over 36%, that for agriculture from 6% to 16%, and that for mining from less than 1% to over 6.5% (Table 61). - 214 - Taxble (1: ST..''iEU AREND 0OFZr OF *OEL:C "IM- NT OlY X I',O AFD S .'C7 q, '960-76 (in percentages) -trcentage Struct=.-r of Aversge Ann, 5;r2. z .h Ra:z Percenta.ge Structure increase 2/ in Real Ter= 2/ 1971-t3/ 1974-7o 19^- ;/ 1.960 19646 156--7CO 197j-T6 9t8-76 '963-70 1;71-73 .968-73 A. ;1.^UC=. Z:^ 3 :"fS- - ':CN Total Nblic Sectcr ico.o 100.0 100.0 100.0 100.3 ;6.4 25.3 20.3 Cener^s1 0overr.-ent 70.0 r a . 58 5 29.0 27,4 18.G T.5 12.9 Central ^aver- t e.t. n.a42.7 3.4.0 30.T 25.4 3 1 6.3 Other n.& 3-& 15.8 5.0 -3.3 -5.1 ' -0,.4 blic -.-ter-rises 30.0 n.1. 41.5 61.0 72.6 13.2 46.5 23.3 r P.CP=7U' n.a. m.a. 5.7 37.0 57.9 51.2 79.3 64.j Cthe:~~~~~~~14 M. f a. 315-' 24 .o 14.7 4.2 27.6 '3.' B. S77C=2^ 3Y -~,70¢ Total .^ebl2c Sectzr 2.00. '00.^ 1_0.0 201C.0 16.L 25.3 20.3 .-n±ity 2rod=cig sectors 10.1 24.1. 2.7 68.3 98.o _.7 612.3 4L._ ''etro-Iet= 4/ 5.7 36.5 51.2 78.9 6i,q X'ining d 0.5 0.8 6.6 e4.s 5a. 71 Industryr and eou;ris o l0-513.3 10.7 6.7 -25.2 66.5- .6 Aiculturv 9.6 10.8 6.3 15.9 36.3 415. . 1.2 ?isheries - - 0.2 2.7 178'.1 17.0 80. 3. :MfrsSt,=tt=V Sectors 52.8 51.3 .8.9 15.2 2.3 0.3 -2.6 -1.0 El ec=,r_c :y T.8 9.1 1 4.o 1 ;.5 l4.5 -7.6 -o.7 c:ro~ and 02-'n :. '5.0 w.2.2 34.9 '0.t .-. 4.2 -1.3 Soceia Sectors 28.5 24.3- 8.6 6.2 2 23.9 _ 8.3 'Vater au:pp7 aod housig 1'7.9 15.3 654 2.4 *41. _-7.6 2.9 Heal:lh - 5.3 3.4 0.7 0..4 26.6 -5.6 9.3 zcacstion 5.3 5.l 1.5 1.9 52.2 1.1 24.1 'hatisectol Pro%r3cms - -1.5 _ - 125.4 - 0ecsntral. A4encias, Socij. ; 215.3 5.0 -3.3 -5.1 ._, Sec.4it7, d. 4cl Oov 0 ts. - - - Othax. 0.2 1.5 I/ Ca.ica.tzd .in eonstant 1963 prices. 2. ?erceatsae diist-brtico cf th. charge of ?ublio ivsttment. 2/} C<:ulztd ~s the s:vge s.enial .1 h rxtee 1let.ven the ad-point3 of the .ee-ye= ave- 4/ ?etrol3'. in 1960 and 19, nd r eAig in .966 are inclided in elec'sricity. 1cvcy2r, the a&Unt :f izvezater.t .h tese sec-Aors : :oC aLd 1066 vas tee14i> 2. 5/ zn 1960 and 1966 epred 'tnr i.-,o= -rutorn. investment expenditures. 1/ Investment in social services (including educa- tion, health, water supply and housing) rose rapidly in the first three 1/ The average public investment during 1974-76 (in real terms) in infrastructure weas 5Z lower than during 1968-70. - 215 - years of the decade but have since fallen back and in 1976 represented less than 6% of total public investment. The change in priorities was the con- comitant of not only the incorporation of large mining and petroleum companies into the public sector; even the structure of Central Government investment expenditures turned sharply in favor of agriculture, fisheries and industry to the detriment of transport and social services. 1/ 197. Another important characteristic of public investment in the 1968-76 period was the heavy concentration on a few large projects. Table 62 below shows that the 12 largest projects accounted for over two-thirds of the increase in real terms of public investment expenditures between 1968 and 1976. 2/ Ten of these twelve projects (which absorbed one-third of the 1968-76 investment) were in commodity producing sectors. 3/ The concentration in large projects was heaviest in the public enterprises (eight projects explain 3/4 of the increase in investment expenditures) although it was not low in the Central Government. 41 Part of this concentration is explained by the inevi- table bulkiness of projects in some sectors, such as petroleum and mining. 5/ In other sectors -- such as agriculture where three projects absorbed 2/3 of the total for the sector -- the concentration was more a reflection of the Government's preference for big projects as a means of trying to achieve more spectacular results quickly, which was an important feature of the Government's investment strategy up to 1976. I/ Appendix Table 5.19 shows that the share of commodity-producing sectors in total Central Government fixed investment rose from 16% in 1965 to 22% in 1968 and to 45% in 1974. Most of the increase was accounted for by the three irrigation projects shown in Table 62 (Majes, Chira-Piura, and Tinajones) although fisheries and industry also showed important gains. 2/ The 12 top projects absorbed 40% of the total 1968-76 investment expenditures. 3/ Only the Mantaro Hydroelectric Plant and the Oroya-Aguaytia Road were in infrastructure. Both projects had been started well befote 1968 (Table 62). 4/ Three irrigation projects -- Chira Piura, Majes and Tinajones -- and the Oroya-Aguaytia Road accounted for one-half of the increase in Central Government fixed investment expenditures between 1968 and 1976 (Table 62). 5/ The construction of the Transandean Petroleum Pipeline alone accounted for 12% of the all public investment expenditures (in real terms) during 1968-76. - 216 - ub! I 62: CN:MCJbTATWI0; Ok' PUBLIC 1hVESTMENT IN TWELYVE UAbs,r PROJECTS SINCE i968 Instit. Percent of Total Str:cture9 f' Invest- _ Ire'rese inS 1/ 1968-70 1571-73 1974-76 1968-76 1969-76 Totz:_1 Publie Se,!tor 100.0 100.0 1GO.O 100,0 00.0 Total of 12 largest projects since 1968 ?J 29.B 33.3 51.2 42.6 68.6 TransAnd'nn Oil Pipeline PE - 0.3 21.6 11.9 :4cantaio Hydroelectric P 13.1 16.3 2.0 7.9 Petroleum Explorations PE - 4.8 4.8 3.9 Cbira-Piura Irrigation &G 0.3 3.8 4.6 3.6 'XJeu Irrigation CG 0.3 0.9 5.9 3.5 S0.Dr'9PiMU e.pansion PE 9.3 0.4 2.2 3.0 prrn Verde I - 1.6 4.2 2.7 Highway Oroya-rauaytia CG 4.3 2.0 0.7 1.7 ;ertilzar Flant PE - 2.5 1.0 1.2 io Copc;or H?einery r PE - 0.7 1.6 1.1 L ?a.--_''a &etroleun Refinery PE - - 2.0 1.1 '.1nejones I-rigation CC 2.0 1.0 o.6 1.0 Centrai Guvt. PrnJects as S of CentrallCcv. total 16.3 L4.6 ;4. 5 24.0 50o. PYblic entnrnriseo pro,4ects As o °' public enterprises total 55.2 I9.7 63.3 63.5 73.2 1/ PZ = Puboli_ temrine; CG = Central Governeent. 2/ 2anked ty umornt of, invest+-,nt (at 1963 constant pric0>s) in the '196-76 period. ,'/ oro,a.w investmoat iue to twalve Droiect.s aS a qhr-. nf t=h trtAl ij re,= in nublic unresnl:2nt beti;ren 1968 And 1976. Sources: latle 5 and Appendix Table 5.40 198. Finally, over 90% of public investment expenditures in- the 1968-75 period was outside the Lima-Callao area. The emphasis on developing natural resources explains part of this decentralized '.nvestment pattern. Investment went where the resource was located: petroleumi in the Northern Costa and the Selva, mining in the Sierra and irrigation works in the Costa valleys. However, deliberate Government policies have been instrumental in the decentralization. Most state-backed industrial, projects have been located in the Northern Costa, while the bulk of infrastructure investment went to Central Highway Project (connecting Lima with the Sierra and the Selva) and to the Mantaro Hydroelectric project, both locatecl principally in the Sierra. 1/ l/ The North was the main beneficiary of public investment (25% of the 1968-75 total) because of oil investments, followed by the South which contains the Majes irrigation project and the Cerro Verde copper mine, both near Arequipa. The Jung:le Region received 14% of total public investment, nearly entirely because of oil exploration and the oil pipeline. The Central Region including Lima received only about 12% of total public investment, more than half of which was spent on the Central Highway project and the Mantaro Hydropower Plant, both located in the Sierra. Lima itself accounted for approximately 6% of public investment in 1975. Finally, approximately one-quarter of public investment benefits -a:ious regions simultaneously (transmission lines, major highways and telecommunications projects). - 217 - b. Achievements and Problems 199. Besides increasing the level of public investment and spreading into the key economic sectors, the Government has attained other objectives with its investment effort. While in 1968-70 the lack of identified projects was a general constraint in most sectors, through extensive preinvestment efforts some of the sectors now have a considerable number of identified projects. The main beneficiaries in this respect are petroleum, minilng, large-scale agriculture, industry and to a more limited extent electric power and water supply. Moreover, much experience has been gained by the central authorities and the individual institutions in managing a large public invest- ment program. Although, as noted below, many costly mistakes have been made, the lessons of the past should help public sector institutions perform better in the future. Some projects also stand out as indisputable successes in modernizing and increasing the operating efficiency of the economy; examples are the large copper refinery at Ilo, the rationalization of the production process at SIDERPERU and the buildup of capacity by SIMA to construct ships of practically any size and at relatively competitive prices. 200. Unfortunately, however, the costs of getting the State rapidly involved in the investment process have been high and the efficiency of resource allocation has been low (in terms of objectives being met). Problems arose in most sectors. I/ The following paragraph will show that projects in the commodity-producing sectors with low or negative returns to the economy, account for well over 35% of total public investment, while projects in the infrastructure and social sectors also ran into serious difficulties that substantially lowered the benefits originally sought. 201. Two of the principal objectives of the State's widespread involve- ment in the commodity-producing sectors were to accelerate the growth of output savings and foreign exchange, and to provide the basis for a more rapid growth of incomes of the poor. 2/ Up to now the public investment program has done little to attain these objectives, and what has been achieved has come at an extremely high cost to the economy. In the 1971-76 period over $1 billion 3/ were spent by the public sector in the search for petroleum and in the construction of the Transandean Pipeline. Unfortunately, as reviewed in Chapter I, the results of the petroleum exploration have been meager and well below the expectations for these investments. Unless new large reserves are found, the return to the economy will therefore be marginal if not negative. 1/ Other Chapters on each of the sectors provide a more detailed analysis on each of the projects. .2/8 See Chapter IV. 3/ At current prices. The share of these two projects in total public investment (at constant prices) during 1971-76 was approximately 20%. - 218 - 202. In agriculture, almost $500 million were spent in 1971-76 on the Chira-Piura and the Majes large-scale irrigation projects. 1/ The Chira- Piura project consisted basically in transferring water from the Chira to the Piura rivers and building an irrigation and drainage network in the Piura valley. The Government concentrated its efforts on the water transfer and the main drainage canal, which were completed (although at a cost four times the original estimate made in 1968), but little has been done about the drainage, which will take at least seven more years to complete. In the meantime, the works will remain basically unused and the additional water flow may even worsen the salinity problem in the lower Piura valley. Therefore, because of its disjointed phasing, the project holds little chance for a positive return on the investments. The problems of the Majes project are even more serious since it is likely to become a drain on Government resources for decades to come. This project involves the tunneling for 140 km of water from the Colca River to irrigate around 57,000 ha in the Majes and Sihuas valleys (near Arequipa). Although serious doubts were expressed on the economic soundness of the project because of the high costs/hectare and the low quality of the lands to be irrigated, the Government decided in 1971 to execute the first phase of the project covering around 27,000 ha. 2/ However, by early 1977 costs estimates have more than doubled to $600 million (or over $26,000/hectare) and the probability is high that costs will continue escalating before completion, now projected for 1984-85. 31 The rate of return for the project has become so negative that, although $243 million had been spent by 1976, serious consideration should be given to redesigning or even abandoning the project. The fact that around 200,000 ha of currently irrigated land remains under-utilized because of inadequate drainage (but which could be rehabilitated at a fraction of the cost of newly irrigated lands), raises all the more doubts about the merits of allocating such huge amounts of capital into the expensive large-scale irrigation projects. 4/ 1/ At current prices. The share of the Majes and Chira-Piura Projects in total public investment (at constant prices) during 1971-76 was about 9%. 2/ In 1971 the costs of the first phase of the project were estimated at $275 million or about $10,000/hectare. The World Bank Agricultural Sector Survey of 1974 estimated that the net value-added in these lands would be at best between 1/3 and 1/2 of the required level to assure a reasonable return. Since then, the costs of the project have been rising faster than agricultural prices. 3/ The costs of the first phase of Majes may reach $1 billion before completion. 4/ For more details on irrigation rehabilitation projects, see Chapter VIII. One such project rehabilitating around 19,000 ha of coastal valleys was in 1976 expected to cost much less than US$1,000/hectare. - 219 - The Government has argued that this was the only alternative to assist the minifundistas near Arequipa. However, there are ways to raise the incomes of the 30,000 families that are to be the eventual beneficiaries at a lower cost than the Majes venture, through smaller projects in agriculture, industry, tourism or services. 1/ 203. Turning to the mining sector, MINEROPERU successfully comp-eted the Ilo Copper Refinery but has run into serious problems in executing tee Cerro Verde I copper mine and has had to stop execution of the Bayovar Phosphate 'tine. 2/ In the Cerro Verde I project, MINEROPERU ventured to use a complex : relatively unproven technology which, combined with inexperience in project management, resulted in delays and more than a doubling of costs to $220 million. In the Bayovar Mining project, MINEROPERU initiated implemen- tation before the economic feasibility had been firmly established. The return on investments in these two mining projects 3/ will be well below initial expectations. 204. Public investment in the fisheries sector totalled around $120 million in 1971-76 3/, but also ran into problems of premature i:;pler.en- tation decisions. Based on a general study of the fisheries sector in 1972, the Government decided to construct four large fishing completes. One is virtually completed (Paita, near Piura on the Northern Costa, considered the largest of its kind in the world) and two are under construction (Samanco near Chimbote and La Puntilla south of Lima). The main problem of the fishing complexes is that they may never be fully utilized for lack of sufficient fishing resources. However, errors of design and other technical and managerial problems have also resulted in heavy losses to the largest fishing company operating in the Paita Complex (PEPESCA), which was financed principally with public investment funds and which was in 1977 in the process of liquidation. 205. Finally, in the industrial sector the Government has acted with somewhat more caution and the level of investment has been relatively low. 4/ But even here the risks are high of not obtaining an adequate return :o the economy from a large portion of the investment. Besides the reasonabLy 1/ The Government also has argued that the economics of the project will improve by using the water to generate up to 270 MW of electricity. However, the Arequipa area will not need this additional electric capacity for the next 20 years and the overall project costs have become so staggering that the benefits of the power component will not much change the rate of return. 2/ For more details on public investment in mining, see Chapter IX. 3/ At current prices. 4/ See Chapter X for more details on public investment in industry. - 220 - successful investments to rationalize the production process at SIDERPERU, the Government has invested in four joint ventures with foreign companies to produce tractors, diesel engines, air compressors and machine tools. All four projects depend on heavy protection and the captive Andean Market. The latter, however, has not yet been assured and the four plants run a serious risk of working at extremely low capac:Lty with high production costs. 206. Investment efforts in the two main infrastructure sectors -- electric power and transportation -- have had taixed results. Much progress has been made in gradually integrating within ELECTROPERU the disjointed power system that existed before and in completing a road link between the Costa, Sierra and Selva. 1/ However, most of the projects have suffered serious delays and cost overruns, caused partly by the difficult and un- predictable topographic characteristics of the country 2/ and partly by limited execution capacity and excessively complex administrative procedures. In electricity, 3/4 of the 1968-76 investment was absorbed by the Mantaro Hydroelectric Project. The first phase of the project (to increase installed generating capacity by 342 MW) was recently corlpleted but suffered delays and cost overruns because of difficult geological conditions and the inexperience of the implementing agency. 3/ Moreover, the preparation of new hydro- and coal-fired projects to satisfy future requirements is seriously behind schedule; this may force the installation of diesel thertial plants towards the mid- 1980s. In transportation, the main problems were in the road construction program. 4/ The extreme topography forced a costly and time-consuming "trial and error" process in the construction of major roads through the Sierra. 5/ These problems were compounded by slow decision-making and inefficiency of the L1 The Lima-Pucallpa Road known also as Central Highway. 2/ On the topographic conditions of the country see Chapter I. 3/ The full Mantaro Hydroelectric project took four years to prepare and has been in execution for the last 10 years; completion is now projected for the mid 1980s. Costs have up to now more than tripled the initial estimates. 4/ Road construction accounted for about 70% of public investment expendi- tures (at constant prices) in transportation. Road maintenance accounted for another 12%. 5/ Particularly large problems were encountered in the La Oroya-Aguaytia portion of the Central Highway. - 221 - Ministry of Transport. 1/ Since the early 1960s, the experience acquired in road construction in Peru is enormous, but it has not yet been reflected in a swifter execution of projects. The return on road investments has been further limited by inadequate maintenance. 2/ Many roads have to be periodic- ally reconstructed at a high cost. Inadequate research prior to embarking on large projects has also affected the railways. After consolidating most of the existing railway systems within one public enterprise -- ENAFER --, the Government started an extensive rehabilitation program of the railways in 1972 Cor which around $50 million had been spent by the end of 1976. The rehabili- tation program was based on a study of the technical considerations to ensure i continued operations of the present railway network. However, the study did not review whether alternative modes of transportation could be more economical, nor did it establish the economic justification of the investments now underway. 207. The Government's investment strategy since 1968 has attached a lower priority than previous administrations to the social sectors. :mportant efforts were made in education, but they were constrained by inadequate plan- ning and execution capacity. In 1970 the Government started a ;-.prehensive reform of the education system, aimed principally at increasing the relevance of training. Important objectives of the reform were achieved (suc;l as the decentralization of the school system); others took much longer than expected (such as the development of new curricula for secondary education). A sizeable project was started in 1972 to expand and modernize school buildings. However, project execution was virtually paralyzed until 1975 because of the lag in developing the new curricula, without which the new I/ The inefficiency covers most of the process of road development from studies to construction. The following are some specific examples of the problems that have appeared. In 1974, a-contract for geological studies which are required to be able to design parts of the Huanjco- Aguaytia highway was supposed to have been awarded; by mid-1977, this had not yet been done and construction of those sections of the road had to be delayed indefinitely. Similarly, funds for repaving the Neshuia- Pucallpa highway have been available since 1974 but no action was taken for three years. Meanwhile, the road had deteriorated to such an extent that repaving was no longer possible; the highway will now have to be completely rebuilt. In both cases the delays have caused cost increases (rebuilding the Neshuia-Pucallpa road is at least three times as expensive as repaving) and the unnecessary payment of commission fees on foreign loans contracted to finance the road. Another source of delays, and cost increases, is the Ministry's slow compliance with previously agreed procurement procedures leading to lengthy and repeated efforts to develop acceptable procurement documents. 2/ Maintenance expenditures rose from $4.5 million in t198 to an average of $11 million in 1971-75. However, this is to be spread out in a network of 10.000 km of all weather roads and 30-40,000 kms of tracks. - 222 - school buildings could not be designed. Never:heless, even when this problem was resolved, the project was slowed down by a switch of the responsibility for school construction from the Ministry of Education to the Ministry of Housing, and by the exceedingly slow process o:- decision-making in general and of calling for bids in particular. By now, the costs have more than doubled compared to the original estimate of 1972. 208. There are various causes for the pubLic sector's low efficiency of resource use in its investment program. In the first place, the State spread out too thinly into too many sectors and tried to achieve spectacular results in too little time. The emphasis was on promol:ion of projects with too little attention given to the execution capacity of the agencies involved. Many of the new public enterprises were not yet prepared to perform the role that was thrust upon them. The predilection for very large projects, the push for a qu_ck involvement in the full detailed management responsibility of the projects, and the desire to introduce new but risky and unproven technologies further compounded the execution capacity probLems. 1/ 209. Another major cause of the low efficLency was the relatively little attention given to economic criteria in the decision-making process of many projects. Foremost were physical considerations of exploiting natural resources or satisfying a domestic need. 2/ The main objective was to increase units of output rather than minimizing the costs per unit of output and/or maximizing the benefits to the economy per unit invested. A premature heavy political backing of projects before their detailed designs and/or economic feasibility had been established resulted in hasty decisions and large cost miscalculations. 3/ In the petroleum I/ The bias in favor of large projects is evLdent in practically all sectors. The push for full management responsibilities occurred too fast, particularly in MINEROPERU, although in retrospect, the decision of the Government to stop temporarily the participation of outsiders in new petroleum explorations also meant :hat PETROPERU has had to carry an excessively large share of the e:xploration losses and that the exploration process was substantially slowed down. Examples of excessive risks on technology are the Cerro Verde Project and a bagasse newsprint plant (now being buil4: by INDUPERU, see Chapter X). 2/ The Bayovar Phosphate Mine and some of the investments in industry and agriculture are examples. 3/ The large irrigation projects in agriculture, the Cerro Verde project, the Bayovar complex, and the fishing complexes are examples of hasty decisions. Cost miscalculations have occurred in most of the projects. - 223 - sector, particularly, the Government took high risks and, admittedly,-- partly because of bad luck--did not attain the expected results. 1/ A distrust of traditional project evaluation criteria was a related problem. If a project aimed at attaining a social or political objec- tive, it was in many cases pursued, with little regard whether that was the least-cost solution. 2/ The economics of projects also suffered from the lack of cost-consciousness of many of the executing agencies. The policies of price controls and subsidies, the strong pressure to achieve visible results and heavy State interference in their management heavily eroded the accountability and the financial management responsibilities of puiblic enterprises. As financial deficits became chronic, the incentives of the enterprises to control costs lessened since the Treasury was expected to cover the deficits. Finally, the efficiency of the public investment program suffered from the inadequate phasing and programming of projects. An impor- tant example of this problem is the Chira-Plura Irrigation project. 210. These problems did not arise because of a lack of formal control mechanisms. On paper, an intricate and sophisticated mechanism exists for processing projects and coordinating the public investment program. Each executing agency or enterprise submits its investment proposals to the Sectoral Planning Office (SPO) of the Ministry overseeing operations in that sector. After establishing sectoral priorities, the SPO submits a prcposed investment program to the National Planning Office (INP). The latter reviews the aggregate set of projects to assure that they agree with the development goals laid down by the Government, that their feasibility has been established and that the investment program is financeable and consistent with the macro- economic realities of the country. Once approved by INP, a project goes back to the sectoral Ministry, which issues a resolution authorizing COFIDE to put together a financing plan. Since COFIDE takes a financial risk in the project, it is also supposed to review the feas'ibility of the project. Then the Ministry of Economy and Finance reviews the budgetary and foreign debt implications of the project. Finally, if foreign borrowing is needed, an authorizatiolL is requested from the Cabinet. The project review process is complica-ed at each stage-by a long series of coordinating committees and commissions. Furthermore, even once a project gets moving, an intricate system of cLear- ances, controls and cross-checking exists in all execution procedures (bidding, import of goods, etc.) involving many offices in various ministries and public agencies. 1/ On the other hand, it could be argued that the decision to construct the pipeline should have waited until more well-drilling had occurred. Likewise, the Government might have placed a larger share of the risk on private associates. 2/ The argument was that since growth and economic objectives were not the only objectives pursued, traditional evaluation criteria were not applicable. What is argued here is that regardless of the objective, a systematic evaluation of the project should have been made to ensure that the objective was being met at the lowest possible cost. A glaring example of this problem was the Majes project. - 224 - 211. For all its apparent thoroughness, in practice the project processing system has various flaws preventing its proper functioning. In the first place, most of the large projects tended to gain an irresistible momentum of their own after having been broadly identified. In some cases, the Government exerted its political weight behind the projects by passing legislation declar- ing them of national priority before studies had been concluded. 1/ In other cases, INP had little chance of influencing the project effectively since by the time it received the completed feasibility study, the institutions involved in the project had already committed large quantities of human resoulces to it and millions of dollars had been spent. In both cases the sector;3 found ways of bypassing the planning system which made the screening process irrele- vant in practice. Moreover, INP lacks the appraisal capacity to conduct a thorough evaluation. In 1976, its Projects Department consisted of 14 techni- cians who in theory had to review and coordinate all public investment proposals. A large amount of their time is spent on administrative functions so that little actual critical project analysis is possible. 212. COFIDE has a greater appraisal capacity and proved to be the most effective institution in reviewing projects critically. However, its influence is limited by various factors. COFIDE is involved only in the projects of public enterprises and not those of the Central Government. At the same time, since it intervenes at the tail-end of the project screening process, COFIDE receives projects that have ofi:en gained even more momentum than when they were submitted to INP. Finally, the largest projects exceed COFIDE's financial capacity and in supporting :hem it acts only as an agent of the Central Government, which reduces its scope for influence. 213. Another problematic aspect of the investment planning process that has caused problems is that INP s work remained divorced in practice from the macroeconomic and financial mangement which was done by the Ministry of Economy and Finance and the Central Bank. In short, IIiP drew up the spending plan after which it was the role of the Ministry of Economay to find ways of financing it. This was not considered an issue in the early .970s since the problem was mainly the lack of prepared projects and since the petroleum euphoria had raised high hopes of abundant availability of i-inancial resources. However, since 1973 the pace of project execution accelerated markedly while the financial capacity of the public sector and of the economy contracted sharply. Only since 1976 has INP tried to establish more effective priorities, al- though the proposed spending levels still tend to remain far beyond the 1/ Examples of this problem are the petroleum pipeline, Cerro Verde, Majes, Bayovar and the fisheries complexes. - 225 - financial realities of the country. 1/ A related problem is that what little analysis of the macrofinancial consequences is done, its horizon does not usually look more than one or two years in the future. The critical consideration is whether the fiscal and foreign exchange budgets can support the expenditures in that period. Much less attention is given to the medium-term implications of the investment decisions. However, in the large projects the bulk of the expenditures usually comes after the second year of execution. The result has been that in 1976 and 1977, even if no new projects were commenced, the demand for funds for ongoing p-rojects kept the expansionary pressures on the budget high and made it exceedingly difficult to control expenditures without stopping or massively slowing down projects. 214. The complex investment planning system neither succeeded in weeding out projects with little benefit to the economy, nor in keeping the aggregate public investment expenditures within manageable proportions, but it did create problems in project execution. The long series of authorization and clearances -- most of which were formalities more than a review of substance -- tied up an important share of the already weak administrative capacity of the executing agencies. At the same time they lead to costly delays in project execution. This is one more case in which the source of the problem :s not the objective of the controls (necessary screening and coordinating of investment), but rather the inefficiency of the control mechanisms. 2/ 215. One final issue of the investment strategy is that by concerntrating on large-scale projects in the commodity producing sectors it contributed little to the reduction of poverty and under-employment. The bulk of the investment was in capital-intensive activities and employment creation was not a criteria in the choice of projects. Moreover, those investments that might have generated permanent jobs were in projects of unlikely economic viability (such as the fisheries complexes and the Majes Irrigation Project) or that ran into serious implementation problems (such as the 1/ The budgetary proceedings of 1975 and 1976 illustrate this problem. Investment proposals by the SPOs for 1975-76 were in real terms double what was spent in 1973-74. However, during 1975 the sectors spent without a ceiling, limited only by their spending capacity, since the 1975-76 investment budget was approved only in late 1975. Moreover, the level of investment budget was rapidly surpassed with the approval of supplementary budgets. The lack of an effective system establishing spending priorities had made it very difficult to control aggregate expenditures since when faced with the choice of stoppi,rg a project or approving additional expenditures the Government usually opted for the latter. Similar problems have recurred in 1977, and an investment budget was not issued until the second half of the year. 2/ The problems of public management are reviewed in more detail later in this Chapter. - 226 - Chira-Piura Irrigation Project). Except for a small project for irrigation works and some limited investment by SINAMOS, 1/ the rural poor in the Sierra benefitted little from the investment program. Only in the first half of 1977 did some serious project preparation work get effectively underway in rural development projects. Although the Government has done much to improve the legal status and the organization of the urban poor, little investment funds and human resources have been allocated to benefit them directly. As mentioned above, the absolute amounts in real terms and the share of investment going to health, education and other basic social services declined. c. Conclusions 216. Since 1968 the Government has successfully increased the importance of public investment in the economy. Through this tool the State can now influence the direction and momentum of the development process significantly. The-efore, successes and failures in the public investment strategy and the degree of its efficiency of resource use have crucial implications for the financial, economic and social prospects of the country. Much experience has been gained in the last years, the basis of an institutional framework has been laid down and some important projects have been completed. None- theless, large errors have been made; some were the inevitable cost of the learning process, others could have been avoided. The public investment management and planning system was overwhelmecd by the increase in the level of investment, the number of projects and the diversity of sectors. It was, therefore, not successful in stopping projects that either did little to meet objectives pursued or exceeded the implementation capacity of the executing agencies. Probably one of the most important lessons from the past public investment experience is that it is equally or more important to ensure that maximum benefits are obtained from each sol invested than raising the level of capital formation in the economy. The relatively few accomplishments in the infrastructure sectors and the low priority attached to programs affecting the poor is another flaw of the public investment strategy. The Government is aware of these problems and is now in the process of reviewing its public investment strategy and the planning system that manages it. C. FINANCING OF INVESTMENT AND PUBLIC SAVINGS 217. Chapter V noted that there has been since the early 1960s a long- term deterioration in the savings performance of the economy. In periods of financial consolidation and export boom, the share of income saved increased, but these were short-term cyclical movements around a long-term trend of decline. In taking over a large share of the most productive assets in the economy and in raising its investment to over half of the country's fixed capital formation, the State has taken on a major responsibility for mobi- lizing a large share of the savings necessary to finance development. Un- fortunately, the State has been less than successful in achieving the latter. 1/ SINAMOS - National Service for Social Mobilization. - 227 - As noted in Chapter V, by trying to follow sharp anticyclical policies it has been able to lengthen the periods of rapid expansion of the economy, but has worsened the amplitude of the depression periods and has significantly contributed to the inadequate long-term savings performance of the economy. 218. The financial performance of the public sector has gone through two distinct periods since 1968. After the expansionary fiscal policies of the 1964-67 period ended in a major financial crisis in 1967-68, the Government embarked on a program of stabilization in 1968-70 with beneficial short-term consequences. The focus of the action was on the Central Government since it still accounted for the bulk of public financial flows. New tax measures combined with a sharp slow-down in the growth of current expenditures allowed the public sector to more than double the share of its revenues saved from less than 9% in 1968 to well over 19% in 1970 (Table 63). As a result, the share in GDP of public sector savings almost tripled from 1.8% in 1968 to 4.6% in 1970. Although public sector fixed investment in the three years was below the peak reached in 1966-67, it was still a moderate 4% of GDP. The share financed domestically rose drastically from around 1/3 in 1968 to 85% in in 1970. The good performance was all the more remarkable since there were the years when the Government was formally initiating its strategy of sharp structural changes in the ownership patterns and development objectives of the economy. In this period the Government policy reflected a keen aware- ness that the long-term viability of the tranformations in the economy demanded that it be as much as possible self-financed, with the distributions in cash flow and operations in some sectors made up largely by surpluses in other sectors. 219. Unfortunately, the financial improvements were short-lived since in 1971 a deterioration started once again that was to gain momentum in :.973. While the public investment boom took-off in 1971 and reached over 8% of GDP and half of fixed capital formation in 1974-75, public investabLe sav:.ngs collapsed and became negative in 1975 (Table 63). The result was that: the over- all deficit of the public sector as a share of GDP rose every year from 0.8% in 1970 to an all-time high of 10% in 1975. Although grace periods on the massive foreign borrowing contracted allowed the burden of foreign debt amorti- zations on public sector finances to decline temporarily, the total financial requirements rose to almost 12% of GDP in 1975. 1/ This extremely wice public sector gap was all the more untenable since there was little room to finance it with non-inflationary domestic borrowing, for in the same period private savings were also declining. Massive recourse was made first to foreign borrowing whose gross disbursements rose from 3.2% to 7.6% of GDP between 1970 and 1974. As even this became insufficient, the pattern of financing turned in 1975-76 to domestic borrowing, a large share originating in the Central Bank (Table 63). 1/ Total financial requirements are equal to overall deficits plus amorti- zation of foreign debt. The latter, as a share of GDP, increased from 2% in 1969-70 to 2.5% in 1973 but then fell to 1.9% in 1974-76 (Table 63). - 228 - 63abi G1: IND1CATOIOJ 0CF I'lJ1i1IC Cl3EMP 'INAIC71, 1968376 i968 1969 '1970 17 1972 1973 1974 1'975 1976 A. OPZ9&TIO,S AS PF TCi;'r OF GDP Current Revenues _/ 20.7 23.1 23.8 23.0 24.2 24.14 32.1 31.7 33.1 Current Expenditures - 19.3 19.1 l9.4 20.2 21.5 23.3 29.8 32.7 32.8 Investi'ole Savings 2 1.8 4.5 4.6 3.2 2.9 1.2 2.9 -0.9 0.9 Capital Expenditures 5,4 5.7 5.4 5.6 6.o 6.7 9.2 9.1 8.9 (of which fixed investment) (3.71) (4.0) (44.0) (4.5) (4.9) (5.5) 8.1) (8.2) (7.6) Overail Ealar.ce -3.6 -1.2 -0.8 -2.4 -3.1 -5.5 -6.3 -10.0 -8.0O Amortizaticn of Fcreign Debt - -2.4 -2.0 -2.0 -2.4 -2.3 -2.5 -1.9 -1.9 -1.9 TJtal Funds Miquired -6.0 -3.2 -3.8 -J4.8 -5.'4 -8.0 -8.2 -11.9 -9.9 FDrcign. Tinancing, gross 3/ 3.6 3.6 3.2 3.1 3.8 5.8 7.6 7.4 4.9 Co.t.estic :?_ ....... cina, net 2.4 -0.'4 -, _,4 1.7 1.6 2.2 ,c 6 3.5 5.3 3. S-RtUCTV'R CF FTNAOcIUG CFs ONV'dSClOT '(3F T(TAL) Capital Expenditures 100 0o :30 1'00 100 0C0 100 100 1Cr) r.nrestable Savings / , 34 79 85 58 49 18 32 -10 9 'let foreign :irancirng 22 27 23 L2 26 50 61 60 33 Liet iczescic financing 44 6 -8 30 25 32 T 50 5T C. "'Cr RM'.MO I1tMS Investable Savings % Revenues _/ 8,9 19.5 19.2 13.9 12.1 4.8 9.3 -2.8 2.5 Tr.nvozt.blae ?ublic Savings ' of r!:S 21 14.0 33.7 28.7 22.1 21.0 9.4 26.8 -12.0 10.2 0zers11 deficit I Expenditures 14.5 4,8 2,6 9,1 11.J l18, 16.0 23.9 19.5 ' / 0ur. nc re.,,u,s inclu-le ,r- . saIs r mahlic enterpri_ces sadd current expenditures include 'roto out.ajs of public enterprises. 2/ Investable savings = Current account balEance and capi *l revenues. Li in .973 and 19714 excludes pre-rayments. .- Re" eminis" include curretit and capital revenues but exclude borroving. c/ -,IT -- rcs:; \:dLti'oral Sasving9. ':;turoes: Appendix Table 2.1, 5.1, and 5.7. As analyzed in Chapter V, this imbalance in public sector accounts was one of the key causes of the serious financial and economic crisis Peru is now going through and the extremely heavy borrowing that was incurred has mortgaged a large share of the limited future benefits thac: may be reaped from the invest- ment made. 220. The financial problems of the public sector during 1971-76 were partly due to the rapid growth of investment, although the main cause was the inadequate savings effort. As discussed earlier, public investment in 1971-76 probably expanded well in excess of project management capacity of the public - 229 - sector and the efficiency of resource use was low. However, the hypothetical illustrations in Table 64 suggest that strictly from a financial point of view the level of investment could have been achieved without causing major disruptions to the country's finances if a moderate savings effort had been made. The retrospective calculations may appear somewhat academic, but they are useful in highlighting the relatively small effort that would have to be made to restore equilibrium to the public sector finances. Hypothesis II in Table 8 shows that if the share of public sector revenues saved had been 15% in 1971-76 instead of the actual 12% the overall deficit would have remained below 4.5% of GDP, clearly a manageable level. 1/ This level of savings would have Table 64: MLLUSTIRATION OF no-PACT OF IGI LEV= OF SAVINGS O'S ?'JLIC SEOCT^ F7naczs, 1968-76 Actual a y o t h e t i r a 1 r9,0-7O L971 1972 12973 LZ74 L575 19?6 L97:-76 : ' ?_ib2.io Thveszab2.e Sravi = 3.5, of :ovases . ;e Say'i-s I Reverues L3 25 1". 14 :1 '2 ' 12 :av:c--Ie Sa Ci-gs S Capital Expet--Itures a2- 63 5S 53 33 39 39 43 Ove-:': Iei fc_ % of GD? 0.9 2.1 2.5 3.2 5.7 5.6 2.6 4.6 _ -- overL1 d.e'ficit between actual - as 's o? GS 2/ - 0.3 o.6 2.3 0.6 4.4 2.6 2.2 - :'" ior. dI 76s - -2 40 212 0 61? 375 1 - 197L-76 acc =ulated as % of foreigp debt 3 37 : -: : 'c tn-vestLe Savigs = 5S of Revenues. :--res:.ble Sa-r.ns t 5D? 4.6 3.5 3.6 3.7 1. .3 5.2 4.5 Mve;.S1b.- 5av inFg % Capital Epeoditu-es 82 63 61 , 55 53 52 5c 56 .ver-3_ Deficit % of 00DP 0.9 2.12 2.4 3.0 4.3 4.3 3.0 3.6 sifere=." _a . v'era1 d.eafcit betveen actuaL -4~ ; 7f a 2? - 0.3 0.7 2.5 2.o 5.7 .T . 3.2 - -ir''io- dollars. / 2.2. 53 23C 232 799 593 192t - r9_76 acc, =eated as S of fore_.gn debt - 5 2. Ass_ess t^a: oob3ic ioest.ble savt-gs 'at:d save reacled 3.5% of GOP durisng :971-?6 compared to 5.6% of GOP in 2/ Cecverte_ a: a-rerage official excharge rates for the 7ear. 3/ ?ore : debt oastaoding and disbursed. TI/ Assces thpt ;tblia investable savings would have reiLched 15% of revenues duri_, 21971-76 com=ared to 1;. eff reve:nes in 1965-70. Sown:es: Tab1e 7, Appendi.x abla 2.1, 4.'., 5.7, 5.9 end Mission estimates. 1/ The share of public sector capital expenditures financed by investable savings during 1971-76 would have been 56% instead of 16% (Table 64). - 230 - been still well below the 19% of revenues achieved in 1969-70. Even if only half of these savings had been translated into less foreign borrowing, the external debt (outstanding and disbursed) wou:d have been at least 25% lower by the end of 1976. Hypothesis I shows that even with a more modest public savings performance of 3.5% of GDP during 1971.-76 -- well below the 4.6% attained in 1969-70 but more than double the L.3% actually achieved --, financing the same level of investment would have required large but still manageable levels of foreign and domestic borrowing. The paragraphs below review in some detail the causes of low savings and, in exploring some important avenues for a long-term permanent improvement, show why the above hypothesis would have been well within the realm of the feasible. a. Public Enterprises as Savers 221. Public enterprises have become the !aost important group within the public sector both in terms of the share of the country's resources they manage and in terms of the key sectors of the economy they control. A major handi- cap in evaluating their performance is the lack of information on their operations. The analysis in this section is based on partial information from a sample of the largest enterprises. This should be sufficient to identify orders of magnitude, trends and main issues, but is subject to error and should be interpreted with caution. Developing a more efficient flow of detailed information from the enterprises to the central authorities should be given high priority, if monitoring of their operations is to become adequate. 222. Table 65 shows that the financial performance of the public enter- prises has sharply deteriorated since 1973. Between 1968 and 1972 the public enterprises were still relatively unimportant but as a whole their financial position improved: their own investable savinigs 1/ were equivalent to around 20% of revenues and financed around half of capital expenditures. In 1973-75 the volume of their operations expanded rapidly and the share in GDP of their capital expenditures tripled to 5% in 1975. EIowever, at the same time their own aggregate savings became sharply negative. 2/ All the investment, there- fore, had to be financed with Central Government transfers and foreign and domestic borrowing. 3/ In 1976, an important improvement was achieved, although the enterprises' own savings remained at less than 5% of their revenues and financed only 16% of capital expenditures. 223. The main causes of this unsatisfactory financial performance can be grouped into three issues: (i) price policies; (ii) quality of financial management; and (iii) the special situation of the fishing industry. 1/ Own investable savings defined as currenl: account balance plus capital revenues net of government transfers. 2/ As a share of GDP, own investable savingsi of the public enterprises fell from around 0.9% in 1970-72 to -0.3% in L973-74 and -2.8% in 1975. (Table 65). 3/ See Table 65. - 231 - UKble 65: MrDCATOPS OF F=A5CE OF SPtJC 3:¢R;S=, 196I-76 !968 1969 1970 1971 197E2 1973 19-4 1.35 19-5 A. O?PATtI0NS AS 2F.C OF GDF Cvo Cxr-en.t Revenues 2 1.*8 3.7 1.0 4.4 5.1 5.a 1.8 13.S 16.2 " r-e=t Expenditures 1.9 2.9 3.1 3.8 4.3 6.2 12 . . 16.0 Cn Capi,'t ?eveues 1/ _ 0.1 0.1 0.2 0.2 O.i .6 ;. 0.5 I3vestat~_e sav_ngs -' 2/ 0. 0.9 1.0 0.8 1.0 -0.3 -C.3 -2.8 c., 0er.trvl Sovernrrt Trp-nsfers - c.S 0.5 o.6 0.9 0.8 1.3 C.. 2.1 1.3 Total &:nvestable Saviags 0.5 2..1 1.6 1.7 1.8 1.0 0.4 -0.7 2.0 Cajita. Sxoenditures 1o.5 2.3 2.0 1.5 1.7 3 2 1.7 ' 5.' 1.7 (Of "rnic; ~ixed invest_ett) tl.3) (2.0) (1.5) (-.2) (1.6) (2.8) Overall be' wece -1.0 -C.3 -0.4 0.2 0.1 -2. 2 4.3 5.3 -2 .7 Foreign Fisa.rciog, "let 0.3 1.0 0.° 0.3 0.7 2 1.3 az-est&c Firar~cirg, .:et O 0.7 -0.1 -0.3 -1.3 0 -.8 0 . B. STJCaZ.' .? 0 03 :E3TrN , CF c ) ra,ita1 lxnensdit'res iOO 100 'CO 100 100 * :C '0 0co %n '-zreste73t Sar'.gs 1/ -4 34 17 53 59 16 Ce ,:al 0Soveo--et 7ransfers 2/ .3 23 29 65 5c :-5 22.t For-ig_ 7i.aci3g 24 41 33 5 4 2 5 25 Niet ::estic Fi czor,6 42 2 -S -74 -, 3 8 28 C. MDtC.11L:T'0 e 0rur nvestle savsngs % 7eflue -2.7 17.4 24.0 1-.6 1?.2 -s. -3.G -23.2 4.4 Overel' 2-alace as , of Epeadit-es -43'.2 -28.9 -21.3 -1'.2 -i.7 -3 5 -29.2 -0.0 _19.' 1 j %cludcs Central ;overrcn* Transf!ers . 2/ lZ.cludes c-_rr--t and cemital trasfers. 7 rC. >-4eS !urrent 1.C campital expendit-.xe:: but excludes dlebt servri-:e. Sc-r-es: Apzenaix Tables 2.1, 5.2., ar.d-3.6. i. Price Policies 224. Although the stated objectives of the Government have been that the public enterprises should be fina-ncially independent with their prices reflecting costs, the actual policies followed have been different. with the objective of insulaving the domestic economy from foreign price fluctuations and to dampen inflationary pressures, the Government has followed a relatively inflexoble policy of price controls. The group of enterprises most seriously affected by the price controls have been those involved in marketing c petroleum (PETROPERU) and foodstuffs (EPSA). Domestic prices of petro:..- products were already among the lowest in Latin America in 1973. 1/ IhTO had limited PETROPERU's capacity to generate savings, but the latter were -or ->. less sufficiently large to finance almost 90% of the company's invest-Ment `n 1/ See chapter I for a discussion of the role of petroleum pricing poli2.o : the overall context of energy policy. - 232 - 1971-72. 1/ However, when the world price of petroleum quadrupled in 1973, the Government failed to adjust domestic prices (Table 66). Since at the same time the country's dependence on imported oil was increasing, PETROPERU's small surplus between income from sales and costs of production became an ever-growing deficit that reached almost 1% of GDP in 1975. 2/ In June 1975, the Government made the first important changes in the prices of gasoline and other derivatives (Table 66). However, the weighted average of price changes increased PETROPERU's revenues by only 37%, while a 57% increase would have been required to eliminate the losses in petroleum marketing. 3/ Finally, during 1976 the prices were increased sufficiently to close PETROPERU's current account gap, although a substantial economic subsidy remained. 4/ The solution of the financial problem of PETROPERU was only temporary, however, since with the crawling devaluation of the sol after September 1976, the costs in soles of imported petroleum rose and a current deficit emerged again. In dollar terms, the price per gallon of regular gasoline fell from $0.77 in September 1976 to $0.66 in May 1977 (Table 66). In June and October 1977 the Government increased prices to more realistic levels. However, the prices are still lower than in most other Latin American countries. The sharp fall in PETROPERU's savings since 1973 has coincided with enormous investments in petroleum explorations and the Trans-Andean Pipeline, which had to be financed almost completely by massive foreign borrowing. 5/ 1/ PETROPERU's investment expenditures in 1971-72 (at current prices) totalled S/.2.5 billion of which S/.2.2 billion (or 87%) were financed by internally generated savings. 2/ The following table shows the magnitude of PETROPERU's losses during 1974- 75 marketing petroleum products (in perceat of GDP): 1973 1974 1975 Incomes from saleS 1.9 2.4 2.8 costs L/ 1.8 3.0 3.7 Balance O.i O. -0.9 1/ Includes costs of production of P-71t0PERU's pdtroleum, purclhases of crude and refi:ning '.nd marketing costs. Source: PIE2ROP8U and Mission e-tint ites. 3/ The June 1975 price changes increased the average price per barrel of PETROPERU's sales from $6.7 to $9.2. In :he same period, the average cost per barrel was $10.5. 4/ Economic subsidy defined as the difference between the world price of crude plus transport and refining costs and the actual sales price. 5/ The disbursements of foreign loans required to finance PETROPERU's investment accounted for 44% of total project-related foreign loan disbursements during 1972-76 (Appendix Table 4.5). - 233 -. tsable S6i: FE'M3 PSU 8AL= ?:-_CES 0x? S=C= F.ET 2I0C'0, 1967-77 Dec, Je June OLec. J=re ja. J,e :"ae 0cc.2' :c:. 1967 1968 1969 1974 1:975 I976 559X6 2977 :97- _9? in soles er gliorn 6.B6 1o.o0 9.00 9.00 125.00 23.C3 50.00 7' 00 75-.30 Gav:0':?e (95 oct.) 9.66 16.oo i6.Co 15.00 3C.00 3C.'0 4s.00, SO..o 80.20 30.2 K:tar~S6ez 1,40 4.40 4.4°0 4'4o 5.50 6.20 7.44 'O0.00 i 0.3C 2..2 Oi=z. 3s. 2 4.43 4.33 4.33 4 ,R 5.40 5.85 7.15 .7.00 2Z._, 2__ .:' i:'70. 5 3.10 3.10 3.0 3.10 4.0 5,50 506 '.3 s.0.o F -e :;o. 6 2.68 2.68 2.68 2.68 3.'0 5.00 6.oo *1.03 _..30 '4.,^ In co cents loer ga21oa aas:).I ne (o3 Oct.) - l 26 23 23 33 51 77 94 93 as3inr.s (95 oct.) 25 41 41 39 67 93 123 99 11 11 11 - 12 14 2. 2'2 0C _i-sael .; . 2 121 11 11 11 12 ' 3 11 27 22 .hel 2i2 ' :,o. 5 B 8 8 8 9 2.1 0 20 16 Fl:e2.Oi' Nc. 6 7 7 7 7 8 U1 9 is 2- .-_U-ze: MX istr7 of :ergy az3. MILnes, Minijstry of 2conor aad Fianr.ce, an Mtis"ior. is-imates. 225. The deficits of PETROPERU are the single priacipal cause for the financial difficulties of the public sector in recent years. An important indicator of the savings potential through adequate petroleum pricing is that if PETROPERU had been allowed to adjust its prices to international levels in 1975, the investable savings of the consolidated public sector as a share of GDP would have been 1% instead of a negative 0.9% and the overall deficit 8% instead of 10%. 1/ 226. ' Chapter I argues the urgent need to curtail the growth of energy consumption through more adequate pricing policies. Besides the finan:ial and energy considerations, there is little social justification for subsidizing the prices of petroleum products. The 1971-72 ENCA survey showed that 70% of gasoline is consumed by the wealthiest decile of Lima's population. By transferring income to the richer groups, the gasoline subsidy goes sharply against the efforts of the Government to distribute income more widely,. In fact, as discussed in more detail in Chapter VII, this report strongly recom- mends that besides adjusting domestic prices to international levels, a stiff consumption tax should be imposed on gasoline. 1/ The average price per barrel of PETROPERU's sales was $8 in DecembLer 1975. The economic price (reflecting the international price of crude plus costs of refinery, transportation and marketing) was $14.3 or 78% more. Had revenues been 78% higher, the operating surplus would have been equivalent to 1% of GDP instead of the 0.9% deficit incurred. - 234 - 227. The losses of EPSA, EPCHAP and ENCI in the marketing of foodstuffs and fertilizers 1/ was the second largest cause for the low savings perfor- mance of public enterprises. To protect consuLmers from the sharp increase in prices of foodstuffs in 1973-74, the GoverrLment decided to subsidize prices. By the end of 1974, domestic sales prices covered less than 60% of the import costs of wheat, rice and vegetable oils (Table 67). During 1975 and 1976 domestic prices were increased sharply several times to try to reduce losses. 2/ Nonetheless, as in the case of petroleum products, these increases were not sufficient to catch up with the import costs in soles which kept rising with the exchange rate devaluations. 3/ Since the middle of 1975, prices of fertilizer have also been heavily subsidized to stimulate agricul- tural production. 4/ As a result, the losses incurred by the domestic market- ing enterprises remained at almost 1% of GDP during 1974-76. The subsidies on the consumption of essential foodstuffs may have had a limited justification as a short-term measure to isolate the economy from the fluctuations of world prices. It is doubtful, however, whether they contributed much to dampening inflation. The temporary slowdown of price grcwth was offset by the inflation- ary impact of the financing of the subsidies. However, in the long-run, there are other reasons, besides financial considerations, that argue against consumer subsidies as they have been applied in Peru. A first problem is that the subsidies favor mainly urban consumers and do not discriminate between income groups. Although wheat and ric.e are important components of the diet of the urban poor, the largest share of the volume is consumed by the middle and upper income groups who, therefore, absorb most of the subsidies. Thus, the income transfer to the poor intendec. by the subsidy is very cost- ineffective. A second problem is that price c.ontrols have discouraged domestic production and may have encouraged an increase, in the import-intensity of domestic consumption. The artificial lowering of the price of wheat has caused a shift in the pattern of consumption away from other food substitutes 1/ Wheat, rice, vegetable oils and fertilizers account for the bulk of the subsidies. 2/ The table below illustrates the large increase in foodstuff prices in soles in 1975 and 1976 (prices before June 1975 - 100): Before June Jane Jan. July 1975 1975 1TI6 1976 Wheat flour 100 128 150 226 Rice 100 100 124 150 Yeget&ble oil 100 125 125 166 Source: Kission estinates from vorkir.2 sheFt orenered by the Ministry of iconomy end Finance. 3/ See Table 67. 4/ Imported fertilizers are subsidized by 4C% (through ENCI). - 235 - of equal nutritional value but which the country has a considerable potential to produce. For example, the production of quinua, which is one of :he few cultivable crops at an altitude over 3,000 meters, has practically d:Lsappeared because of the distortion in relative prices created by the price subsidy on wheat. Table 67: WuGII oTIMATI OI SUWIDY OF MICTSOF or ssoe :ODSTUFTF, 1974-76 (actual, priev as percent of cont) December tecuar July 19T1 1975 1976 Wheat 57 74 T7 Rice 60 69 85 Vegetable oils 59 45 50 Sourcee Mission estirlmtcs from workiac ctict pr:parcd by thc e-niutzy Of Economy awd Finance. 228. Until 1975, the public utilities companies vere also seriously affected by the price controls policy, although aggregate magnitudes of their operational financial deficits were small compared with those of PETROPERU and of the main marketing companies. Public enterprises in power, 1/ water supply and sewerage, 2/ and transportation 3/ all suffered large declines in operating surpluses between 1970 and 1975 because of the unchanged tariffs against rapidly growing operational costs. This constrained their investment capacity and, in the case of ELECTROPERU, large transfers from the Central Government were required to finance its investment. In 1975, the Government started to adjust tariffs to gradually bring t:hem in line with operational costs and to cover development costs. This has markedly improved the financial position of these companies in 1976. 229. In sum, the problems created by the past pattern of price control and subsidies seem to outweigh by far the benefits that may have been attained. Controlling prices may work for a short period in dampening inflationary -res- sures but in the longer run the distortions in relative prices is not condu- cive to an efficient allocation of resources in the economy. Allowing, the public enterprises to i?dopt pricing policies that reflect their costs and the economic value of their goods and services could increase public sector savings by at least 3-4% of GDP. 4/ Whatever specific objectives are pursued with price subsidies (income distribution, for example) they can be achieved 1/ ELECTROPERU and ELECTROLI1MA. 2/ ESAL and ESAR. 3/ ENAFER (railways) and ENAPU (ports). 4/ This, of course, assumes that parallel measures to ensure cost- consciousness and efficiency are taken. - 236 - more effectively and at a lower cost with direct subsidy programs that could be easily financed with the additional revenues generated by more adequate pricing policies. ii. Financial Management 230. The second major cause for the low efficiency of resource use and the small generation of economic surplus by public enterprises is the general weakness of their financial management and the lack of cost-consciousness in their operations. All countries that have developed a large network of public enterprises have been faced with the issue of how to manage them to assure that operational efficiency is maximized but that company objectives do not run counter to national economic goals. Most public enterprises should achieve financial independence; the Government has reiterated this objective on various occasions in recent years. However, even if this is achieved, the efficiency of resource use could remain low because of the monopoly position most enterprises enjoy in the domestic market. Therefore, Government controls are needed to set down performance criteria for their operational efficiency and their investments. 231. A wide range of controls has been established, but these may have done more harm than good in pursuing the objective of efficiency of resource use. The most potent instrument of Government intervention has been through the price controls discussed above. By forcing the enterprises to accumulate operational losses and by not requiring any performance conditions, the Government has eroded cost-consciousness and financial management. There is little incentive to cut costs and target for a maximum rate of return on in- vestments if the enterprise is certain that the losses incurred will be covered by the Treasury. Even if the subsidies were a cost-effective alternative to attain a-Government objective, their accounting should be clearly separated from the company accounts. Other existing Government controls add little to an effective monitoring but do curtail excessively the company's autonomy for efficient management. The Government not only intervenes in the appointment of board members and top and middle level staff, but it also restricts the autonomy of boards to approve contracts and screens all applications for business travel abroad of the enterprise's staEf. Excessive Government intervention in such detailed internal affairs of the enterprises has also made it difficult to attract qualified staff to top managerial positions. On the other hand, efficiency criteria have been set for none of the firms, and as mentioned above there is a dearth of information needed to assess efficiency in terms of the objectives for which the enterprises have been established. 1/ 1/ Much information is requested from the en:erprises by the Government. Budgets must be submitted to the Ministry of Economy and Finance (in 21 forms) and quarterly reporting requires aniother ten forms. The problem is that the information is provided with Large time-lags, is of low quality, and is not very useful for effec:ive monitoring purposes. - 237 - 232. There is much that Peru could learn from the successes and failures in managing public enterprises in countries of Western and Eastern Europe. 1/ In general, this report recommends that a simpler but more effective system of monitoring be established. The objectives of the enterprises and a set of criteria to judge the performance in achieving these objectives should be clearly established. For the commodity-producing activities which substitute imports but have a monopoly in the domestic market, iaternational prices should be used to judge the degree of efficiency. If the differences between local and foreign prices is substantial, the Government can decide whether the benefits pursued by the enterprise are worth the economic losses in- vulved. If this were the case, the Government can set realistic performance targets for the enterprise to reduce costs; if these are not achieved, the Government would have valuable information in its hands (which it does not have now) to decide whether to allow higher prices, to reshuffle the composi- tion of the Board and of management or to liquidate the public enterprise. Equally simple and effective performance criteria can be set for public enterprises in public utilities and marketing. Once efficiency criteria are set and the performance targets are agreed upon, boards and management should be given far greater freedom of action and responsibility. Control devices should be the periodical comparison of performance against the fixed effi- ciency criteria and targets, rather than the plethora of restrictive idevices presently in use. iii. The Fishing Industry 233. One last cause for the large operating losses of public enterprises in 1973-75 was PESCAPERU. When the fishmeal industry was nationalized and PESCAPERU was created, the industry was in very bad economic condition. The installed fishing and processing capacity was excessively large and manv firms were on the verge of bankruptcy. PESCAPERU reduced the installed capacity drastically but was unable to cut the excessive labor force (at least 10,000 fishermen on the payroll were idle). In addition, PESCAPERU had to make large cash payments to the previous owners and had to take over a large amount of short-term obligations that the private companies had accumulated. Al.though these short-term debts were consolidated into medium-term obligations, they remained a major drain on the enterprise's finances. The continued adverse ecological conditions limiting anchovy fishing compounded PESCAPERU's problems. As a result, the company incurred heavy losses in 1973-75, estimated at around 0.3% of GDP. In June 1976, the Government decided to return the fishing operations to the private sector, thereby removing the overstaffing, one of PESCAPERU's major burdens. Although bad fishing conditions continue to weaken the company's financial performance, it should not incur major losses in future years. 1/ Particularly France, Italy, U.K. and Spain in Western Europe and Hungary and Yugoslavia in Eastern Europe. - 238 - Trble 68: IXDIACnuS OF SOCALSECURCtITY FINMICaZ, 1968-76. i968 _ ro 1973 1Q74 1975 1976 Contrifbutons to .ocial Security as: Percent of GDP 1.9 1.9 2.4 3.3 3.0 2.3 Percent of General Gov't tax revenues 12 :.2 15 19 13 16 Investable Savings of Social Security as: Percent of GDP 0.5 0.,7 0.9 1.5 1.4 1.0 Percent of ?ublic Sector rnvest. Se'r. 26 ;.6 T4 50 152 16 Overall 2alance of Social Security as: Percent of GDP -o.4 o 4 0.8 1.0 1.3 0.9 Percent of Public Sector balance: -10 8 12 11 13 10 1/ Public sector investable savings were negative in 1.975. Social security savings vcre 1.52 times the overall public sector cegatime saviz:s. Source: Appendix Tables 5.9 and 5.21 b. The Social Security System as Saver 234. Since 1970, the Social Security Administrations (SSA) has been the only public sector institutions has consistently generating a growing amount of savings. These savings have been channeleci through the banking system to help finance the large deficits of the rest of the public sector. 1/ Table 68 shows some indicators of SSA's savings. The large surpluses have resulted mainly from the lag between the growth of contributions to the pension fund and actual retirement of workers. 2/ However, unless the financial policies of the SSA 6fctated by the Government are charged in the near future, its cash position will turn from the present surplus to a rapidly growing deficit during the next decade. The SSA has estimated that this may occur as early as 1980. 235. The financial structure of the SSA is based on the principle of capitalization of contributions, where the actuarial calculations assume that the assets of the SSA will accumulate in real terms to allow it to pay services and pensions in the future. The problem is that the opposite is occurring. By requiring the SSA to maintain the bulk of its funds in fixed 1/ The Social Security Administration was reformed in 1973. It consists of three main system: (1) the health insurance for blue and while collar workers (about 65% of revenues); (ii) the,national pension system (45% of revenues) and (iii) national insurance system against work-related accidents and sicknesses. 2/ In 1974 1,300,000 persons contributed to the pension fund, but only 70,000 beneficiaries received payments. - 239 - term deposits in the Banco de 'a Nacion, the Government is gradually decapi- talizing it. 1/ Unless the SSA is allowed to place its funds at substantiall, positive real yields, its financial collapse is inevitable. 2/ If the present cash reserves do not grow in real terms, they will soon be drained by demands from a rapidly growing number of beneficiaries. Moreover, there are plans now under study to extend the coverage of the social security benefits to the families of workers and to rural areas. 3/ While these objectives are import- ant elements of a strategy to reduce poverty, they would further increase the deficit of the SSA under present financial policies. 4/ 236. The present pattern of financing of the SSA has other negative con- sequences on the Government's development objectives. The main ones are that the rate structure is highly regressive with respect to incomes and the burden that social security payments place on payrolls favor capital inteasive 1/ Most of the financial assets of the SSA are placed in fixed term deposits in the Banco de la Nacion. These deposits earned only an annual interest of 7% up to mid-1976. A small share is deposited in the Banco Hipotecario and the Caja de Ahorro de Lima earning about 9% per year. With the general increase of interest rates decreed in 1976-77 the yield to the SSA has been increased to about 12%, which is still far below the 40-45% yearly inflation of the last two years. The SSA does little fixed investment. However, given the low yield of their financial placements, some of their real estate investments (such as a large hotel/office complex at Lima, the "Centro Civico") may yield higher results. 2/ In fact by depositing all its funds in assets bearing negative real interest rates, the SSA is transferring resources to the Banco de. la Nacion, and through it, to the rest of the public sector. Some con- sideration has been given to allowing the SSA to place a share of its funds in Central Government investment bonds. Although the yield paid on these bonds is higher than those of Banco de la Nacion deposits, they are still well below the inflation rate. Thus, this would not be an adequate long term solution. 3/ At present the SSA covers only workers themselves (and not their families) in urban areas. 4/ Related to the financial problems, have been the difficulties found in setting an efficient administrative arrangement for the collection of contributions and the serious delay with which audited accounts of the SSA are completed. (in April 1976, the audited balance sheets for 1973 of the SSA were still not completed.) Although in 1973 the existing social security systems were unified into one, the contributors are still identified in 2 or 3 separate accounts, one for each service pro- vided. (Part of the problems found have been difficulties with the computerized data processing systems.) This makes difficult the pay- ment by the contributor and complex the accounting in the SSA. It also prevents a close control of evaders. - 240 - Tabic 69: PERU: FSTIYATED CONTRIBUTIONS TO THE SOCIAL SECURITY StSTT2.M (percent of selary) vonthlyi cAlary or Wages Soles 10,000 20,000 50,soo 75,0CO 100,000 Dollars 1/ 142 285 T15 l,OT7 1,421 5lue-coller 16.5 i6.5 15.3 13.4 12.3 Hes-Ith insurance 2/ 9.0 9.0 9.0 9.0 9.0 Pension fand 3/ 7.5 7.5 6.3 4.4 3.3 Viite-collar - Public sector 13.0 'l.9 8.3 5.7 4.3 He-Qth ins-"rance 4/ 5.5 I,4 2.0 1.3 1.0 Ponsion fNo1i 3/ 7.5 3.5 6.3 14.4 3.3 White-.co1a.r - Private sector i4.0 12.7 3.6 6.0 4.5 Health insurance 5/ 5 5.2 2.3 7 1.2 Penzsion fund 3/ 7.S 7.5 6.3 4.4 3.3 1/ Converted at SI. 70 = $1 9 ss of salary (3/4 tayed by employer). 3/ Zmployer cays 5% Up to a maxiium salary of S/.37,500; emplo/ee pays 2.5 .p to a taximum salary of SI. 56,250. 4/ Total assessment is 5.50,arounad half by the employer. The employer's contribution i.s limited to the first S/.12,000/ncath, the employee'sl contribution is 1imited to the first S/.24,000/month. 5_' Same as in footnote 4 except that total contribution iii 6.5%. Source: Soiial Security Adminiatration and ,-ission estimates. economic activity. Table 69 illustrates the regressive impact of the rate structure. A blue collar worker with an income of SI. 10,000/month is assessed 16.5% for contributions to the SSA, while a public sector employee with an income of SI. 75,000/month is assessed only 5.7%. The main causes for this are the blue-collar workers are charged a higher rate thaa white- collar workers for health insurance and there are ceilings on the salaries on which assessments for the pension fund and the bl-ue-collar health insurance are based. 1/ Even if differential treatment of blue-collar and white- collar workers in health contributions were justified because of the higher health risks of the blue-collar workers, it is difficult to justify on equity grounds. Eliminating this discrimination and the macimum ceilings would in- crease revenues and would have a powerful redistributive impact by shifting the cost burden to higher income workers. The high rate on low income blue collar workers increases labor costs for positions which most of the poor are seeking. More generally, it is questionable whether the financing of the social security system should be based on payroll ta:ces in a country like Peru, where the growth of the labor force has been exceeding the capacity of the 1/ The ceiling is a common feature in countries where the social security system works on a capitalization basis; it has niegative income distri- bution effects. - 241 - economy to create jobs. Adding the payroll tax and the contributions to the pensions of the Government to the social security payments, taxes on payroll already amount to over 4% of GDP. 1/ 237. In sum, the SSA is rapidly becoming a very important part of the public sector and now generates a high share of the economy's savings. What occurs to the SSA will be one of the key variables in determining wheth the public sector and the economy are successful in increasing their savings rates. With the present financial arrangements, prospects are not good. In the short run, it is therefore urgent to take measures that allow the SSA to achieve a high return on its investments in real terms. In the longer run, would be advisable to thoroughly review the pattern of financing of the SSA and to eliminate the distortions in resource allocation created by the taxa- tion of labor income. Consideration should also be given to financing the social security system out of general tax revenues, although this could only be done after a reform of the tax system guarantees a strong built-in elasticity. 2/ C. Local Governments and Decentralized Agencies as Savers 238. Given the highly centralized structure of the Peruvian State, loca governments are relatively unimportant. Tax revenues of local governments account for only 1.2% of total General Government tax revenues. 3/ The only local governments of some importance are those of metropolitan Lima; they account for about 80% of the revenues and half the expenditures of all local governments in the country. Mainly because of the extremely weak tax admini tration, the revenues have been growing much slower than GDP and expenditure and the local governments have become more dependent on Central Government transfers. 4/ The local public welfare organizations 5/ have sizeable incom from lotteries with which they mainly operate hospitals, but their revenues do not cover current expenditures. Finally, a large number of cecentral-zed 1/ The payments to the SSA amount to around 2.3% of GDP (See Table 68). T' contributions to the Government's pension fund and the payroll tax toge account for around 1% of GDP (See Appendix Table 5.11). Another indica of the weight of these taxes on salaries and wages is that they account about 1/4 of General Government revenues. (See Appendix Table 5.21). 2/ On the tax system, see Chapter VII. 3/ See Appendix Table 5.21. Local governments cannot levy taxes; all taxation decisions are decreed by the Central Government. 4/ See Appendix Table 5.3 and 5.5. 5/ Beneficencias Publicas. - 242 - 0:b1a 70: ::Z;-A0RS ^F-_1A.ECS 'OF LnOAL Zi.VrTr;S =,A__ _A 7.:i S .r:nr . . 5, --9 c-5 pererentage Af 3:7\ 'vn :7_--emt e.-ernue3 - '_.5 1.6 1.3 1.0 . ' .o -. Z-uorot 0:.:en_:eo'es 2.9 2.5 2.1 1.9 2.0 1._ 2.3 C.- Z ita' e- e-' 0Z1'.". L C.'.L : Ooa:r2. ov-er=ent ::ase3 2/ 2.9 1.9 1.7 1.1 .2 1.1 :.2 '.1 2.9 .:ttn. ':_nesteb.le Sevirgs 1.9 :.. 1 0.2 0.2 0.2 ^. 0.1 ~~?-s:a Zxcc-.z_t-e3 '~~.0 0.1 A., 0.1 , 0.3 0.5 ,. ^.2 I v , a I a . ?.: t _ Ce 0.3 1. 0.. 6 -0.2 -'.2 - :,' _;oQo^ es Ceat. vTe er.ert ::-.sfers. Oi 0r:1o5' ;s 00e2n Cfa tn nsf:-rs. '.02r:20:Ane::Fx Lables 2.1, 5.1, 5.3, - 5.5. agencies in education and other public services depend almost exclusively on the Central Government for financing. As a whole, these three groups of public sector institutions incur a deficit of around 1% of GDP practically all covered by Central Government transfers (Table 70). d. The Central Government as Saver 239. Table 71 summarizes the main features of Central Government finances since 1965 and shows that they have broadly mirrored the trends for the whole public sector. Following the crisis of 1967-68 the Government took important action in 1969-70 to boost revenues and cut current expenditure growth. 1/ The improvement was dramatic: the share of revenues saved (current account balance) rose from 2% in 1968 to almost 16% in 1970 and the overall deficit was cut from 3% to 1.3% of GDP. 2/ Excluding the transfers to the rest of the 1/ Current revenues as a share of GDP rose from 15.3% in 1968 to 16.4% in 1970 (Table 71). The average yearly real growth rate of current expenditures fell from 5.6% in 1966-68 to 1.9% in 1969-70 (Table 72). 2/ The improvement was mainly caused by revenue increases. In fact, in spite of the sharp cuts in current expenditure growth, total expenditur.s (in- cluding debt service) grew in real terms by 5.9%O yearly in 1969-7t' compared to 4.4% in 1966-68 (Table 72). The growth was mainly caused by 1Y.0% growth of debt service payments (in spite of the formal debt resc!,edulings in 1968 and 1969 and also by the recovery of fixed investment frowi the depressed levels of 1968-69. Table 71: DIDICAT'ORS OF CEINTRAL mVIOKMENT TINANCES. 1965-76 -'ine rr. I o a r 1 y a v e r a S a a Th*rte-ysr average overeat 1965 -196d 1969 -1970 1971 1972 1973 1974 1975 1976 9W 0-b 973-19-7T--g A. OPI RATION4S AS PERCiFX OF GDP C -:roat Re-ecute 15.2 15.5 15.4 16.4 15.8 15.5 l$.o 15.5 15.4 13.6 15.8 15.3 14.6 15.0 (of wrich tax revqpueO) (13.2) (13.5) (13.6) (14.3) (13.4) (13.3) (13.1) (13.3) (13.8) (12.0) (13.8) (13.3) (12.9) (13.2) C.-rent Frnoaditur.e .~oxcl. tranaf. to pub. sect.) 12.5 13-9 11.9 12.7 13.2 13.5 13.5 13.0 14.1 13.7 12.5 13.4 13.6 13.5 C.n SaV:4ga 2.7 M .5 6 2.6 2.0 11. -o. 1.° 1.0 1.7 tarrert traesrferr to rest of public sector 1.7 `.1 16 1.0 0.9 0.9 1.1 1.0 1.5 1.0 T . C_ .-nt Account enc 1.0 03 1-° 2.6 1-7 1.1 0.4 - :~° -1 5 1.8 0,5 .o 4 0.4 C--ital xpetadiCree 4.5 3.4 2.9 3.9 4.7 4.8 4.3. 4.5 4.9 4.6 3.4 4.6 4.7 4.4 of jich fixed ilasetnent) (2.2) (1.4) (1.4) (2.0) (2.8) (2.9) (2.3) (2.9) (2.8) (2.5) (1.7) (2.6) (2.7) (2.5) kof bicb tr&nafera tc rost of public oector) (2.3) (1.4) (0.9) (1.3) (1.1) (1.1) (1.2) (1.0) (1.4) (0.8) (1.2) (1.2) (1.0) (;.1) C erall 3alance e35 =O -1.0 7-° -3- 7 -3 - -6.1 -1.6 -3-7 -5.1 __O B. H)Ql i JND IT1S ic:al transfer, to roeu of public sector % GDP 4.0 35 2.5 2.4 2.0 2.0 2.3 2.0 3.2 2.2 2.7 2.2 2.14 2.4 C-. -.. ave / A current revenues. 17.5 15.8 22.9 22.3 16.4 13.2 9.8 15.2 8.7 -o.6 27.8 12.8 6.6 11.1 C-.ront accoant bal. Y% current revenues 6.o 2.0 12.3 15.9 10.5 7.2 2.6 8.9 -3.0 11.1 19.0 6.4 -3.3 2.2 C 2:ent account balance 7. fixed irveatment 41.3 23.7 134.7 126.9 59.2 38.6 16.5 47.2 -16.5 -60.2 197.4 37.3 -17.7 13.4 .c m rerditures inl, aeb; awort. % GDP 20.0 20.3 18.3 20.1 21.8 22.3 22.0 20.8 23.0 21.7 19.6 22.0 21.9 21.5 2' Inc, ins current tranafer to the reot of the public sector So.r:c; AceaDtiX Tables 2.1, 5.1. 5.2, end Central bank t' - 244 - public sector (which were cut from 3.5% tc 2.4% of GDP), the Central Government achieved an overall surplus and was able to finance all of its fixed investment with its own savings. The interval of financial consolidation was followed however, by six years of steady deterioration. 1/ While the Central Government's own savings dropped until they disappeared in 1976, 2/ current transfers to the rest of the public sector once again expanded rapidly and the current account balance became sharply negative in 1976. 3/ As fixed investment expenditures recovered to the 1965-66 levels and capital transfers to the rest of the public sector remained high, the 1975-76 overall deficit reached the unprecedented level of over 5% of GDP. 240. The single major cause of the 1971-76 financial problems of the Central Government was the unsatisfactory performance of revenues. In fact, the Central Government's share of domestic resource use has increased somewhat during the decade but much less than what could have been expected from the aggressive interventionist role of the State since 1968. Total Central Govern- ment expenditures (including debt amortization) grew in real terms by about 6.5% yearly during 1971-76, not much faster than GDP, and by 1976 they were equivalent to 21.7% of GDP compared to 20% of GDP in 1965. Conversely, revenues grew only 2.1% annually in real terms during the last six years and as a share of GDP they dropped from 16.4% to 13.6%. 4/ Given the financial realities of sluggish revenue growth, expenditure growth was excessive. However, if the elasticity of revenues with respect to changes in GDP had been equal or some- what above one (as is the case in many other developing countries), the level of expenditures could have been financed without incurring an unmanageable overall deficit. 5/ 241. Restoring buoyancy to the revenues of the Central Government is there- fore an essential condition to attaining viable long-term financial balance. For this reason, Chapter VII is totally dedicated to a review of the main 1/ 1974 was an exception in the six-year period, mainly because of the extra- ordinary export performance that boosted tax revenues and temporary cuts in current expenditures. But even in 1974 the performance was worse than in 1969-70. 2/ Government's own savings are calculated before transfers to the rest of the public sector. Table 71 shows that the Central Government's own savings fell from 3.5% of GDP in 1969-70 to -0.1% of GDP in 1976. 31 The current account balance as a share of GDP fell from 2.6% of GDP in 1970 to -1.5X in 1976. 4/ See Table 71. 5/ Had current revenues averaged 16.4% of GDP during the entire 1968-76 period, the average overall deficit would have been 2.6% instead of 4% of GDP. - 245 - characteristics, problems and possible long-term inprovements of the tax system; only the principal conclusions of Chapter VII are given here. By international standards the tax burden in Peru is low and the buoyancy of tax revenues with respect to the changes in GDP has fallen from about 1.18 in 1950-65 to only 0.93 in 1968-76. _1 The built-in elasticity is much lower. The main causes for this low elasticity are the generous tax holidays and exemptions, which have increasingly eroded the tax base, and the weak tax administration. In addition, because of the poor performance of taxes on domestic transactions, the tax system has remained excessively dependent on foreign trade movements. 2/ Besides not generating sufficient revenues, the tax system does not contribute to the attainment of other Government objectives. Its impact on resource allocation is negative by subsidizing inefficiency, tilting investment sharply against labor intensive activities and failing largely in guiding resources towards sectors of high Government priority. Finally, the tax system is at best neutral in redistributing income within the economy. Chapter VII therefore concludes that a major reform of the tax system is required to increase elasticity and reduce distortions in resource allocation and income distribution it now encourages. Detailed suggestions for tax changes and improvements of tax administration in Chapter VII show there is room to increase tax revenues at least by 4-6 percentage points of GDP and that raising the Central Government tax burden from the 15% of GDP averaged in 1968-76 to 17-18% of GDP should not be an excessively difficult task. 3/ 242. Tables 72 and 73 show the main characteristics of the economic and functional classification of Central Government expenditures since 1965. The most important issue that these tables highlight is that the share of expend- itures directly related to development endeavors has been declining. By far the largest increases in expenditures have been in national defense and debt service. 4/ Their combined share in total expenditures almost doubledl from 22% 1/ See Chapter VII. 2/ Tax income from foreign trade transactions accounted for 65% of the increase of total tax revenues between 1973 and 1975, see Chapter VII. 3/ As a share of GDP revenues from income taxes could be raised by 2.4 percentage points (by reducing personal exemptions, tax holidays and deductions and toughening tax administration); taxes on property could be increased by 1 percentage point and excise taxes on gasoline and luxury consumer goods could be increased by 1-2 percentage points. In addition, more effective tax administration could increase yie:lds of most other taxes. For details see Chapter VII. 4/ Appendix Table 4.5 shows that less than 45% of the foreign loan commit- ments debt accumulated since 1970 have been related to development pro,ects. Table 71: INDICATORS OF CENTRAL GOV1RNHENT FIJNANCES 1965-76 r a * 1 r r a v * r a t * a e avSra#n averag 1965 1965 1969 1970 1971 1972 1973 1974 1975 1976 156-70 1971-7) 17"...77 A. OPERATIONS AS PERCENT OF ODP C_ttRevenues 15,2 15.3 15.4 i6.4 15.8 15.5 15.0 15.3 15.4 13.6 15.8 15.3 14.6 15.0 'of 'dich ttx revenue.) (13.2) (13.5) (13.6) (14.3) (13.4) (13.3) (13.1) (13.3) (13.8) (12.0, (13.8) (13.3) (12.9) (13.2) C-_ront rpeLditurow (.x l. tranef. to pub. sect.) 12.5 13.9 11.9 12.7 13.2 13-5 13.5 13.0 14.1 13.7 12.5 13.4 13.6 13.3 C;r "S":-ag 2.7 2-4 A : 2.6 2.0 i5 2.,3 1. ~-o.l I 1.0 1-7 .arreit trr. efers to root of public 6a4tor 1.7 2.1 1 1.0 0.9 0.9 1.1 1.0 1. Y T.5 1 C -7re Acco..Zt BOaI4c& 1.0 0-3 1 2.6 1.? 1.- 0-4 -3 -05 -1.5 1.8 . -04 0.4 C- ta_ £;xTnditure6 4.5 3.4 2.9 3.9 4.7 4.8 4.3 4.5 49 4.6 3- 4.6 4.7 4.4 ',f f fixee investment) (2.2) (1.4) (1.4) (2.0) (2.8) (i.9) (2.3) (2.9) (2.8) (2.5) (1.7) (2.6) (2.7) (2.5) ,ct .'cut tr,-gfor. to reat of public seotor) (2.3) (1.4) (0.9) (1.3) (1.1) (1.1) (3.2) (1.0) (1.4) (0.3) (1.2) (1.2) (1.0) (1.1) C-erkil dfauenc -5 -:! - 1.0 -1-3 = J 0 -3-9 -3-2 :5-4 -6- -1 .6 s z =4.c B. M£MORASIDJH ITW.S -si rarafero tu zst o~ pu~tli ..ector %i GDP 4.o 3.5 2.5 2.4 2.0 2.0 2.3 2.0 3.2 2.2 2.7 2.2 7_ ) L VA* ef;se ~ urront revenues. 17.5 15.8 22.9 22.3 16.4 13.2 9.8 15.2 8.7 -o.6 27.8 12.8 6.6 1. C;. -t accoant al. c current revonuee 6.o 2.0 12.3 15.9 10.5 7.2 2.6 8.9 -3.0 11.1 19.0 6.4 -).3 2.2 C xrert accwnt bal,nca A fixed inve6tment 41.3 23-7 134.7 . 126.9 59.2 38.6 16.5 47.2 -16.5 -60.2 197.4 37.3 -17.7 13.4 - u.experniturse icl. dabt arort. % GDP 20.0 20.3 18.3 20.1 21.3 22.3 22.0 20.8 23.0 21.7 19.6 22.0 n1.9 21.5 * Ledt cu-rret trt aferb to tta r*at of other public sector c lne . a6c xrrent tracoCtr to the rest of the public eector 5 :o: ApFeo.dix tables c.1, 5.1. 5.2, and Central Bank ON Table 72: SONOHIC CLASSIFICATION OF CENTRAL GOYiVEUIN1T XPENDIITURES INCLUDING DEBT SERVICIE, 1968-76 Percentage Structure Average armual growth ratee in real ter i51 968 1970 19' 199766-69-70 19wi-74 19V5-76 19Uig7W Wages aAd naleriea 35.4 30.8 28.5 25.8 24.0 -0.4 2.1 4.8 1. 3 2.2 Vation-ki defense 13.8 19.1 15.9 16.7 22.1 16.2 -3.4 8.9 20.7 1C.5 rsn ers27.3 27.1 21.4 1'7.8 21.5 4.2 5.9 2.7 15.8 L'+bt service '.4 11.4 i6.8 19.6 17.0 21.9 19.0 11.6 -2.4 12.9 Ti.xod investment 11.1 7.1 10.2 13.9 11.5 -9.9 26.6 16.2 0.9 6-.3 Cther expendituires 4.o 2.6 7.2 6.2 3.9 -9.8 75.4 3.4 -15.9 5.6 TC-A!,. DITURES 100.0 100.0 100.0 100.0 100.0 4.4 5.9 ? 5 .0 Debt a-cmrtizatiog 5.9 9.5 12.0 11.3 9.2 22.4 19.3 5.9 -S.9 10.3 Otter expenditures 94.1 90.5 88.0 88.7 90.8 3.0 4.4 ?-7 6.3 5.6 Current expenditures 71.4 - 68.4 67.0 6 1 A 6.9 Z4 . C 7 Intere6t paymenta 2.5 3.9 4.8 8.2 7.8 20.6 18.3 22.8 1.8 17.3 Iastional defenne 13.8 19.1 15.9 16.7 22.1 16.2 -3.4 8.9 20.7 10.5 Otner cmrrent 55.1 50.9 47.7 42.1 39.8 1.7 2.5 4.1 2.2 2.8 Caniti1 expendituree 22.7 16.6 19.6 21.7 21.1 = 14.9 10.2 2 / Total expenditurea (current expenditurea capital expenditures from Appoud'-- Table 5.2 + AmgrtizAtion of debt from Central Ba-ik). o/ Lv'Isted with GDP implicit deflator. S2rcen: Appendix Tafble' 5.2, and Central Bank. Table 713: mJCTIOAL CLARICATION OF CElNTLAI OVNERNMINT XPMDIT2RES IcLUDIoNG IEBT SERVICE, 1965-76 1/ TOTAL WMITUHR.s CURREN EXPIJDITU2ZS Percentage Struttre - F-ceut of GDP Average annual growth rate at conetant prices i/ 196?5 19t 197 19?3-74 1976 - 165 396B 1971 1974 974 70 19?11-74 1925-26 1966-74 TCriAL 100.0 =AO 100.0 100.0 9AO 20.01 22.1 2.1 224 100.0 100.0 100.0 100.0 -10o.0 1 w;lEVPMNiT .XPENDIT. 54 0 48.0 °6.7 46.6 10.8 2 4 10. 52.8 48.1 24 44.1 40.0 2.) 6-9 2.0 .ocia1 Sectors 36.4 32.8 26.2 25.3 n.-. 7.3 5.8 5.5 5.4 n.a. 46.6 41.4 35.2 34.7 n.a. 1.5 -5.9 6.5 n.A. 2.0 Education -(25.7) (23.3) (18.6) (19.3) (5.1) (4.1) (4.0) (4.1) (33.5) (30.6) (26.7) (26.9) (2.5) (-6.3) (7.9) (2.8) Healtb (.l1) (6.8) (5.5) (4.7) (1.4) (IaZ) (I.1) (1.0) (9.9) (9.0) (7.2) (6.6) (2.1) (-6.7) (3.6) (0.8) other (3.6) (2.7) (2.1) (1.3) (0.8) (0.5) (0.4) (0.3) (3.2) (1.8) (1.3) (1.2) (-13.5) (4.1) (-3.4) (-5.3) Economic Sectowe 17.6 15.2 18.8 16.9 r.-. 3.5 3.3 3.8 3.7 n.e. 6.2 6.7 8.3 7.1 n.-a 8.o 9.2 5.0 6.9 Jgriculture (5.2) (5.8) '5.8) (7.2) (1.0) (t.1) (1.2) (1.5) (1.9) (3.7) (4.0) (2.6) (30.9) (3.8) (-2.5) (9.1) Industry hnd Ca_. (0.7) (0.7) (0.8) (0.8) (0.1) (0.1) (0.3) (0.2) (0.2) (0.2) (0.5) (0.5) (10.1) (38.9) (17-8) (19.5) Fnergy and 11ines (2.9) (2.7) (2.5) (3.0) (o.6) (o.M) (0.6) (o.8) (o.5) (o.4) (0.6) (o.8) (1.1) (14.0) (19.4) (11.8) Tansaportation (8.8) (6.0) (9.1) (4.3) (1.8) (1.5) (1.5) (0.9) (3.?) (z.4) (2.8) (2.2) (-S.6) (9.4) (1.7) (-0.2) Fiaheries (-) (-) (0.5) (1.6) (-) (-) (0.2) (0.2) (-) () (o.8) (0.4) (-) (-) (20.3) () Naltisector progra - - 1.7 4.4 a.&. - 0.1 1.1 0.9 1.9 2.3 - - 24.3 MT SSRVICE 8.4 11.1 16.8 19.6 17- 1- so 4.5 43 -8 _ 8.4 12.5 11.1 20.6 18.3 22.8 1.8 OffER 2 6 4A I 2£ 8 _A -_ 2 4 6.1 7 o 46.6 46.2 48.9 8.2 3.9 1.4 Hatioal Defense 13.8 19.1 15.9 16.7 22.1 2.8 3.4 3.2 3.5 4.7 19.4 25.8 23.0 25.0 31.8 16-.2 -3.4 8.9 20.7 8.4 Other 23.8 21.S 20. 1 1 7-1 L.0 A .9 2. ;4.3 _ _0.8 2 1.a .0. rI.i u.8 12.2 -1.3 2.8 2.3 T Tot*l expenditures tro Tale 16. 2/ Deflated with GM implicit AefIter. Source Appeadix Tabl*s 2.1, 7.2, .15 and C*atral hak. I Cx J=. - 249 - in 1965 to almost 40% in 1976. The increase in national defense expenditures was particularly rapid in 1975-76 (averaging almost 21% per year in real terms) and they now absorb 4.7% of GDP compared to about 3.5% in the early 1970s and. 2.5% in 1965. The 13% average annual real growth of debt service payments since 1965 is the result of the succession of many years with large overall deficits. Since the bulk of borrowing occurred in the 1973-76 period debt service payments will continue to grow at an even faster rate as the grace periods start to ex- pire in 1977. 243. Excluding national defense and debt service, Central Government ex- penditures grew at an annual rate of only 3.6% in real terms in the 1966-76 period. 1/ During 1968-74 the growth of transfers to the rest of the public sector and to the private sector was substantially reduced, but went up sharply in 1975-76 mainly because of the large deficits accumulated by public enterprises. 2/ However, only a small share of the transfers support develop- ment-oriented expenditures; most transfers to the private sector are pensions and most transfers to the public sector subsidize consumption. Fixed invest- ment grew faster than total expenditures in the 1968-76 period, but as a share of GDP in 1975-76 it was still below what it had been ten years earlier. Moreover, as discussed earlier, the returns from public investment will probably be low in terms of contributing to the development objectives of the Government, at least for most of the larger projects. 244. The functional classification in Table 73 shows a decline since 1965 in the importance of social services and transport in Central Government expenditures. 3/ Expenditures on these sectors calculated as a share of GDP dropped from over 9.1% in 1965 to 6.1% in 1974. 4/ While population wcls growing at almost 3% yearly, expenditures on social services grew only 2% yearly. 5/ After contracting sharply in 1969-70 (in real terms) expenditures on education have been growing at rapid rate during the 1970s, but as a share of GDP they remain below the level attained in 1965. 6/ Other basic sccial 1/ See Table 72. 2/ See Table 73. 3/ The share of total expenditures allocated to social services and trans- port fell from 45% in 1965 to 30% in 1974 while the share of current expenditures in these sectors fell from 50% to 37% (Table 73). 4/ As a share of GDP, between 1965 and 1974 the expenditure on social ser- vices fell from 7.3% to 5.4% and those on transport from 1.8% to 0.9% (see Table 73). 5/ Therefore, expenditures on social sectors have substantially shrunk since 1965 (in real terms). 6/ Current expenditures in education fell at an annual rate of over 65% in 1969-70 and grew at a yearly rate of almost 8% in 1971-74. As a share of GDP, total educational expenditures fell from 5.1% in 1965 to 4.1% in 1973-74 (see Table 73). - 250 - services (principally housing and health) hava received even lower priority, judging by the expenditure patterns. Outlays on these services were practi- cally at the same level in real terms in 1974 as they had been in 1965, and their share in GDP dropped over 40% from 2.2% to 1.3%. As discussed above, developments in transport were constrained by the low amount of funus allocated and by the limited administrative capacity. 245. The expenditure increases in agriculture, industry and commerce, fisheries, and energy and mines were substantial but their impact on accelerat- ing economic growth or on other development objectives of the Government have probably not been high. As a share of GDP, Central Government expenditures in these sectors rose from 1.7% in 1965 to 2.8% in 1973-74. 1/ About half of these expenditures were for fixed investment (mainly for large-scale irrigation projects); an additional quarter were capital transfers supporting investments in the rest of the public sector. 2/ Current expenditures account for the balance of outlays in these sectors and financed mainly the creation and rapid expansion of the Ministries of Industry, Commerce, Fisheries, and Energy and Mines. 3/ However, most of the money spent in these new ministries went to administer controls imposed on private and public sector activity, while little went for the expansion of services supporting and promoting development. Some examples may illustrate this. In the agricultural sector, the bulk of current expenditures and staff went to support the Agrarian Reform process. The latter deserved important attention, but the assistance to farmers through extension services was negligible. 4/ In incustry and commerce, current expenditures went to administer the intricate system of controls of industrial investment and of external and domestic trade, but little was spent supporting small-scale industry and promoting Peruvian exports. 5/ 246. The performance of Central Government operations and finances also suffered from management problems that have kept the budget from remaining within the targeted ceilings and to maintain the pattern of expenditures desired at the beginning of the budget period. One of the problems was the practice started in 1969 of having two-year fiscal periods. The theoretical objective of the two-year budget was good: it would permit a longer time framework in which to plan and allocate expenditures. In practice it did not work. The principal problem was that the two-year budget did not allow for a sufficiently close control of expenditures: by the end of the first year a large share of the two-year allocations had already been spent making it 1/ See Table 73. 2/ As mentioned before, the bulk of these Lnvestments will do little to foster rapid and efficient growth of the economy. 3/ The Ministries of Industry, Commerce and Fisheries were established after 1970. Table 73 shows that the current expenditures in industry and commerce, in energy and mines, and in fisheries, grew by over 17% annual- ly (in real terms) during 1971-74. 4/ For details see Chapter VIII. 5/ On industry see Chapter X. - 251 - necessary to authorize larger spending levels for the second year. Spending units became accustomed to the somewhat theoretical nature of budget aLlo- cations, since they could always request supplementary authorizations, particularly in "strategic sectors". As supplementary authorizations were ad-hoc decisions made throughout the budget period, their accumulation tended to change the pattern of expenditures sharply. Moreover, since there was no effective machi.nery for establishing "hard" priorities, faced with the need to match projected current revenues closely with expenditures at the time of budget law approvals, the Ministry of Economy and Finance often made unrealis- ric expenditure cuts which would be untenable during the budget period for political or economic reasons. 1/ A third problem is that the discretionary margin of the authorities in managing the expenditure levels has been falling: the rapidly growing debt service burden has become incompressible in the short run; the concentration of public investment in a few large projects with multiple-year execution periods and large cost overruns makes it difficult to trim investment expenditures; the losses of public enterprises are control- able by the Government, but in the absence of measures the budget takes them as given; and while over the period as a whole Central Government wages have not grown rapidly in real terms, cutting wage costs would demand a reduction in the number of employees which is politically difficult to achieve. In view of these problems, the Government decided to return to the traditional one year budget for 1977. Less progress has been made in establishing an orderly and efficient system for making priorities and reducing rigidities in the expen- diture pattern. By mid-year the overall size and composition of the 1977 budget was still unknown. To balance the budget, the Government was forced to decree across-the-board cuts in expenditures which are effective but highly inefficient since they cut expenditures in high priority and low priority pro- grams alike. 247. In sum, there clearly is room to rationalize the budget process and cut expenditures by r.educing non-developmental outlays, by shaving the size of the bureaucracy (through streamlining or elimination of ineffective and unnecessary controls), and by increeasing the efficiency of resource use in investment. Nevertheless, the urgency of improving and better maintain:.ng the transpz:tation network and increasing the provision of education, health and other social services --if the objective of reducing poverty is to be effect- ively pursued-- will exert expansionary pressures on Central Government expenditures. Allocating the same share of domestic resources to these serv- ices as in 1965 would increase expenditures by almost 3 percentage points of GDP. In addition, as mentioned earlier, the burden of debt service will grow rapidly in coming years. Therefore, although much more benefits could be derived from each sol spent by the Government, it is doubtful that the aggre- gate level of expenditures as a share of GDP can be cut while at the same time responding to pressing demands for public services. This emphasizes the cru- cial importance of revenue action as an essential condition to improve the savings performance of the Central Government and a prerequisite of success- ful poverty programs. 1/ Examples of items often excluded from the original two-year budget law were salary increases, financial subsidy implications of price controls on food and petroleum, project cost overruns, and local counterparts to foreign financed projects. - 252 - D. MAhNAGEMENT CAPACITY OF THE PUBLIC SECTOR 248. The problems faced in generating savings and achieving an adequate efficiency of resource use are only two dimensions of the broader issue of the capacity of the State to plan and direct the economy and manage the public sector. When the State started to increase :.ts role in the management of the econoay in the early 1960s, it was constrained by the lack of qualified and experienced technicians and administrators. There was no tradition in Peru of a prestigious civil service; the heavy bureaucratization and inefficiency, and the few incentives involved in a public service career meant that most of the scarce trained personnel preferred to join the private sector. The problems, of course, have become worse since 1968 because of the zeal of the State to attaiuing direct administrative cont:rol of so many functions in the economy. I/ The determination of the State to change development patterns and the desire to get things done d:.d attract many young, well- educated, and highly motivated recruits into the public sector. Many of them had a strong theoretical background, but little practical experience. They had great confidence in the State's capacity to do anything if the proper organization were set-up, and were responsibLe for the design of sophisticated but extremely complex organizational systems for public management. Finally, they were generally skeptical of economic efiiciency considerations, placing priority on political objectives. 249. Various parts of this report analyze specific problems and shortcomings that have appeared in the management of the economy by the public sector. This section addresses these problems in a broader context of public management: (i) aggregate planning, economic management, and management of the public sector; and (ii) efficiency and State contro;. on economic activity. a. Aggregate Development Planning 250. Efforts at economic planning by the State began in 1961-68. The National Planning Institute (INP) was created and two national development plans were prepared. 2/ However, little practical use was made of the plans, mainly because of the still limited role of t:he State in the economy and of the limited political support for planning efforts. 3/ This changed after 1/ Amongst others, the State aimed at: devreloping an effective macro- planning system; creating and managing a large number of state enterprises; developing and managing a large investment program in practically all sectors; and setting-up increasingly det:ailed controls on general economic activity. 2/ The first plan was prepared by the Central Bank in 1961 (Plan Nacional de Desarrollo Economico y Social del Peru 1962-71). INP was founded in 1962. The second plan was prepared by DNP in 1966 (Plan de Desarrollo Economico y Social 1967-70). 3/ Another reason was that the two first p.ans were completed shortly before new governments introduced significant changes to the develop- ment strategy. - 253 - 1968 and central planning became a major objective of development policy. I/ In the first days of the new regime, the top military leaders prepared the Plan Inca, which laid down their new political objectives. Although its contents were not publicly revealed until 1974, they guided Government action until 1975 during the so-called "first phase of the revolution." 2/ In 1976-77 a new document was prepared -- the Plan Tupac Amaru -- laying down the broad objectives for the "second phase of the revolution", that is to last until the early 1980s. 251. To make the planning objectives of the Government operational, a complex machinery has been established. At the top are the Presidential Advisory Council (COAP) and the National Council for Economic and Social Development (CNDES). 3/ They translate the general objectives established by the broad political plans. (Plan Inca or Plan Tupac Amaru) into specific policies. Under them, INP prepares and monitors the execution of plans, manages and coordinates the public investment program, coordinates the activities of government agencies at a regional level (through its four region- al offices 5/); and participates in the preparation of public sector budgets. 4/ 1/ See Chapter IV. 2/ The most important aspect of the Plan Inca is that it laid down the series of structural reforms to be pursued by the Government. However, these general plans did not specify when and how these reforms were to be achieved nor any quantitative targets. 3/ COAP is made up of the top military chiefs and the ministers. CNDES comprises the President of the Republic, the ministers, the head of INP, the presidents of the Central Bank and the National 3ank and the * head of COAP. 4/ The four regional offices are ORDECENTRO at Huancayo, ORDESUR at Arequipa, ORDENORTE at Chiclayo and ORDEORIENTE at Iquitos. These deal with t'le Central, Southern, Northern and Amazon regions respectively. 5/ A more complete list of INP's functions (according to the law that governs it) includes to: (i) prepare and present to CNDES periodically short-, medium- and long-term plans and monitor their implementation; (ii) par- ticipate in the formulation and evaluation of the public sector budgets and ensure that they agree with the targets and objectives of the develop- ment plan; (iii) present to the cabinet periodical reports about the country's economic and social situation, and propose fiscal and other measures to assist the fulfillment of the targets and objectives of the development plan; (iv) coordinate social policies for structural charge; (v) oversee the selection, preparation, financing and evaluation of public sector investment projects; (vi) advise on international aspects of development policy, including technical cooperation; (vii) participate in the formulation of policies for science and technology; (viii) orient and coordinate the operations of the SPOs; (ix) collaborate with the National Council for Integration (CONIT) in defining the policy of regional and sub-regional integration; (x) promote studies relating to development planning. - 254 - INP also orients and coordinates the work o:. the Sectoral Planning Offices (SPOs) within each Ministry. These are responsible for formulating sector strategies, policies and investment programs and for carrying them out after inclusion in the national plans and budgets. 1/ The Ministry of Economy and Finance and institutions under its control (particularly the Central Bank) participate in the planning process, preparing the public sector and foreign exchange budgets and the monetary programs. The planning system also includes a National Institute of Statistics (INE) created in early 1976 to centralize data required by the Government. Finally, :he work of these basic organi- zations is coordinated and supplemented by a growing number of statutory and ad-hoc committees. 252. INP prepares two main type of planis. First, there are the medium- term Development Plans of which two have been prepared for 1971-75 and for 1975-78 2/. These plans are framed within the general long-term strategic objectives defined by COAP, and formulate detailed objectives and targets sector by sector. Second, there are the short-term Biennial Plans which up to 1976 were supposed to precede the preparation of public sector budgets. These adjust and refine the targets of the medium-term Development Plans. To monitor the implementation of the Plans, IN"? prepares quarterly and annual reports, which are reviewed by the Cabinet. 253. The planning system that emerged hlas substantially heightened the influence and power of INP and of the SPO iL each Ministry. Both have been responsible for much of the voluminous legislation enacted since 1968, and designed and executed many of the structural reforms of the last eight years. Today the sophistication, experience and know-how of Peruvian planners is far better than what it was before 1968 and provably compares favorably with planning systems in other Latin American coulntries. 254. Nevertheless, the planning process also has various serious short- comings with important negative consequences. In the first place, the planning process gave little attention to the actual resource balance of the country: it was unrealistically optimistic on the availability of some resources and gave little emphasis to other resources that remained underutilized. Excessive optimism was shown on the availability of financial resources to finance the plans and or human resources and experience to manage them. In spite of rapidly widening savings and foreign exchange gaps after 1971, the planning system was not successful in adjusting targets and spending plans to the 1/ All policy and investment proposals from Central and Local Government organizations and public enterprises go through the relevant SPO before being considered for implementation or financing. Government relations with private enterprises operating or wishing to operate within a par- ticular sector, are the concern of tha: sector's SPO. Applications from private entrepreneurs for investment or government licenses must go through the SPO. Each SPO has to repo-t periodically to INP on the progress of sector plan implementation. 2/ For an overview of the objectives pursued in the 1971-75 and 1975-78 Development Plans see Chapter IV. - 255 - financial realities and in avoiding the boom-bust cycles. 1/ On the contrary, it probably widened the gap and resulted in the serious financial crisis that the country has been going through since 1975. 2/ Part of the problem is Institutional. In spite of intricate coordination mechanisms established, the preparation and management of plans by INP remained divorced from financial realities of the country which were the responsibility of the Ministry o' Economy and Finance and of the Central Bank. The emphasis of INP was on how resources should be spent, not on how they should be mobilized to finance the plans. 3/ On the other hand, until 1976 INP was the only agency that did financial and economic proiections. 4/ However, these projections tended to 1/ Since, as reviewed earlier, the share of investment and of resource use managed by the public sector rose substantially since 1968, and since the private sector was subjected to numerous controls, it might have been expected that the State should be able to fine-tune developments in the economy more rapidly and more directly than before 1968. 2/ See Chapters IV and V for an overview of macroeconomic trends since 1968. 3/ The problem was more of policy and of the mechanics of coordination than of the basic organizational framework. The plans prepared by INP men- tioned the savings problem and a large number of committees exist in which the financial problem, as viewed by the Ministry of Economy and Finance and the Central Bank, should have been transmitted to INP. En reality this did not occur and in the actual execution of the plan little attention was paid to the widening savings gap. 4/ The economic projections by INP have principally revolved around two large econometric models (known as INP I and INP II). These were used in the preparation of the two medium-term plans. In practice, these models have been mainly interesting academic exercises but with little practical use. In the first place, they depend on a very large amount of data whose generation requires the work of a large number of people --making diffi- cult the permanent updating of its projections-- and whose quality was anyway unreliable. Thus, they did not serve the function of providing decision-makers with a simple consistency framework with which to test the impact of current macrofinancial policies and decisions. In the second place, for all their size and comprehensiveness, these models were not useful to estimate the implications on the economy of major structural changes introduced by the Government (such as the impact of the agrarian reform on the incomes of the poorest half of the population). In the third place, these models were not useful to test out the long- term consequences of alternative policies on issues such as population, energy, etc. - 256 - be highly optimistic about future resources available 1/; little sensitivity analysis was made of the possible impact of unforeseen exogenous factors; and they were not regularly updated to see how the changing present conditions were affecting the panorama in the future. Therefore, the proJections were not a useful tool in guiding the economic decision-making process. Only since 1976 has the Central Bank started to produce, on an adhoc basis, medium-term macroeconomic projections used as an input in decision making. At the same time, the human resources and experience in management within the public sector were excessively strained by over-ambitious objectives and the complex organizational structure. Typically within the INP and in many ministries and public sector institutions there is a small group of qualified and motivated tecnicians and managers, but they are overburdened and their effectiveness is diluted by having to participate in a large number of boards and committees, and by the excessive number of routine procedures. 255. Little practical emphasis was given by the plan on reducing the underutilization of labor, probably the resource in largest supply in Peru. The plans mentioned labor absorption as a general objective, but the strategy and policies the plans laid down gave it little priority. 256. A closely related issue to the resource balance question is the little amount of long-term planning that has been done. A fairly clear vision of the organization of the economy and society to be attained in the long run exists; much less is known however on the constraints that the resource base and the weight of present decisions will create in attaining this vision. 257. The planning process also gave a low priority to maximizing the efficiency of resource use in attaining objectives. In part this stemmed from skepticism toward traditional efficiency criteria as a way of assessing the desirability of specific policies or programns. However, as is argued in detail in other chapters, whatever the objectives, many of the policies and programs pursued would have been different if the eccnomic costs of alternative means of attaining them had been calculated. It is no exaggeration to say that unless the problem of efficiency in resource use is given more priority, it will be virtually impossible for Peru to maintain a rapid pace of development. 258. Another problem of the planning process is that it is "planning from above" with little consultation being sought from outside the public sector. Even with the best intentions the process can become divorced from the percep- tion of development goals as seen by the ultimate beneficiaries at the base. This in turn, can make it difficult for the latter to comprehend the develop- ment objectives and participate in their swift implementation. 1/ The initial positive results of the petroleum explorations and highly optimistic assumtions on the timing anc. foreign exchange generation capacity of the public investment program were important causes of these optimistic projections. It was widely believed that even if no measures to improve the savings performance were taken, the export boom after 1977 would be so dramatic that it: would allow financing both the servicing of the external debt that was being incurred and rapid economic growth. - 257 - 259. Finally, the macroeconomic planning and decision-making process has been hindered by lack of reliable data. Earlier pages of this Chapter have briefly reviewed the problems in public sector data, other Chapters have highlighted the problems in other sectors such as national accounts, agriculture, industry and the balance of payments. INE was created in 1976 to tackle this problem. It is too early to evaluate INE's operations, but if it is provided with sufficient resources and is able to gradually upgrade the quality of data it may prove to be one of the most significant contributions to improving the planning and decision-making process. Particular attention should be given to avoid the experience in other countries where autonomous central statistical offices have often devoted too much of their energy to what may be intellectually interesting as a statistical exercise but may be far less rewarding to the country, than technically less sophisticated exercises required to enhance the quality of decision making. Thus, while the independence and integrity of the institution must be maintained to ensure an accurate portrayal of reality, the work programs of INE should be closely supervised by the decision-making institutions to ensure that the highest. priority data are being produced. 260. The problems of management of the internal operations of the public sector are closely related to those of aggregate economic planning since it is through the public sector that a large share of the plan's objectives are implemented. The little attention paid to the balance of resources and to the efficiency of resource use in the planning process has been an important cause of the shortcomings in managing the public investment program, the operations d of the public enterprises, and the Central Government budget. b. Controls of Economic Activity 261. One important characteristic of the planning and management system set up in Peru since 1968 is the widespread confidence in the ability of ad- ministrative controls to influence the resource allocation process. Starting from the assumption that the free operation of market forces was a principal cause for the unbalanced development and skewed wealth and income distribution in Peru, it was believed that the State should not only acquire a large share of the strategic resources in the economy directly but should monitor closely and restrict activities outside the public sector. Controls have been imposed on practically all aspects of economic activity. Other chapters of this report discuss the specifics and the consequences of many of these controls. Not: only has the number of controls increased rapidly but the procedures involved in many of the controls have become more complex. The cost of the controls to the economy have been very large both directly, because of the large staff required to administer them, and indirectly, because of the loss of fluidity in the decision-making processes. 262. Have the benefits achieved by the controls outweighed the costs? The experience up to now has been very negative, since the controls have done little to foster the attainment of Government development. Some examples illustrate the problem: (a) The complex procedures in obtaining Government clearance for establishing and operating firms have not changed the structure of industry in favor of intermediate and capital goods, but may - 258 - have encouraged more concentration of industry in a few large firms and in the Lima-Callao area. (b) The State's controls of agricultural marketing have not increased efficiency. (c) Foreign exchange controls have failed to stop capital flight, and the increasingly restrictive import-control mechanisms have not been able to keep merchandise imports within planned limits. A complex machinery to control foreign borrowing has not been successful in curbing the expansion of external debt, although it has contributed to costly delays in loan approvals. (d) The widespread price controls have not been conducive to an efficient allocation of resources nor have t:hey contributed to reduce inflation in the longer run. The exchange rate was not moved and became substantially overvalued in the 1968-75 period; this penalized exports, subsidized the upper income groups who spend more on imports than others, and was part:ly responsible for the financial crisis the country is now going through. The policy of negative real interest rates, repressed the development of the financial system and distorted resource allocation by subsidizing investment and encouraging capital intensive activities. The long lags between increasing costs and adjustments of public utility tariffs held back the expansion of basic public. services and infrastructure. Price controls on petroleum products and foodstuffs were not consistent £ with a viable long-term energy poLicy, discouraged agricultural production, and the financing of :he resulting losses fueled inflation. Finally, the broader price controls on many services and industrial goods discouraged production. 263. Overall, the controls used have not favored a more equitable distri- bution of wealth and income nor a reduction of poverty. In fact, they may be having the opposite effect. Controls on industrial investment and location, credit and interest rates, and prices are biased in favor of large established enterprises, making it extremely difficult Eor a new small enterprise to get established and succeed. Restrictive labor legislation protects excessively those already employed but, combined with t'ie main instruments of industrial policies, discourages labor-intensive industries. The slow development of public services caused by inadequate tariff policies penalized the poor. For example, the urban poor not covered by the formal water system end up paying far more for water transporation by truck than higher income persons in the system. 264. The Government should thoroughly review the present controls to assess their efficiency in achieving objectives. Controls have become so complex, interlocked and widespread that a partial analysis of specific groups of controls may not be sufficient. The tying up of scarce management and administrative resources in complex and ineffective Government controls has become a major constraint to rapid development and is a major cause for the problems encountered in aggregate planning and public sector management discussed above. It is probably true that in the present legal framework and with the sharp disparities in opportunities that prevail in the economy, che - 259 - forces of the free market will not tend toward an equitable distribution of income nor a rational allocation of resources. However, the present patterns of Government interventions instead of equalizing opportunities and allocating resources to the attainment of the desired objectives at the lowest costs are having the opposite effect. The Government's desire for a heavy State inter- vention could be attained much more effectively if most of the present bureau- cratic controls were eliminated and replaced by simpler and more sharply focussed mechanisms. Other chapters discuss specific recommendations in the various sectors. E. CONCLUSIONS AND FUTURE PRIORITIES 265. This Chapter has discussed the main characteristics of the dramatic expansion since 1968 of the role of the State in the economy. The State has become the largest entrepreneur controlling a large share of the most pro- ductive assets, accounting for around half of all fixed capital formation and of all exports and it has set up mechanisms to closely monitor and steer activities outside the public sector. The main problems that have constrained the State's efficient attainment of its objective and the rapid development of the economy were found to be: (i) deficiencies in the management capacity of both the general developments in the economy and the functioning of the internal operations of the public sector; (ii) insufficient generation of savings to finance the expanding activities of the public sector; (iii) low efficiency in the utilization of resources by the public sector; and (iv) an excessive number of inef'ective and bureaucratic government controls. The financial crisis the country has been going through since 1975 has been partly caused by these problems but is only a symptom of the broader problems of public sector management. The next few years will require tough austerity in public sector finances, but unless the more fundamental problems are faced the improvements will not be long lasting. It should however be emphasized that the analysis in this chapter does not assess what the relative size of the state should be in Peru's economy; instead it concludes that the state should avoid controlling for its own sake and should allow state enterprises to be run effectively. - 260 - CHAPTER VII: THE TAX SYSTEM 266. The previous Chapter discussed how the inadequate growth of Central Government revenues has been an important cause for the low savings perfor- mance of the public sector. Unless more permianent increase in the level of public savings is attained, efforts to increase the rate of investment in the economy will rapidly lead to financial problems. This chapter takes a more in-depth look at the tax system and identifies the main weaknesses and ways of improving them. Two main objectives are kept. in mind in the analysis of the tax system: how to increase its elasticity *rith respect to incomes, and how to improve its impact on resource allocation in the economy. A. ON OVERVIEW OF ISSUES IN THE TAX SYSTEM 267. From a revenue-generating point of view, the most serious problems of Peru's tax system are its low elasticity to changes in income, its heavy dependence on foreign trade, its complex structure and the weak tax adminis- tration. 1/ 268. Indicators of the low buoyancy of the tax system and its deterioration since 1970 are shown on Table 74. The marginal tax rate in- creased from 9 to over 15% between 1951-55 ard 1966-70 but fell to 11% in 1971-76. The buoyancy of the tax system, estimated at a respectable 1.18 for 1950-65 fell to 0.94 in the 1968-76 period. 2/ Moreover, the built-in elasticity of the tax system was found to be well below I in various studies. 3/ Thus, buoyancy has been achieved, only because of continuous 1/ The 1971-75 Plan stated that the Government would ".... undertake an integral reform of the tax system aimed not only at providing the Government with the necessary financial resources to expedite development but also aimed at supporting the measures to redistribute income, direct economic activity, develop an efficient system of tax administration, and at the same time change the tax consciousness of the country". (See Plan Global 1971-75, op. cit., p. 43, mission translation.) 2/ "Buoyancy" measures the elasticity of actual tax collections with respect to GDP. "Built-in elasticity" measures the elasticity of tax revenues with respect to GDP without discretionary changes in the tax system. 3/ A detailed IMF study measured the built-in elasticity of the tax system at 0.72 in the 1960-71 period, 46% below the buoyancy estimate of 1.12. This large difference between the buoyancy and elasticity estimates shows how slowly revenues would have grown in the absence of discretionary action (See Chand, S. and Wolfe, B. "The Elasticity and Buoyancy of the Tax System of Peru 1960-71: An Empirical Analysis", DM/73/55, ITMF, July 1973, pp. 11-14). Information was not available to estimate the built-in elasticity for the more recent 1968-76 period. However, since the buoyancy fell from 1.12 to 0.94 (measured using essentially the same methodology) in spite of important changes in the tax structure (parti- cularly income taxes in 1968, sales taxes in 1973 and export taxes in 1976) the built-in elasticity is also likely to have fallen below the 0.72 estimated for 1960-71. _ 261 - Table 74: P?u- mUicATRs OF LOw-CTIM CnANGES iN TAX 5UBflG AMD SlOYANCY, 195o-76 Tax R8evenues Charge in Tax Revenues as o as Z7 of Buoyarey GDP Change in COP Estitate- Tax Effort Indicators 1951-55 9.8 9.2 1956-60 10.1 14.4 1960-65 12.8 14.8 1966-70 13.8 15.2 1971-76 13.0 11.0 Buoyancy Estimate&s' 1950-65 b/ 1.18 1960-711/ 1.12 1965-s71-' 1.04 1968-76!' 0.94 *1 Estimate of rate of change of tax revenues to rate of change in income. Normally cal- culated with double logarithmic function relating tax revenues to income. a, Schydlovsky. I., coumenta on the paper by S. Hunt, Distribution, Growth and Etono='c Behaviour in Peru, ia C. Ranis (ed). Government and Economic Development, Yale Unu- versity Press, Nev Haven 1971, pp. 416-28. el Cband, D. and Wolfe, B., "The Elasticity and Buoyancy of the Tax System of Feru !V>O- 71: An Empirical Analysis," DfF, DM173/55, Wlashington, Df.C., July 1973, p. 11 1I ClAP, Subcorite sobre Peru, Estudio de Is Secretaris de la OFA, Washington, D.C., 1972, p. V-6. Sources: See footnotes and Appendix Tables 5.11 and 5.20. Table 75: PMiD: ESTMAUT OF BUOYANCY OF CTBRAL COVl..'EsT TAX REVENUES, 1968-76 (in percent) Percentage £ Structure Suoyancvl 1968 1976 1963-76 Total Tax Revenues 100.0 100.0 0.939 Taxes oan ncoma 26.6 22.7 0.876 Taxes oan Property 6.8 6.3 0.823 TaxeS on Goods and Services 37,5 48.2 ;.098 (Gemeral Sales Tax) (24.9) (32.9) (1.187) Taxes on Foreign Trade 29.5 24.9 0.325 (Taxes on Tzports) (23.6) (17.9) (0.654) i Estimated with simple double logarithmic functions relatirg actual tevenues to GDP. Sources: Tables 2.1, 5.11 and Misaion estimates. - 262 - changes in tax laws. Compared co other developing countries, Peru's tax effort is low, despite the Government's inc::eased participation in the economy. 1/ 269. Two main factors account for the Low buoyancy of the tax system: the erosion of the tax base through generous tax exemptions and the ineffective tax administration. In 1968, a general personal and corporate income tax was introduced replacing an archaic schedular income tax which had taxed only certain types of income. With a maximum tax rate of 65% on personal incomes, markedly increased both progressivity and the elasticity of the tax system. However, these beneficial features were neu:ralized by extremely high per- sonal exemptions and generous tax holidays ':o corporations. In addition, tax evasion appears to be large because of inadequate tax administration. The overall result is that in the 1968-76 period, the buoyancy of income taxes was only 0.88 (table 75). Similarly the buoyancy of import duties with respect to GDP was only 0.65 despite an increase in the average tariff level as well as faster growth of imports than of GDP. The :elated causes were a shift of the structure of imports away from final consumer goods and massive exonerations of import duty payments to promote industrial development. The only dynamic element in the tax system over the last decade has been indirect taxes on production and consumption. Benefitting mainly from the replacement of stamp taxes by a general sales tax, these indirec: taxes grew faster than GDP and accounted for well over 60% of the increase in tax revenues in the 1968-76 period (Table 76). The introduction of the general sales tax has probably been the most important improvement in the :ax system in recent years, both because of its large revenue-generating potential and its less adverse effects on resource allocation, and also because it is easier to administer than the cumulative and complex stamp taxes it replaced. 270. The second feature of the tax system that stands out from a revenue- generating point of view is its strong dependence on foreign trade. Although the importance of the taxes directly levied on imports and exports appears at first view to have decreased somewhat (Table 76), Table 78 shows that the dependence of the tax system on foreign trade is large and may be growing. In 1975, an estimated 2/3 of the corporate income tax revenues came from exporting companies (mainly in the mining and sugar industries) and 1/3 of the general sales tax revenues came from foreign trade operations. In total, therefore, 1/ Another IMF study of trends in taxation in developing countries found that Peru's tax effort for 1966-69 was below average. (This study took into account three factors in comparing tax efforts: per capita income, the relative importance of international trade in the country and the relative importance of the mining industry in GDP). (See Chelliah, R., "Trends in Taxation in Developing Countries", Staff Papers, Vol. XVIII, IME, 1971). Since 1969, taking into account contributions to the Social Security System and taxes collected by local gover'rents, total tax revenues of the General Government peaked at 16.9% of GDP in 1974 but fell to 14.5% by 1976 (See Appendix Table 2.21). Table 76: PERU: ChANES IN STRUClURE OF t:FNTIjtA. COVFUItWFNT TAX RFVENUES. 1968-76 (In Peceent) Percentage Structure of., Total Change In Revenues 1969 1970 1973 1975 1976 1969-70 1971-73 1974-76 Total Tax Ikavenue _ , 9- 100 _"J .1i05 .1(pO IOQ -M . Taxes on lcome 27 36 34 32 23 62 2H 12 Income Tax 21 26 21 n.1 . 0 2 5 n.m. Other 6 10 a 11 n.m. 22 3 D.&. Taxes on Property 1 S 5 5 6 1 4 8 Taxes on Goods ad Services 38 35 43 41 48 28 64 53 General Sales Tax 25 22 30 30 33 13 S2 36 Ex1cl Texee 9 9 10 a 12 9 13 15 Other .4 4 3 3 3 6 1 2 Taxes On l*0reign Trade 30 24 20 25 25 10 8 30 Taxe on Imports 29 24 20 23 18 9 ---8 16 Texes on Exports I - - 2 7 1 - 14 Source: Appendix Tableb 5.11 *ad 5.12. LA) Table 77: PEHIO: ShARE OF GDP AND REAL GROW0I RATES OF CENTRAL GOVERHMENT TAX REVENUES. 1968-76 (In Percent) Ig68 1969 1970 19JI 1972 1973 1971. 1975 1976 1969 Ig70 1971 1972 1973 1974 1975 1976 Percent of GOP a/ Percent Growth In Real Terms c/ Total Tax Revenues 1 3 6 1i 13.4 5[d 13. 8 12.0 1 .1 . -1.2 4 631 10.8 6 0 -1to5 Taxes on Income 3.6 3.8 5.1 4,0 4.5 4.4 4.9 4.4 2.7 8.2 47.9 -17.5 17.8 5.8 20.9 -9.2 -35.5 Income Tax (2.9) (3.2) (3.8) (2.1) (2.6) (3.4) (3.8) (2.9) (n.a.) (15.6) (26.4) (-24.7) (3.6) (38.7) (21.9) (-23.2) (n.a.) Other (0.7) (0.6) (1.3) -(1.3) (1.9) (1.0) (1.1) (-5) (n.a.) (-22.2) (178.9) (2.2) (46.8) (-41.4) (17.6) (39.8) (n.a.) Taxes on Property 0.9 1.0 0.7 0.7 0.7 0.6 0.7 0.6 0.8 10.2 -17.3 -2.7 5.1 0.7 -13.1 23.0 25.8 Taxes on Goods and Servlces 5.1 5.0 5.0 5.4 5.4 5.6 5.4 5.6 5.8 0.9 9.2 13.5 5.9 11.3 4.9 7.0 5.7 Sales Tax (3.4) (3.1) (3-1) (3.4) (3-4) (3.9) (3.8) (4.1) (4.0) (-4.5) (7.9) (13.9) (5.8) (24. 0) (7.5) (10.1) (-1.4) Excise Tax b/ (1.2) (1.5) (1.3) (1.4) (1.4) (1.2) (1.2) (1.1) (1.5) (31.6) (-10.7) (16.6) (3.6) (-2.7) (2.4) (-2.7) (37.7) Other (°.5) (0.4) (o.6) (0.6) (O.6) (0.5) (O.4) (0.4) (0.3) (-36.9) (123.6) (5.2) (11.4) (-24.0) (-10.4) (6.3) (-l0. 6 Taxes on Foreign Trade 4.0 3.8 3.5 3.4 2.7 2.6 2.5 3.5 3.0 -0.7 -1.8 4.5 -16.0 2.7 5.4 40.8 -11.8 Impart BDtleS (3.9) (3.7) (3.4) (3 .4) (2.7) (2.6) (2.3) (3.2) (2.1) (-0.5) (-2.4) (5.5) (-16.6) (2.9) (-1.9) (4t1.6) (-30.5) Export Duties (0.1) (0.1) (0.1) ( ) I ) ( ) (0.2) (0.3) (0.9) (-8.5) (8.9) (-17.1) (11.3) (-4.6) (250.5) (33.0) (175.6) -/ - less than 0.05%. - b/ Excise taxes on gasoline, alcoholic beverages and tobacco. c/ At 1963 constant prices. Soorce; AppoJdlx Tabtle 2.1, 5.11 and 5.13 - 264 - revenues originating from foreign trade accourted for almost half of the tax revenues in 1975 and for 65% of the increase in revenues between 1973 and 1975. Because of their relatively easy collection and quick yield, the Government has at various times introduced spe!cial taxes on foreign trade in periods of financial crisis. A special surtax on nonessential inocrts was introduced in 1968 and a special 15% emergency tax on traditional exports was introduced in 1976. l/ The latter measure accounted for almost 30% of the increase in revcnues in 1976, averting a more serious collapse of Central Government finances, but further increasing the tax system's dependence on foreign trade. 271. The main problem with this close linik between tax revenues and foreign trade (particularly traditional exports) is that it amplifies the impact of the cyclical swings of Peru's primary exports on the domestic ecoaomy. In periods of buoyant export growth, Central Government revenues grow rapidly .tal stimulate an increase in the level of its expenditures (particularly current expenditures). Then, with the periodic and violent downturn of exports, tax revenues fall. But since expenditures are difficult to cut, their financing with foreign borrowing or credit from the monetary authorities fuels an excessive growth of aggregate demand. This worsens the financial crisis and the depth of the inevitable recession that follows. 2/ If tax revenues depended more on domestic transactions (and as will De discussed below there is ample room for this), the cyclical impact ofr export swings on Central Government finances would be temoered. The reliance of foreign trade taxes also runs against the Government's objective of reducing foreign dependence. Table 78: EsT3A,D TAXZS O0 FCR!IGN TIDE, 1973-75 In 3i11io: S/. ercenc o£ Tocal 1973 1975 1973 1975 Total Tax Revenue 16- R.9 C1000 '00.0 Tota1 Tax" on Foreign TridA 17.4 38.0 37.3 Ia.2 TAxes on Importa 9.1 18.2 19.5 23.0 lndependene Taes on !xports 0.3 1.8 0.5 2.3 Ineome Tazes A4ginst ExportJ a/ 0.6 4.5 1.3 5.7 Income Taxss of e±ning and Sugar Enterprlaes bI 5.0 6.1 10.7 7.7 Sales 7az on Toregng I rnasctlona 2.5 7.5 5.3 9.S &/ Eport taxes payed an advance payments of the income tax. b/ It Is estamted thAt the mi±nIg and ,u%ar sectora together accounted for 55: of pasyents from the corporate income ta. Source: Table 5.11, ?Uis±etr of Economy amd 3Ance and Mi3io= _ tinstes. I/ A small 10% tax on nontraditional exports (based on export price in- creases) was introduced in 1974. 2/ This is what occurred in the mid-1950s, in 1966-67 and in 1975-76. - 265 - 272. Other problems of the tax system from a revenue-generating point of view are complex structure and weak administration. Three taxes (income, sales and ad-valorem tax on imports) account for over 70% of tax revenues, while many other taxes yield minimal revenues and distract administrative efforts away from improving the performance of more important taxes. For example, there are 25 indirect taxes on production and consumption that together produced only SI. I billion in 1975 (around 1.2% of total tax revenues). 1/ The specific problems of tax administration are reviewed in some detail below. 273. Besides not raising an adequate amount of revenues, the tax system has not been an efficient tool to redistribute income or to improve resource allocation in the country. Although the legal rate structure of many of the taxes suggests a progressive tax system, much is lost in unnecessary deductions and exemptions. At the same time an elaborate and complex set of tax holidays has resulted in indiscriminate subsidies with little selectivity and have discouraged the use of labor in the economy. B. ISSUES IN MAJOR TAXES a. Taxes on Incomes and Profits 274. In June 1968, the unitary personal and corporate income tax as it exists today was introduced. The 1968 tax reform distinguishes between per- sons and corporations and between residents and non-residents. In formalj terms, this was one of the most significant steps in modernizing the tax system. However, other concomitant measures have weakened the potential impact of income taxes on revenue generation, income distribution and overall resource allocation. i. The Personal Income Tax 275. The 1968 income tax reform established a progressive rate structure and aimed at taxing equally all sources of income. 2/ However, after a some- what rapid growth in 1968-70, personal income taxes fell as a share of GD? to less than 1% in 1973-75 and have not proven to be an important component of tax revenues contrary to initial expectations (Table 79). 3/ They now account 1/ Many of the small taxes have a negative yield, costing more to admin- ister than what they produce in revenues. Some taxes that could be eliminated are suggested below. 2/ The 1968 law established tax rates ranging between 3 and 42%. The maxi- mum rate were raised to 55% in 1972 and in 1976 to 65%. 3/ As a share of GDP, personal income taxes peaked at 1.9% in 1970 but had fallen to only 0.9% by 1973. - 266 - for less than 1/4 of total Central Government tax revenues. 1/ The main reasz for this poor performance is that very few people pay income taxes and those who do, have an effective tax rate well below the theoretical nominal rates. As discussed below the main problems are, or. the one hand, the generous Ry-"- tions and deductions and, on the other hand, considerable evasion and weak tax administration. Table 79: 5UCu. OF TfRZ INCOME TAX, 1970-75 1970 1973 1974 1975 In millions of SoLes Tota. Income TaxL 10.8 12.8 18.8 20.5 Corporations 6.2 9.7 14.3 15.5 P?y. aainst expozrs (0.7) (0.6) C1.7) (4.5) 3P,rect paynts (5.4) (9.1) (12.6) (11.0) Perxows 4.6 3.1 4.5 5.0 Percent of total Total Income T.L 100.0 lOC.O 100.0 100.0 Corporations 57.1 76.1 T5.9 75.7 ay'. against exports (6.9) (4T.7) (8.9) (21.9) Direct paymts (50.2) (1.4) (67.o) (53.8) Persons 2.9 23.9 24.1 24.3 Percent of GDP- Total Income T&x 11 36 4.2 3.! Corportions 2.6 2.7 3.2 2.7 Pay. aainst exports (0-3) (O.!) (0.4) (0.8) Direct payments (2.3) (2.6) (2.3) (1.9) Persons 1.9 0.9 1.0 0.9 1/ Includes tax axesties wtLich are included in "Other income taxes' an Tables 2, 3 and Sourte: .Miistry of Ec=nw and Finance and Appendis Table 2.1. 276. Tables 80 and 81 give indications of which income groups paid personal income taxes in 1972 and how they compare with the patterns of income distribution in the country. 2/ The most salient points are: 1/ See Table 79 and Appendix Table 5.11. 2/ 1972 was chosen for the focus of the analysis of personal income taxes because of detailed data on who paid taxes and on income distribution were available. (The income distribution data are from the ENCA house- hold survey). Then little progress has been made, judging from aggregate receipts, which have not increased as a share of GDP (Table 79). - 267 - (a) Only around 3% of Peruvian families paid any income tax and the bulk of taxes (96%) were paid by the richest 1.5% with net annual incomes over $5,150 (S/.200,000). 1/ Moreover, 10% of the taxpayers (7000 families with incomes over S/.500,000 2/) contributed 65% of total personal income tax revenues. 3/ (b) The number of taxpayer families is a small fraction of the total in each income group, particularly in the middle in- come g.oups (Table 81). Only around 19% of the families that according to the ENCA survey were in the S/.180-200,000 income group paid taxes. The same shares are 26% for the S/.200- 260,000 income group and 50% for the over S/.261,000 income group. This suggests that there may be much underreporting for incomes in the S/.180-260,000 income groups where a large share of professionals, small commercial establishments and rental incomes from real estate are concentrated. 4/ (c) Effective income tax rates are much lower than nominal rates. The average rate for all taxpayers was only 11%. 5/ The 1/ Table 81 shows that there were around 2.6 million families in 1972 while there were only 70,422 personal income tax payers. Table 80 shows that the 37,922 tax payers with net annual incomes over S/.501,000 soles paid 96% of the actual income tax collections. (Exchange rate used was that prevailing in 1972, i.e., $1 = S/.38.70). 2/ S/.500,000 equivalent to $57,471 in 1972 dollars at the official exchange rate. 3/ Actually the number of persons filing on income tax return were 230,000 in 1972 although only 30% (or 70,000) had any taxable income. As dis- cussed later, this large share of tax returns with no income only com- plicate and increase the costs of tax administration. 4/ Table 8 also shows a striking similarity in average annual incomes wichin each income group between the taxpayers and the total number of families (comparison of columns "e" and "f" or "g" and "h" in Table 8). This can suggest two things: (a) that the average income of the taxpayers is nlot higher than that of those who do not pay taxes and that, therefore, the former are not necessarily richer than the latter; and (b) that the underreporting or possibly tax evasion may be more a problem of number of taxpayers rather than of underreporting of income, since those who file tax returns seem to report an average income similar to the average for the group as a whole. 5/ See Table 80. - 268 - Tab le 80: DISTRIBUTION OF PER'OiAL icomz TAX PAYERS, 1972 cwoe Group Number Net Taxable Calculated Effective Nominal (in 000 soles) of tax- Income Income Tax Tax1 Tax 2 payers Rate - Rate ------in n,Lillions of soles-S- 0-5O 449 14 7 0.5 3.6 3.8 51-70 1,105 68 17 1 1.5 6.4 71n100 2,063 180 47 3 1.7 10.4 101-150 15,258 1,915 360 26 1.4 16.3 151-200 13.555 2,352 634 55 2.3 20.7 201-500 ,1,241 9,394 3,770 691 7.4 31.7 501-1000 5-495 3,644 2,249 732 20.1 37.7 1001-5000 1,235 1,891 1,523 627 33.2 49.1 5001 *L2 more 21 151 144 73 48.3 55.0 Total T0422 19 8851 2209 11.3 ---Per:ent of total-- - 0-50 o.6 0.1 0.1 - 51-70 1.6 0.4 0.2 0.1 71-100 2.9 0.9 0.5 0.1 101-150 21.7 10.0 4.1 1.2 151-200 19.2 12.0 7.2 2.5 201-500 44.4 47.9 42.6 31.3 501-1000 7.8 18.6 26.5 33.1 1001-5000 1.7 9.6 17.2 28.4 5001 and more - 0.8 1.6 3.3 Ttal 100.0 100.0 100.0 100.0 / Calcalated tax as percent of net income V Calculated from the tax schedule for the upper interval of each income group. Soure!e Ministry of Eceoso!y and Finance and Mission estimates. Table 81: COMPARISO OF INCOME DISTRIBUTION WITH INCOMZ TAX PAYERS, 1972 Level of Income Distribution Distr. of Taxcayers c. d. Family Income Number of Income lumber of Set Income as 5 as S Average income (000 soles families (million Taxpayers (million of of (in million soles) year) soles) soles) a. b. 1/ ! e Total - Taxpayers - * b c d a f t h 180 - 20 49,400 10,300 9,171 1,827 18.6 17.7 0.20 0.20 201 - 260 23,400 5,700 6,137 1,439 26.2 25.3 0.24 0.23 261 and over 54,600 25,000 27,502 :2,590 50.4 50.4 o.46 0.46 180 and over 127.400 41.000 42,810 15,856 33.6 38.7 0.32 0.37 Total Peru 26,000,000 To.422 2.7 Source: Ministry of Ecouo0y and Finance - 269 - discrepancy between effective and nominal rates is parti- cularly large in the group of taxpayers with net annual in- come between S/.201,000 and S/.500,000 1/ which encompasses around 45% of taxpayers; while the nominal rate was almost 32%, the effective rate was only around 7.5%. But even in the highest income groups, (with net annual incomes over S1.501,000) which pay a large part of the taxes, the difference are considerable: the legal rate was 47% while the effective tax rate was 34%. 277. The very generous personal exemptions is the first reason why so few persons pay income taxes. These exemptions are not only among the highest in in absolute amounts in Latin America 2/ but since 1972 are indexed to the changes in minimum legal salaries. 3/ In 1975, a married couple with two children needed an income of about $6,250 (or around seven times the country's per capita GDP) to pay any personal income taxes (Table 82). 4/ Not more than 6% of Peruvian families have this level of income. 5/ Thus, because of the personal exemptions alone, most of the families in the richest quartile (except for the above mentioned top 6%), who have been the main beneficiaries of the Government's wealth distribution reforms, do not pay income taxes. 1/ In dollars, equivalent to between $5,150 and $57,500. 2/ The following table shows estimates of personal exemptions allowed in some other Latin American countries in 1972 for a married taxpayer with two children (in dollars at official exchange rates): koe=;~~~~~~I .G r Urrugay 67.2 3razil 72.8 Ecundor 32.0 Olo£.omi18. ^ 313.5 vexlezuL 8- .3 Soure: K±n.iztr7 of S=ocW anA zine and Xisuiou est+i:t*s. 3/ Between 1972 and 1975 the Lima Consumer Price Index rose 58% (see Appendix Table 9.4) and personal exemptions 56% (See Table 82). 4/ Seven times the per capita CNP in 1973 is equivalent to $43,400 in the United States, $11,480 in Argentina, $11,410 in Venezuela, $5,320 in Brazil and $3,080 in Colombia. (Source: World Tables 1976, IBRD pp. 496-698). 5/ See Table 82. Source: ENCA household survcy. - 270 - Table 82: r:=SONAL Mtl A MM=:IZ .A'AXYR WIIT TWC T IWflO , 1972-7T Per~~ooa.3 .=P ut.o f tern2a arcent of per capita eSe zti=l3 tc zc omvq 5/. 3 5 /.1' CD?' per cthan 2/ rno- eyeM-,t. - in 00- - 1972 164.o 4.24 20.9 8 6 1973 172.8 4.47 7L.5 T n.a. 1974 216.0 5.58 30.0 7 n. 1975 255.1 6.25 3T.2 7 t.c. I/ Convertod using the fol1owig ezhamige rates: S/.38.T( $3 for 1972-74 nd Z/.40.33 = ,I fcr 19T5 2/ GDP tfro Appendiz Table 2.1 di7i-dd by -ot&I po;ulatlon (rmr 15T2 from. Appendix Table 3.1 and ^or 1973-75 aasses a ttl nopulatica Trowt. rate of 2.9, per annum). 2/ Income distribution data from ECA Saurce: Appendix Tabl1! 1.1, 2.1, !inistr. o° noo and F1ne nm aid est±cates. 278. In addition to high personal exemaptions, certain types of income are taxed either at proportional rather than progressive rates or at lower average rates, and a long series of deductions are granted. Just a few examples of the more important deductions are given here. To encourage the channeling of savings into the public sector, all interest income from any type of loan to the public sector is tax exempt and 35% of the investments in COFIDE bonds can be deducted from income tax payments 1/. To also channel savings and stimulate investment in the private sector, a taxpayer can deduct from his from his net income: 50% of investments in practically any type of industry 2/; the first S/.50,000 of dividends from sharess; and any interest from the first S/.150,000 deposited in the Mortgage Bank or in savings and loan associations. 3/ To stimulate residential construction, owners of houses built after 1969 and valued at not more than S/.0;7 million soles are exempted from payment of the income tax on the presumptive rent and of the real estate tax. Finally, to reduce the differential oetween salaries in the private and public sectors, the latter's employees are allowed to deduct 30% of their salaries from net income, and to deduct all their contributions to social security from income taxes. 1/ There are three limitations to obtaining the income tax deduction from investments in COFIDE's bonds: (a) the investment must be in type "C" bonds; (b) the total amount invested zannot exceed 50% of the taxpayers' net income; and (c) the taxpayer cannot make use of other tax incentives to reinvestment. 2/ The law specifies that the investments must be in industries of the first second or third priority (See Chapter X for an explanation of the priority structure). A large share of industrial investment falls into these categories. 3/ Banco Hipotecario (Mortgage Bank) and Mutuales (Savings and Loan Associations). - 271 - 279. The differential treatment of sources of income and the numerous deductions reduce both the number of taxpayers and the progressivity of the effective tax rates of those who actually pay income taxes. It is difficult to evaluate how successfully the deductions have achieved their stated pur- pose. But the overall effect of the cumulative biiild-up of deductions has probably been that some deductions neutralize others, with a large number of persons obtaining a Government subsidy but with little impact on changing the patterns of resource allocation. Moreover, some of the deductions may riot be the most efficient means of attaining the objectives. Allowing the general level of interest rates to rise to positive level in real terms and possibly establishing a limited number of differential interest rates is a more efficient way of stimulating the holding of financial assets by the private sector and their channelling to the desired institution or sector. 1/ In- creasing public sector salaries to approximate those of the private sector, may be more efficient than the generous deductions now granted to public sector employees. 2/ These deductions reduce the equity and indiscriminate treatment of sources of income that should be aimed at if an impact is tD be made on improving income distribution. The rationale and efficiency of the deductions to encourage private sector activity is discussed in more detail below. Suffice it to mention here that some of these deductions may have been justified at the peak of private sector uncertainty when basic reforms of ownership structures were being implemented. Their justiflcation in the future is doubtful, given the extremely generous deductions that are also given on corporate income taxes. 280. Related to personal income taxes is a tax on payrolls introduced in 1973. 3/ The tax is levied on all salaries of private and public sector employees and on incomes of independent professionals. 4/ Because of the ease I/ It is often argued that increasing interest rates would be costly for the public sector. However, the income tax revenues now lost plus the distortions on capital markets and in the equity of the tax system are more serious losses. 2/ While at the upper technical levels Central Government salaries are -lot competitive with the private sector, this may not be the case at the lower levels which have received large salary increases in recent years. 31 The tax consolidated two earlier taxes: a payroll tax and a contribution to the National Health Fund (Fondo Nacional de Salud). The payroll tax is remnant of the system of schedular taxes (which taxed different sources of income at different rates) that existed before the income tax reform of 1968. 4/ The tax rates are 3.5% on salaries (2.5% paid by the emplover and 1.0% by the employee) and 1% on incomes of independent professionals. The bulk of the revenues come from the tax on salaries. The tax is collected and used by the National Health Fund although it is formally considered a Central Government revenue. - 272 - of its collection, the tax yields 2/3 as much as the personal income tax. 1/ However, the tax is regressive and increases the cost of labor. 281. In sum, many changes could be made to laws affecting personal income taxation in order to increase the revenue yields (from the 0.9% of GDP in 1975 to at least 2-3% of GDP) and to pursue other development objectives more efficiently. This report recommends the following changes: (a) Modify the schedule of income tax :rates. At present the schedule starts with a 6% rate for incomes under S/.30,000 and after 19 levels reaches a maximnum rate of 65% for incomes above S/.5 million. They could usefully be reduced to not more than six levels that start at a minimum rate of 12% with a maximum rate not above 55%. This would simplify tax administration and, while maintaining a sharp progressivity, would tax a higher share in the high middle income groups than now. (b) Reduce drastically the levels of personal and family exemptions so that at least the richest 10% of the population are taxed. This could be achieved relatively fast by freezing the present level of exemptions in absolute terms for a few years (i.e. by temporarily not indexing the exemptions to changes in minimum salaries). (c) Eliminate a large share of the deductions and exemptions now granted. In particular, consideration should be given to eliminating the: (i) deductions for reinvestment in industry (government subsidies and incentives should be concentrated on the corporate income tax); (ii) exemption of income from interest on loans to the public sector and from shares; (iii) exemption of presumptive income from new residential construction; and (iv) exemption of income from deposits in the Mortgage Bank and the "Mutuales". Whatever objectives are sought by these deductions, they can be more efficiently pursued with other instruments. (d) Eliminate the special treatment accorded to public sector employees. (e) Establish minimum presumptive incomes for independent professionals. One possibility is to use the number of years of membership in the professional registry as a proxy for income and cross check it with information from the Social Security Administration. 1/ In 1975 the personal income tax yielded 6% of Central Government tax revenues and the payroll tax 4% (Appendix Table 5.11). - 273 - (f) Eliminate the regressive payroll tax because of its undesirable impact on the cost of labor. Given its present importance for total tax revenues, elimination would have to be gradual as of the personal income tax increase. ii. The Corporate Income Tax 282. The 1968 tax reform also introduced a progressive rate schedule for the corporate income tax that now ranges from 20% for profits below S/.0.1 billion to 55% for profits above S/.1 billion. 1/ Corporate income taxes account for around one-fifth of Central Government tax revenues, but the tax burden is only 2.5% of GDP. 2/ Yields are low principally because of the generous and widespread tax holidays granted, although the rate structure has other defects that tend to favor the larger corporations. 283. Tables 83 and 84 give indicators on companies and sectors which contributed most to revenues from the corporate income tax in 1973. 3/ The following are the most salient conclusions: (a) A small number of companies pay most of the corporate income tax revenues. Although around 15,000 companies filed income tax returns in 1973, over one-half did not pay taxes, because they either declared losses or claimed to have no taxable income. 4/ Of the 7,469 companies that did pay, 2,051 gener- ated 97% of total revenues, and the two largest corporations 1/ The 1968 income tax law specified a tax schedule that went from 20 to 35%. However, in 1969, the rate schedule was exDanded and increased for profits above S/. 0.5 billion. The rationale for expanding the rate schedule was to be able to capture fluctuating revenues in the mining and sugar exporting sectors. Another feature of the Peruvian corporate income tax is that a part of the revenues (Table 79) are paid as export taxes to be charged against income tax liabilities. 2/ In 1975. See Tables 79 and Appendix Table 5.11. 3/ 1973 is the most recent year for which detailed information on the corporate income tax is available. Little has changed since then in :he legislation of the tax (except for the inclusion of cooperatives and public enterprises in 1976) and 1973 was not an unusual year (like 1974 when mineral prices boomed or 1975 when mineral prices collapsed). 4/ The share of companies declaring not to have taxable income is high: 80% declared losses. However, the large number of tax returns that have to be processed without producing any revenues; increase the costs and distract the attention of tax administration. - 274 - in the country -- Southern Peru Copper Corporation (SPCC) and the Cerro de Pasco Corporation (now CENTROMIN) -- paid almost 40% of the total. 1/ (b) The effective tax rates are low and the differences between net incomes and taxable incomes are very large. The average effective tax rate for all compan:es is only 26%. 2/ Had the effective tax rate been equal to the estimated nominal tax rate (around 37%), tax revenues would have been about 40% higher. 3/ The larger the corporation, the lower its effective tax rates and the larger the difference between its net income and its taxable income (excluding the two largest mining companies). 4/ While relatively small companies with net incomes below S/.0.5 million 5/ had an effective tax rate of 26% and no difference between taxable income and net income, large companies with net incomes in the S/. 100-500 million range 6/ paid less than 18% in corporate income taxes, and on average their taxable income was only 46% of their net income. (c) The mining sector paid over one-half of all corporate income taxes and had the highest effective tax rates. Industry was taxed a lower share of its net income than agriculture or services and its rates of taxable income to net income were well below those of any other sector. 1/ 14% of companies filing a tax return paid 97.4% of the taxes. 2/ Excluding the two largest mining companies, the average effective rate falls to 21.9% against an estimated ncminal rate of 31.3%. 3/ This is a hypothetical case, but it illustrates how much revenue is lost with the erosion of the taxable base of the corporate income tax. The 40% estimated difference in revenues is equivalent to 1% of GDP. 4/ Even the tax base of the two largest mLining companies has been considerably eroded. Against a nominal tax rate of over S1%, the effective tax rate was only 37.2% implying a loss of arotnd S/.1 billion in tax revenues (equivalent to about 2% of 1973 Central Government total tax revenues). Table 83). 5/ S/.0.5 million equivalent to around $12,900 (at 1973 official exchange rates). 6/ S/.100-500 million equivalent to $2.6-12.9 million (at 1973 official exchange rates). Table 83 PJz sm3: o0 comTs acme T3C A sX ZYm U uUM2 -~~~~~ ~ ~ ~ M AMBY 1913 roae xTaa.f 3cinal Ratio of t"z an (J= V~~it taC. 3 Txble U ti-M %z al i*u* w (Lo~4joe "* aml. m In- AEesseed 3/ rue t- to ant mom seter Sol") d. ~~~~~~~~I -!i b- . A. BI LXM 01 !ICU -in .iLUm- -p.romt- O - 0.50 5218 5800 80 206 25.7 25.7 100 0.51- 50.0 1999 11322 8976 263k 23.3 29.3 79 -- - 50.1 - 100.0 35 3661 2757 873 23.9 31T.7 75 100.1 - 500.0 15 6353 2954 1137 17.9 38.5 2o 500.1 ad ovr 2 83T7 6057 3113 31.2 51.2 72 TS21 IL69° MU= 2156 122 I/ 26 1 2u- M -Nircet of -Vta1- 0 - 0.50 72.5 2.6 3.t 2.6 0.S- 50.3 26.8 37.1 11.6 33.1 50.1 - 100.0 0.5 12.0 12.8 11.0 100.1 - 500.0 0.2 20.8 13.T 1%.3 500.1 and or 0.1 2T.5 28.1 39.1 Totat 100.0 100.0 100.0 100.0 3. BY s Agiculture and 106 1208 917 223 18.5 24.3 76 fisberi.. Wixing 95 122I6 8718 4i61 3i.0 47.6 n1 T.zd-zMr7 1925 9101 i35a 1414 15.5 32.5 2a Czwtction 1S6 410 1042 36 8.5 9.0 98 Co=erce 2917 4602 41433 1433 31.1 32.3 96 Trs.msport Lad 271 701 612 154 22.0 25.2 87 Ccz=micstioos Fersonal and 1950 216 2052 524 22.0 25.5 92 F4enc. Se*r. 3o0 fp.kci.d 19 55 46 16 29.1 31h.8 8i Ttal T69 30511 21569 7962 1/ 26.1 36.9 71 -percet of total- Agrlou1tre ad 1.41 4 .0 2.3 2.8 1J rishen4ss miziag 1.3 20.1 20.6 52.3 5 ndustry T27.8 29.8 20.2 17.8 21 0ocstrntion 2.5 1.3 1.9 0.5 _ .rVe 39.1 15.1 20.6 18.0 Truanport and 3.6 2.3 2.d 1.9 Commuicatimsr .5 ? 'rs Zal &-ad 26.1 7.2 9.5 6.6 , L ?insLc. Ser. xo . specirie4 0.3 0.2 0.2 0.2 Total 100.0 100.0 100.0 100.0 100 1/ Tx ssaessed oat eqn.. to ctual collections shown on TAbl*79 become of advane p.in7ftU. / At Cl.statv. '963 pri*e*. Source Ainiur of Z2o%o and inance - A*pendix Table 2.16 - and X4issioa estiates. Table 84: ?M: ctAs3TYcaTzo OF!T 3COYS MP Co0ORA m3co TAmAE BY LarEM OF ZN,M AND SZB02, 19T3 (I pernetages of total) of incomne ni3 llioo soles lotal T l 0 . ,1- 50.1- 100.1 Ta:ble lumber of Sector 0.50 50.0 100.0 and o0r OcoeU Tapayers Ag; c<1 , = an d ishetrles 0.1 2.2 0.7 1.1 4.0 l 4 .%Czing a.a 1.3 2.5 35.6 210._ 1.3 ;_c:r try :0.4 15.2 6.o1 7.9 z r 27. 2onastr.tian 0.1 0.5 0.0 0.7 13 2.5 Coo nrca 1.0 11.2 1.3 1.6 15; 39.1 tsno. Lad icmioz 0.1 1.3 0.0 1.0 2i. 3.6 Serv%al 03.8 2.3 1.3 0. l 7.2 26.1 Other 0.0 0.2 0.3 0.0 0.2 0.3 tal Tasable ooC 2.5 37 1 12.0 48.3 lW.o toa 3mber of taxpayers 72 5 26.3 3 l000 2JI oo~c-o Sir7 ,,C 511n_tr and -i-.See -eb e !'3e _i 1 -o TL69 tcy-::r_ (Sea 7&!:'a S33 Sc=it-tc f i ir of =d r c ewce oe ssiom- estirAez - 276 - 284. The main conclusion that emerges is t:hat the taxable base of the corporate income tax has been sharply reduced by generous tax deductions and exemptions. The most important set of tax hol:days are those meant to stimulate investment and economic growth. Through individual seccor laws, practically every sector receives generous government subsidies via tax holidays, particularly of income taxes. 1/ The second set of large tax subsidies are related to the reforms of ownershiip structures since 1968. To promote the participation of workers in the profits and ownership of firms, a share of net corporate income (ranging from 10% in mining to 25% in industry) is deductible from taxable income; 2/ and to stimulate their development many of the new public enterprises and all the agricultural cooperatives were exempted from the corporate income tax until 1976. 3/ 285. The cost to the Government of these incentives has been large. Excluding the unpaid taxes of the public enterprises and the agricultural cooperatives, the cost to the Government of corporate income tax holidays alone rose from 0.9% of GDP in 1972 to 1.3% in 1974 and accounts for around 35-40% of total potential corporate income tax revenues (Table 85). The most favored sector has been industry, which in 1973 absorbed almost one-half of the subsidies granted, reducing its tax bill by 60% (Table 86). 4/ 286. Has this high level of subsidies been an efficient way to foster growth and development in the direction of thE Government's objectives? Very little in-depth study has been done of the impact of the income tax holidays. However, the analysis of industrial policies in Chapter X and Tables 87-88 suggests that tax holidays may have stimulated a higher level of capital formation than would otherwise have occurred, but their overall impact on development has probably not been very positive for the following reasons: 1/ The tax holidays granted from import duties, personal income taxes,. property taxes and sales taxes are discussed ir. other parts of this chapter. Concerning corporate income taxes the following are the main tax holidays. Firms in first, second or third priority industrial subsectors can reinvest tax-free between 65-85% of their incomes (depending on priority) if they are at Lima and between 85 and 99%, if they are outside Lima (See Chapter X for a definition of the industrial priorities). Mining firms pay corporate income taxes in lieu of all other taxes, however, 40% of the profits or S/.300 million (which- ever is lower) are tax deductible if reinvested. The share of profits reinvestable tax free is 75% for tourism and between 55 and 100% for fisheries. 2/ In the case of industry, 10% of the firmLs' net income is distributed to the workers in cash and an additional 1'% is capitalized in shares for the workers. 3/ However, given the inadequate pricing policies:a -d efficiency problems few public enterprises have profits (See Chapter VI), and a way to effectively tax the cooperatives has nol: yet been developed. 4/ Mining accounted for 40% of the taxes foregone, and its tax bill was reduced by 30% (see Table 86). - 277 - (a) The tax holidays are not channeling resources towards the Government's high priority sectors. The policy of using tax holidays to stimulate investment dates back to the first mining laws in the early 1900s and to the industrial law of the late 1950s, but since 1968 their use has increased rapidly. The principal stated objective is to guide invest- ment and resource allocation to the priority sectors and geographical regions. But the sequential pyramiding of incentives granted to practically every sector has reduced most of their selective impact on resource allocation. As shown in Chapter X, the tax holidays considerably increase the financial rate of return of projects to entrepreneurs, but do not create significant differentials between, for example, projects of first or second priority in industry or between Lima and the rest of the country. (b) The structure of income tax holidays discriminates against small-scale enterprises by granting larger subsidies to larger enterprises. This is demonstrated by the hypothetical examples shown in Tables 87 and 88. The examples show that a small firm with a pre-tax net income of SI.50 million ($0.7 million) encs up paying out of its own funds 80% of the costs of the tax deductible reinvestment and of the contribution to the indus- trial community. At the other extreme, a large firm with a pre-tax income of S/.1.25 billion ($17 million) pays out of its own funds less than 1/3 of the costs of the industrial community and 42% of the tax deductible reinvestment. This bias has various negative effects. It implies a transfer of resources from the rest of the country to the larger firms, with obvious regressive implications. Perhaps more important this differential subsidy goes counter to the Government's stated objectives of distributing wealth more widely and promoting grassroots entrepreneurial developments in the poorer income groups. To the extent that small enterprises, which use local resources, skills and technologies, can be efficiently developed, they are put at a disadvantage vis- a-vis large-scale capital intensive firm. (c) The tax holidays discriminate against labor intensive by concentrating the subsidies on investment in capital goods. (d) The tax holidays, by encouraging reinvestment within the same corporation, restrict the mobility of capital to projects of highest return and thus foster inefficiency in resource use. 287. The above analysis suggests that the present system of income :ax holidays should be urgently studied and thoroughly reformed. With the iiifor- mation available it is not possible in this report to determine if incentives are needed or not to achieve the desired levels of investment. If they are needed at all, however, the present level is clearly excessive. Moreover, a reformed system of income tax incentives should remove some of the major distortions on resource allocation. One possibility is to grant investment - 278 - Oble 85: F.CAL COT 3? 11CE"P/S GRAY1¶M VOUG11 V,CZ:' TA=-S, 197T-T4 Actul ?otenti&l Income Tax noot 2aid tax rac Reinvc'stment ther Total revenues revenu a. b. c-a+b d. eOc+d -- in =ill_onz o: soles--- 1972 1553 1042 2595 4426 7021 1973 2903 162 4524 7o62 12he6 19714 3.150 277O am 10971. 1689i pertet or total -- 1972 22.2 .148 370 630 100.0 1973 23.3 130 363 637 100.0 19741 18.6 164 350 650 100.T I/ Zncl_des tvxes mot collected fcr: (a) reinvest=ent in the seze fira; ('a reiare-C_erl.t in other ri:=s; (c) investneat in CO?F- ;3y-ve "C" bonds; and (d) investrmet in rqionc.l baonts. 2/ Includes taxes not co" ected for (a) contributinns to scierce and technol;r' re-earch (ITNIJS-C); (b) wrkerz Profit-shariza; and (c) reinvestment by Ildustrial Cco nity. / The actual t= revenues ia this table do 06t coincide vith those cn Table ' tccause of advance pe7-mnts. 4/ Frcllminmry estimate. Scurce: 4 s-.r of Zconosr &nd ?ina=ce anc' S Lion esti=-mnzz tax credits with a carry-forward provision instead of tax deductions. The advantages of a tax credit are that it would not discriminate in favor of undistributed profits as a source of funds (that-reduces the mobility of capital) and would also not discriminate in favor of the larger enterprises. In addition, tax credits would simplify and increse the effectiveness of tax administration. Finally, if special benefits are to be given for the promotion of particular goods or for decentralization, they should be offered on a much more slective basis than is currently the case. 288. Concerning the share of profits tha.t are distributed to the workers in cash and in capitalized shares, they should be computed after taxes and not before taxes as is now the case. As shown above, the Government is paying for a substantial share of the wealth and profit distribution that is bene- fiting the workers of the modern formal sectors in the economy. While the participation of" workers in the benefits generated in their firms is a laudable objective, the additional government subsidy seems much less justi- fied, particularly since these workers are not among the poorest half of the population. 289. The second problenmvzS the corporatE income tax is that a progressive rate schedule has little justification on the:oretical and practical grounds. The principal arguments put forward in justifying a progressive rate schedule - 279 - Table 86 s3= 3 or 3wae 2sAX =MQ ONS 37 sCQ, 1973 ?ax Actual Potential ?ercent t AOtAati sot Ta T7a eenue Not TA Rine ?" ~ ~ ~ ~~~~~~~~ t~~~~~.7 9.1 U2.5 10.1 23.2 gel 315.0 h.l 3 9.3 &.A 1.4 1.0 5iorra 6.3 0.1 1.6 3.3 3.0 O.S 3i*rrs IF,' f .0 2r.2?5 3.1 -o" 5.$ . .9 1.2 '. 2.0 Tat &I 6.6 <39 26.9 12.1- 15.5 Sior 3. 37.3 23.7 ?7.S Ccwta 15.3 0.1 l.i4 3.0 3.6 T. 7 laa &M7 5. d1. 2. Si* - 5 71 .6 2.4 rs l~ ~ ~ ~~~~ ~~~ ~~.3 b.40 2.3 I.h 0.53 i-tIto * 1.2 : S,. I ~~1. Ca %a. 0.5 5.2 5.l1. Sierm 9.4 0.1 2.7 li 1.5 Coo% Wi. 9 1.2 S.2 15.3 8.4 13.3 agots T~0 .5 1. + TV ! 2 . 3T, 3SirL 27TJ *.l 12.k 9.0 1.3 0.4 5e1fa 2.1 - 3. .6 D.6 0. 8~~~~~~~~~~~~~. -16 0. -H T S 11 Ca" 12.0 .1.3 3.5 - 1.5 1.6 5 1, 9.3 0.2 3.9 2.3 1.' 2.9 Total LM-77 Z1.2 315 15.5 ?, 73,0- Siar 9?i.3 20.7 30.7 1.5.7 7.0 23.3 Tota / 22.3 %2.0 2k.2 5.5 . Slarr1. 13.0 7 -0. L3D3.0 O Costa 25.7 9.6 6.9 5.T *.7 Q.T 54s '1~~~ ~~~~ ~ ~~~3.2 o.6 3.0 4.9 3.3 . .83 Eu or COATS Sierr 67. - - 3 2.2 3. 2 4 1 *5 S::App-adi. Table. T.215 W1'.3et - 300 - Tble 94: INDICATORS OF USE OF FERTILIZERS, PURCIHASED SEEDS, TRACTORS, CREDIT AND EXSTION SERVICES BY FARm SIZE AND HATUEAL MaIOw, 1972 * Farm Size All Sizes (100 20-100 5-20 .5 A. PERCENTAGE OF FAIMS USING FERTILIZERS,/ Perua/ 15 21 18 20 14 Costa 38 85 66 58 32 Sierra 13 15 18 17 12 B. PERCENTAGE OF FARUMS USING PURCM,ASED SEEDS-/ Costa 42 82 71 64 37 Sierra 16 18 22 22 15 C. TRACTORS OWNTED PEM 1,000 HECTARES OF HARVESTED LANDz- Pertr/ 4.1 11.0 6.7 3.0 1. 2 Costa 12.1 14.2 17.8 10.2 5.2 Sierra 2.9 10.9 8.3 5.8 2.1 D. PERCENTAGE OF FARMS RE:CEITVING CREDIT-/ )?ertu 5.4 19.0 12.8 11.0 3.5 Costa 17.0 69.6 48.3 36.2 11.4 Sierra 2.9 10.9 8.3 5.8 2.1 E. PERCENTAGE OF FARMIS RECEIV2ING EXENlSION SERVICES-/ Peru-t 3.4 13.9 6.1 6.1 2.4 Costa 8.3 41.7 27.4 16.9 5.5 Sierra 2.4 11.0 6.2 4.2 1.9 * No data for 1967 available. However, little change occurred between 1967 and 1972 in use of technology. a/ lncludes Selva. b/ Chemical and organic fertilizer. c/ Share of farms growing crops that use conmercially purchased seeds. d/ Harvested land includes cropl1nd plus area render perennial pastures. e/ SRefers to any type and source of credit wliether public or private. f/ Refers to any type and source o'L cl:etision servRices whetlher public or private. Source: Data froi-7z 1972 Agricultural Co.:--sus proccs_euO1 'n Appendix Tahble 7.21 to 7.30. - 301 - workers were in the richest half of the populations and 50% were in the richest quartile. 1/ However, in their drive towards efficiency during the 1960s, the large Coastal estates mechanized heavily and cut employment dras- tically. 2/ By 1967 they employed only 65,000 permanent and regular part-time wageworkers (less than 4.5% of the rural families) and an unknown number of seasonal wageworkers. 3/ 326. Second in rank were the medium sized farms (20-100 ha) of the Costa. Although they held less than 4% of the croplands, almost all were irrigated and cultivated with modern techniques. 4/ These farms produced about one-fifth of Peru's cotton output and important shares of the crops marketed domestically (particularly rice, maize and vegetables). 5/ They provided little permanent employment, but did hire a substantial number of seasonal workers. 327. Developments in the Sierra -- where most of the rural population is concentrated -- were diametrically different and the agrarian economy has stagnated throughout the centuries. In the colonial period the traditional growing techniques of the Incas were abandoned and massive deforestation took place, 6/ causing erosion in a large share of the scarce good lands. Since then, the Sierra's agrarian economy has suffered from the neglect and the pro-urban bias of Government policies. A policy of cheap food prices has been followed most of the time, affecting little the export-oriented Costa agriculture, but lowering production incentives in the Sierra. Because of little public investment, many areas of the Sierra remained physically isolated and most have had little economic infrastructure and a minimal supply of public services. 1/ See Webb, R., "Politicas Gubernamentales y Distribucion del Ingreso en el Peru: 1961-1971" in Webb, R. and Figueroa, A., Distribucion del Ingreso en el Peru, Instituto de Estudios Peruanos, Lima, 1975. 2/ The sugar estates halved employment and in other crops such as rice and cotton, the remaining labor-tenancy arrangements were eliminated. 3/ See Table 99. 4/ See Table 94 and Appendix Table 7.6. 5/ See Table 93. The shares of marketed production accounted for by the medium-size farms are much higher than those shown in Table 93 which include the production for subsistence of the small farmers (which is not marketed). 6/ Timber was required for mine tunneling and fuel. - 302 - 328. By 1967, around 40% of the Sierra's farmlands were large rentier haciendas, most of them managed with semi-feudal tenancy arrangements. 1/ They had the best pastures, but not the biest croplands which were owned by the medium- and small-farmers. 2/ In gerneral the technology used in these haciendas was primitive, investment by owmers, small, and productivity, low. 3/ Consequently, they were relatively unimportant in aggregate agricultural production, being significant only in sheep, cattle and potatoes. 4/ Labor was provided predominantly by tenants (about 15% of the rural families 5/), who individually farmed small plots of the hacienda in exchange for rent, a share of the crops or a provision of free labor to the rest of the hacienda. An unknown number of minifundistas and landless wageworkers were also employed seasonally. Various studies have argued that the Sierra haciendas grossly underutilized their resources by failing to modernize production methods and by grazing land which was suitable for cuiltivation. 6/ While this was true for many haciendas, the potential for improvement was often exaggerated. Horton found that a large share of Peru's "cultivable but uncultivated" lands was held by the Sierra haciendas, but that most of these lands were in "remote areas characterized by hazardous growing conditions and poor economic infra- structure and markets". 7/ Moreover, sharp variations in the quality of land, 1/ See Table 99 and Appendix Table 7.6. With time a complex variety of tenancy arrangements developed in the haciendas. In a simplified form there were two basic types of tenants. The tenant worker provided free labor to the landlord in exchange for the right to use for his own benefit (for cropping or grazing) a plot of land belonging to the estate. In some cases, the tenancy system also included the paying of a rent or a share of the crop for the use of the land. The tenant farmer, on the other hand, did not provide any free labor to the landlord, only rent or a share of the crops. The distinction between tenant workers and tenant farmers acquired importance after the agrarian reform, since the former in general became full members of the ARE (but kept their plots of land rent free), while the latter in general became independent small farmers. 2/ See Appendix Table 7.81 and Horton., D., op. cit., p. 24. 3/ Table 94 shows that in 1972 less than one-fifth of the Sierra estates used fertilizers or purchased seeds and only about one-tenth received credit or extension services. 4/ See Table 93. The estates over 100 ha in the Sierra accounted in 1972 for 22% of Peru's production of sheep, 8% of cattle and of potatoes. In the other crops the shares were well below 5%. 5/ See Table 99. Around 219,000 families. 6/ See, for example, CIDA, Tenencia de la Tierra y Desarrollo Socio-Econo- mico del Sector Agricola, OAS, Washington, D.C., 1966. 7/ See Horton, D., op. cit., p. 24. - 303 - growing conditions and microclimates within the same hacienda, made it difficult for a central type of management to be more efficient than independent small farmers. For the owner it was therefore more profitable to simply collect rent from the small scale tenant producers than to run the farm with a central admi- nistration. 1/ In these circumstances, the few haciendas that did modernize moved from tenant-mixed farming to extensive sheep-production, the only way to achieve economies of scale. Although sheep output expanded rapidly, total employment in the estate declined. 2/ Employment and value of production could have been increased in some areas, but this would have required a much smaller farm size. 329. Another 6% of the Sierra farmland were in medium-sized farms. With more croplands than in the Sierra haciendas, their contribution to total output was larger, but still not of major national significance. 3/ However, the quality of lands, the technology used and consequently the yields were well below the Costa medium-sized farms. 4/ I/ See Ibid, p. 25a. 2/ The peasants, of course, resisted these changes and up to the time of the land reform had been successful in resisting eviction from many haciendas. (See Horton, D., op. cit., p. 29). 3/ See Table 93. 4/ Table 94 shows that three times as many Costa medium-farms than Sierra medium-farms used fertilizers or purchase seeds; six times as many received credit and four times as many received extension services. The Costa medium-farms also had the highest number of tractors/ha in. the country; six times as high as in the Sierra medium-farms. The following table gives examples of the differences in yield/hectare betweeen the medium-farms in the Costa and those in the Sierra according to the 1972 Census (in tons/hectare). Can=- Sierra zLea 4.8 2.8 maims Cf2nt) 2.8 0.9 visae 21.3 0.5 Iazicy 1.4 0.6 racaoels 1 .9 L5 Other roc cxope 7.3 1.4 Citrus 9.3 3.7 Othatze futs- 5.6 1.5 Y.a5tabla 7.7 5.0 rodder =Wops 22.9 3.5 al/nOa eot includ. casav. El olss not iancudea bm=aS. soure: AAppend±x 7bl 7.26. - 304 - 330. The rest of the farmland in the Sierra and the Costa was held by the small farmers, 70% of the rural famil:Les. I/ Their individual plots covered 11X of Peru's gross farm area (two-thirds of which-were croplands), 2/ but 40% of them also farmed the communal "Lands, extensive natural pastures and woodlands in the highest and poorest parts of the Sierra covering an additional 35% of the gross farm area. 3/ The small farmers were important in the production of crops and livestock, but a large share of their output was for subsistence and was not marketed. 4/ 331. However, almost 40% of Peru's riaral families had less than 2 ha of farmland. 5/ These minifundistas together with the landless workers not permanently employed in the modern coastal estates (around 20% of the rural 1/ Around 990,000 families (Table 99). 2/ See Appendix Table 7.6. 3/ The communal lands were large estates owned jointly by one or more peasant communities. In the words of Horton: "Historically all land within the jurisdiction of a community was the domain of that community, and its use was determined by communal authorities. Pastureland was generally held in common, and all ccmmunity members grazed their live- stock on it. Cropland was typically parcelled among community members for individual exploitation. A small portion of the cropland was often cultivated communally, and the proceeds financed community projects. Each community had a hierarchy of puLblic officials for management of community affairs. Labor pooling was a common form of taxation for community projects, and labor exchanges were common between individual households (especially between close relatives and neighbors)." (See Horton, D., op. cit., p. 170. The L972 Census estimated that the com- munal lands involve 1,344 units (of more than 100 ha) with 6.12 million ha. There are many more communal lands which belong to traditional com- munities not yet recognized by the 'Law. Based on various Government surveys (particularly the "Diagnostico de los Centros Poblados "Rurales", Direccion de Comunidades Campesinas, SINAMOS, 1975), the mission estimates that communal lands total around 4,000 estates and cover 8.2 million ha (Table 99). 4/ The farms under 20 ha produced arouad 40% of rice; two-thirds of maize and sheep; three-fourths of wheat, barley, beans, potatoes, bananas, vegetables and cattle; and 90% of hogs and goats. 5/ See Appendix Tables 7.22 and 7.27. - 305 - families 1/ were the poorest at the base of Peru's agrarian structure. Over 80% of the minifundistas were in the Sierra. A survey in 1971-72 estimated the annual per capita expenditures of rural families at around $125, but the per capita expenditures of 70% of the rural people in the-Sierra did not surpass $86. 2/ A study based on the same survey estimated that even after taking into account regional price differences, these families accounted for the poorest 40% of Peru's population. 3/ 332. The archetype of a minifundista was the family of Indian descent in the higher parts of the Sierra having less than 0.5 ha of very poor cropland and a few sheep. Harsh climatic conditions made cultivation in their plot a constant risk. With practically no access to credit and Government assistance, the technology they used was primitive and had changed little or nothing since the time of the Incas. 4/ Whatever meager surplus they could produce for barter or cash sale, was difficult to 1/ Approximately 26% of the rural families did not own land. Besides the 219,000 tenants and the 95,000 permanent wageworkers in the medium and large estates, there were around 67,000 families employed in seasonal or occasional jobs in farms or in nonagricultural activities (Table 99). The incomes of many of the landless workers were among the lowest in the country, but there were sharp differences that make generalizations difficult. (See Webb, op. cit., pp. 38-43) Horton's survey of various enterprises showed that the daily cash wages (including fringe benefits) just prior to the land reform were around S/.80-100 in the coastal sugar estates and S/.60-70 in the smaller coastal estates producing cotton and potatoes. In the Sierra, on the other hand, one of the most modern live- stock haciendas paid S/.25-27/day and the smaller cropping estates of Cuzco, mainly operated by tenants, paid less than S/.19/day. (See Horton, D. op. cit. pp. 38-43). Figueroa also found that in the coastal sugar estates the average monthly income of the permanent workers was over 2.5 times larger than that of the occasional seasonal workers (See Figueroa, A., "Redistribucion del Ingreso y la Propiedad en el Peru, 1968--1973", in Webb, R, and Figueroa, A., op. cit., p. 134). 2/ ENCA (National Food Consumption Survey). Data prepared by the Ministry of Economy and Finance. Exchange rate prevailing in 1971-72 used for conversion (S/.38.7 - $1); to the extent that this exchange rate! over- valued the Sol, then in dollar terms per capita figures are even lower. 3/ Thomas, V., The Measurement of Spatial Differences in Poverty: The Case of Peru, World Bank Staff Working Papers No. 273, Washington, D.C., 1978. If regional price differences are not taken into account then the 4 million poorest of the rural Sierra account for 83% of the poorest 40% of Peru's population. 4/ Table 94 gives indications of technology use of the farms with less than 5 ha: less than 15% used fertilizers or purchased seeds and less than 2% receive any credit of extension services. The minifundistas with less than 2 ha fared even worse. - 306 - market since road connections were unusable during the rainy harvest season or were inexistent. the work on the plot filled only 60% of the annual working days 1/ of the family and covered slightly over half of their con- sumption. 2/ For survival, the minifundista family depended on seasonal employment in the larger farms -- which was stagnant or declining -- and on rare opportunities for non-agricultural employment. 3/ Their diets were unbalanced and deficient, the majority were still illiterate 4/ and whatever health facilities that existed were distant and hard to reach. In this desperate scenario, migration to an urban or mining center provided one of the only hopes for better living standards, but mining jobs were scarce and many minifundistas could not iiave enough to finance the trip and the survival in the first days in the city. 333. In this scenario of inadequate agrarian structure and rural poverty, the agricultural sector has been for many years in a deep crisis. Political discussions on the problems in the Sierra and on land reform go back to the first quarter of this century. 5/ However, growing peasant unrest and the lagging agricultural performance in the Sierra made land reform a major political issue by the mid-1950s. 6/ Land invasions became a common occurrence J A study by R. van de Wetering (Primera Estimacion del Subempleo de la Poblacion Economicamente Activa Agricola en Areas Rurales en el 1967, mimeo, Lima, 1970) estimated that in 1967 the full-time equivalent employ- ment in agriculture was only some two-thirds of the nominal. Another survey by the Ministry of Labor in 1971 found a similar result. This, however, does not mean that one-third of the labor force in agriculture is superfluous; what it means is that the number of man-days actually worked is only two-thirds of the number available if the full labor force worked full-time all year. In fact, van de Wetering found that there is no absolute surplus of labor (in terms of persons) in agricul- ture; on the contrary, substantial deficits occur in some part of the year. The real problem is the sharp seasonality of employment, parti- cularly in the Sierra where usually only one crop is produced yearly and the cultivation cycle of most crops occurs at the same time of the year. Another study by Figueroa eEstimated that the minifundistas spend only 60% of their working time in their plots. 2/ Source: 1972 ENCA Survey. 3/ According to the 1972 ENCA Survey, almost 40% of the expenditures of the minifundistas are financed with salaries employment. A study in Cajamarca by the Ford Foundation found an even higher share (50%). 4/ See Appendix Table 1.63. 5/ For a brief review of land reform discussions in the earlier half of the century, see Horton, D., op. cit. p. 32. 6/ In the 1956 presidential elections all major candidates advocated land reform in one form or another (see Horton, D., op. cit. p.34). - 307 - in the early 1960s and in 1964 the Belaunde Administration passed Peru's first land reform law. Little progress was made, however, since the law faced strong political opposition and was excessively complex and costly to implement. Nevertheless, the law accelerated the change of expectations in rural areas. Fearing that land reform was now inevitable and only a matter of time, land- owners stopped investment and decapitalized their estates. 1/ At the same time, labor management conflicts sharpened, land profits fell, further discour- aging investment. 2/ By the late 1960s, the point had been reached that the longer the delay for effective land reform, the worse the decapital- ization and turmoil in rural areas. B. RECENT TRENDS IN OUTPUT AND FOREIGN TRADE 334. The uncertainty, the peasant unrest and the pro-urban Government policies all fed each other to lower investment and reduce the growth of agricultural output. Production data on the agricultural sector are scarce and of low quality, 3/ but most studies and the data on Tables 95 and 96 suggest the following trends. Since the early 1960s, aggregate output growth has been lower than in the 1950s and has not kept pace with population growth. Much of the relatively rapid growth in the 1950s was due to an expansion of export crops (cotton, sugar and coffee), while the very slow growth turing the 1960s was associated with contractions of cotton and sugar output. ?roduction of crops for domestic consumption has been much less dynamic throughout the period; earlier studies have concluded that the growth of food output was already slower than population growth in the 1950s. A few livestock products (particularly poultry and eggs, but also pork and milk) have been the only dynamic products during the last fifteen years. The share of agriculture in total GNP has fallen rapidly from about one-fourth in the early 1950s to about one-tenth in recent years. The structure of agricultural output has also changed markedly with livestock products now accounting for about one-third of agricultural value-added. 1/ Decapitalization occured in the Sierra livestock estates by liquidation of stocks and no maintenance investment (for example, Horton estimated that the number of sheep in the Department of Puno fell from 6 million in 1961 to 4 million in 1972, a significant part being sold prior to 1969 (see Horton, D., op. cit., pp. 162-163), and in the Costa by no maintenance of tractors and machinery. The exceptions were the large agroindustrial complexes of the Costa which were explicitly excluded from the 1964 reform law and did not fear expropriation. 21 Horton maintains that it is likely that by the end of the 1960s, many large landowners "saw land reform not as a threat to their economic interests, but as a way out of a bad situation." (Horton, D., .2p cit., p. 46). 31 The IBRD 1975 Agricultural Sector Survey in its Statistical Appendix provides an analysis of the quality of agricultural related data. - 308 - 335. The low investment in agriculture is also reflected in the little change in yields. Between 1961 and 1971 the increase of value- added in agriculture was about the same as the expansion of agricultural land. 1/ Moreoever, the data on yields of individual crops in Table 97 show substantial annual fluctuations but little sustained growth. 336. Against the low growth of output, the demand for food has grown faster than population, probably by well over 4% per annum. 2/ The result has been a rise in Peru's dependence on imported agricultural products. Wheat has traditionally been Peru's largest food import. Because of stagnant domestic production, the volume of its imports has risen at an annual rate of almost 6% since the early 1960s, and import: dependence increased from around 70% in 1961-63, to almost 85% in 1973-75. l/ Other traditional imports that have increased rapidly recently are dairy products, pulses and vegetable oils. 4/ Moroever, other foods in which Peru was self-sufficient are now imported. The most striking case is maize of which significant quantities were not imported until 1973, but of which some 350,000 tons were imported in 1975 (over one-third of domestic demand). Although Peru's beef output is also insufficient to meet the domestic demand, the Government has re- stricted import volumes. This has been a major factor stimulating the spectacular growth of the poultry industry. However, the foreign exchange savings of the beef import controls have been minimal, since the boom in poultry production has contributed significantly to the growth in maize imports. In value, imports of food and animal feeds rose from around $40 million to $50 million in the 1950s and shot up to $110 million by 1965. 5/ 1/ See Appendix Table 2.16 and IBRD 1975 Agricultural Sector Survey. 21 Data on aggregate food demand are not available. However, with total population growing at almost 3% yearly and urban population at almost 5% yearly, the demand for marketed food must have been growing well above 4% yearly. 31 The table below shows estimated average annual production, imports and domestic consumption of wheat (in thousands of metric tons) for 1961-63 and 1973-75: Fod=Xt.i=D Z=pors a2za 1961-63 153 40O 5e2 1973-75 150 763 912 So0e: AcriculturaL Sector Surver 4/ Vegetable oil imports rose from 9,000 tons p.a. in 1961-63 to 30,000 tons p.a. in 1969-72 and to 68,000 tons p.a. in 1973-74. 5/ Source: Central Bank and 1975 World Bank Agricultural Sector Survey. - 309 - TablcA5: PXDICATORS O0 ACRICULTUIAL PRODUCTION, 1950-75 (in percenta4ea) Index 1969-70 a 100 _ . A"erww *inu1 grov1h rntQ 1930-51 1962-63 l969-70 1974-75 - 1962-63/ 1969-70/ 19T75/7 1950-51 194F2-63 19O6-7O Agriculture Value added (at constant prices) 58 91 Ica 112 3.8 .3 .2 Agriculture gross value of production (at conhtant prices) 3.IL n. 100 119 a.a. .. 2.9 Livestock n.-- n.&. 100 148 ,.. u.s. 6.7 other n.-. 81 100 107 e.&. :3.0 1.1 Total 01P **3 72 100 139 5.4 J.L 6.8 Agriculture G$P/Total GNP 2.3 19 gL5 ' 12 Saorce: Appendix Table 2.16 and Central Bank. TabLe 96: S VALTE AND VOLU? IDICES OF AOETCUL'URAL PRODUCTION BY MAOR PRODUCTS, 1962-72 (percentages) VALtTE / VO TL U M E Fercentage Index Average annual Shares 19691 grcvwth rate 1969 1975 1962 1969 19T5 1962-69 1969-75 TOTAL 100 100 _ - MAJOR ZXPORT CROPS _i 12 Sugar 14 14 100 100 2/ 125 - 3.9 Cotton T7 5 157 100 76 -6.2 -4.5 Coffee 3 3 66 100 101 6.2 0.2 OTHER MAJOR CROPS 33 29 Potatoes 1.1 79 100 91 3.4 -1 .6 Rice 6 6 84 100 1. 5 2.5 2.3 Maize 5 5 79 100 109 3.5 1.5 Bananas 3 2 45 100 106 11.9 1.0 ibeat 1 1 120 100 83 -2.6 -3.1 Alfalfa 6 6 85 100 n.a. 2.4 a.&. Cassava 1 1 87 100 T8 2.1 -3.7 MAJOR LI VETOMC PROWCTS 2! 32 meet PoLtr 14' 10 42 100 298 13.2 20.0 Beef 6 6 81 100 120 3.1- 3.0 Pork 3 4 90 100 1ia 1.5 6.2 Mutton 1 1 102 100 98 -0.3 -0.3 Cov silk 8 59 100 120 7.7 1.1 Eggs 2 3 71 100 212 5.0 13.3 OT_E 29 1/ At 1970 tonstuat prices. 2/.Refers to 1968-70 annuaL average, since 1969 alone vms an wiusun.1ly bad year. Source* I3RD, 1975 Agricultural Sector Survey and Ccntral Bank. - 310 - The sharp rise in imported quantities in the early 1970s coincided with an equally sharp increase in import prices in. 1973-74. 1/ Import value there- fore jumped to $357 million in 1974 and, In spite of the gradual easing of prices since late 1974, has continued to rise to over $400-million a year. These figures do not include imports of fertilizer and other farm inputs which have also gone up considerably. 337. Until recent years the rise in import value had been widely offset by the rise in agricultural exports. However, in 1975-76 against $400 million of imports yearly, exports reached less than $365 million. (Table 98). During the 1950s, volume of agricultural exports more than doubled, which was sufficient to offset a sharp fall in prices / and increase value-added by over 60%. Since then, however, export volump has been falling rapidly and in 1975-76 was 40% less than uhat it had been in 1962-63. The principal cause for this has been the sharp contraction of cotton production, responding to low world prices that prevaiLed in the 1953-70 period and to the low investment since 1970. Another contributing factor was the fast rise in domestic consumption of sugar which has curtailed export volume by over one-fifth since 1962-63. 3/ During the 1'970s, export value has expanded vigorously by almost 15% annually, but only because of very favorable prices._4/ In the long-run, the fact remains that unless food import growth can be stopped and/or agricultural export growth rapidly stimulated - both of which are unlikely -- Peru will have to finance an increasing share of its food con- sumption with foreign exchange earned by other sectors. C. RECENT GOVERNMENT STRATEGY AND POLICIES 338. Important changes have been made in Government policies vis-a-vis agriculture and rural areas. The two basic objectives were to accelerate the growth of food output and to improve the distribution of wealth and the conditions of the rural poor. To achieve, this the Government's strategy had three aspects: a massive agrarian reform, higher public investment in agriculture, and an increased intervenition of the State in all aspects of agribusiness. This section discusses and evaluates the success of the principal policies followed. 1/ The import price (CIF) of wheat roset from $70/ton in 1970 to $190/ton in 1974 and the price of vegetable oils rose from $220/ton in 1967 to $550/ton in 1974 (Source: IBRD 1.977 Agricultural Sector Survey). 2/ Caused by the end of the Korean War and the 1959 world recession. 3/ Between 1960 and 1974 the domestic consumption of sugar rose 5% annually against an average 1.5% annual growt:h of production (Appendix Table 3.12). 4/ Although prices of sugar fell sharpLy in 1976, they remain higher in real terms than what they were ten years before. Moreover, the fall in sugar prices was somewhat offset by increases in coffee and cotton prices (Appendix Table 3.3). - 311 - TAbla 97: AVERAGE YZLD FOR SELECTED CROPS, 1961-T4 In tons / hectare Rnd l96 100 / 1961 1965 19T1 97k 19612 965 1971 197k Rice 4095 3l 4010 4So00 100 95 98 112 f.is '1285 1625 650o 1660 100 126 128 129 bat 1005 960 880 9-40 L00 96 88 94 Potatoes 5780 6245 6150 7T0 100 108 106 123 Causava 9980 10950 13365 12140 100 110 134 . 122 Sugar cue 145255 151105 156870 168500 100 104 10- 116 Cotton 1510 1500 1705 1850 , 100 99 113 123 Coffee 480 520 575 620 2100 106 120 129 Alfalfa 36800 48780 4691o 4646o 100 133 127 126 Source: 1lRD, 1975 Agicturlv Sector Sur"e and Apend1 Table 7.45 Table 9S: nDICATOR OF AICULTURAL XO S, 1950-T6 V A L U E v 0 L U M E In Us$ Elliori' Index 1969-70 d100- Avera.e annual -rovth rate 1950-51 1962-63 1969-70 1975-76 1950-51 1962-63 1969-70 1975-T6 1962163/ 1969-70i 975-76/ 1950-51 1962-63 1969-T0 TOTAL i2 195 162 364 67 142 100 82 6.4 -3.2 cotton T7 94 59 66 S7 169 100 46 5.7 -7.3 -12.2 Sg 32 58 50 186 75 129 104V 101 4.7 -3.95J1 0.2 Coffee 2 25 38 79 4 89 100 95 30.5 1.6 -0.8 Wool 13. 10 6 14 148 170 100 .17 1.1 -7.3 2.T 2the/ n.-. T 9 20 n.a. n.a. n.a.. n.a. n.a. n.a. n.a. NAwrundum item Agricutural exports an a of total exports ,/ 53 *5 17 27 1/ Annual averge. 2/ INot availabLe before 1960. wJ Inalue. 4/ lndex of total valom calculated S exports &t 1963 constat prices. / Refers to 1968-T0 umua. average, since 1969 alone vas an unusually bad year. Source: Appendix Table 3.3 - 312 - a. Land Reform: Achievements and Problems 1/ 339. In June 1969 the Government approved, a new land reform law that has rapidly and dramatically changed Peru's landownership patterns. 2/ Latifundlst agriculture was to be liquidated and the property rights trans- ferred to agricultural workers, small farmers and peasant communities. The target was to expropriate 14,500 estates covering 43% of the country's gross farm area (10.3 million ha) by the end of 1977. 3/ The new agrarian structure was to be based on cooperative production units which would stimulate the growth of agricultural output, raise rural income and promote socioeconomic development. i. Main Characteristics of the Government's Model of Agrarian Structure 340. In adjudicating the expropriated land, the Government created a very complex organization of the agricultural sector. This section provides the mini-mu necessary information to understand how the new agrarian reform enterprises (AREs) operate and how the 434,0013 rural families that benefitted in some way from the reform, were distributed. 341. The country was divided into thirteen Agrarian Zones to which many of the functions of planning and executing the land reform and servicing the farmers were decentralized. In turn each zone was to be divided into PIARs (Integral Rural Settlement Projects) according to internal homogeneity in 1/ This section borrows heavily from the 1977 World Bank Agricultural Sector Survey and from the report prepared for the World Bank and the University of Wisconsin by Horton, D., "Land Reform and Reform Enterprises in Peru", June 1974. Another important reference was Caballero, J.M., "Reforma y Reestructuracion Agraria en etl Peru", Documento de Trabajo No. 34, CISEPA, Universidad Catolica del Peru, Lima, December 1976. 2/ Decree Law 17716 of June 24, 1969. The main changes introduced by the 1969 law were: to make the agroindustrial complexes (notably the sugar plantations) subject to the reforn, to lower the maximum size allowable for private holdings and to cult in half the ceilings on cash conpensations to landowners. In addition, the law created a specialized branch of the judiciary to deal with land reform, prohibited all forms of farm rentals, required owners to reside on their properties and decreed that all agricultural workers participate in the profits and management of the enterprises for which they worke,i (see Horton, D., op. cit., pp. 48-52 for a description of the 1969 land reform law and a comparison with the 1974 law). 3/ See Table 93. By early 1976 almost 80% of the target area had already been expropriated. (See Appendix Tables 7.64 to 7.66). Most of the lands that are still in the expropriation process are in the Sierra, particularly in the Cajamarca and Arequipa areas (see Appendix Table 7.65). - 313 - terms of ecology, physical infrastructure and land tenure problems. 1/ The PIAR was to be the basic operational unit in which the different cooperative enterprises and the medium and small farmers were to be integrated, and through them the Government planned to implement future rural development programs and projects. Within each PIAR, central cooperatives (CCAP) were established. 2/ The organizational superstructure of the agrarian sector would be completed with the confederation of all CCAPs into the National Peasant Organization (CCN). 3/ 342. At the base, the Government theoreticians visualized an agrarian structure organized in large-scale units with cooperative forms of ownership and production. The Government saw various advantages in the large-scale units, including: (i) potential economies of scale in marketing, financing, management and the provision of technical assistance; (ii) more rational use of resources (land, labor, water, etc.); (iii) lower costs for improvement of social conditions by concentrating the resident population in larger centers; and (iv) potential job creation through the division of labor and diversification of activities. Collective ownership and work was favored to break the traditional isolation of the peasantry and establish more 1/ The size of PIARs varies depending on the specific conditions of the region. In general, where irrigated agriculture is concerned, each PIAR may cover 200-500,000 ha if soils are low grade pastureland. With the exception of the 12 largest sugar plantations in the Costa and a few very large Sierra livestock haciendas (expropriated prior to the establishment of the PIARs), all other lands affected by the reform have been expropriated and adjudicated through the PIARs. (See Horton, D., op. cit. pp. 58-59). 2/ Central of Agrarian Cooperatives of Production (CCAP). Each individual cooperative (base level enterprise) by law has to join a CCAP. The CCAPs, owned and operated cooperatively by the base level enterprises had the following objectives: (i) plan production, purchase inputs and market output of the members; (ii) establish processing plants; (iii) plan and execute investments in economic and social infrastructure; (iv) stimu- late the economic diversification of the rural areas; and (v) equalize level of productivity and income between base level cooperatives. (See Horton, D., op. cit., pp. 57-70). To achieve the two latter objectives the CCAPs would receive a share of the profits of the affiliated members, part of which they would use for investment in nonagricultural activities and the rest would be redistributed to the members in accordance with social and economic needs. Another aim of the Government in creating the CCAPs was to reduce the financial and administrative burden of the Ministry of Agriculture in providing assistance to the new agrarian reform enterprises. 3/ This third-level cooperative organization was called the "Central Campe- sina Nacional" (CCN). - 314 - socially-oriented attitudes and behavior. 1/ The idea was not only to transform the estates within the expropriated areas into large-scale cooperatives, but to do the same with the smal1 farmers and minifundistas through parcel integration and concentration. The basic organizational model called for the establishment of two general types of cooperatives; the expropriated lands were to be organized into Agrarian Production Coope- ratives (CAPs); most of the rest of the land, mainly held by small farmers, was to be integrated into Parcel Integration Cooperatives (CIP). However, the adaptation of this basic framework to the regional realities resulted in the proliferation -- which still continues -- of a series of different types of cooperatives, making the new agrarian structure extremely complex. 343. The Agrarian Production Cooperative (CAP) and the "Grupos Campesinos": By the end of the land reform around 940 CAPs will have been established on 39% of the expropriated land with around 7% of the rural families. 2/ The CAPs salient characteristics are: (i) ownership and usu- fruct of all productive assets is collective; (ii) members receive a salary and a share in the profits proportionate to tae number of hours worked; 3/ (iii) management is done by worker representatives; 41 and (iv) labor is provided by the members and by seasonal hiring of non-members. 5/ 344. Most of the CAPs were created on estates operated as wage-labor enterprises (mainly in the Costa) and on some of the ex-haciendas partially operated by tenant workers. For these, the cooperativization process left the systems of production basically unchanged particularly since many of I/ These views were expressed by the then Minister of Agriculture and are reported in Horton, D., op. cit., p. 61. 2./ See Table 99. The area in CAPs is 4 million ha or 17% of the gross farm area. The beneficiaries are 120,OCO families (6.7% of rural families). Table 99 shows CAPs and "grupos campesinos" combined into production cooperatives. See Horton, D., Op. cit. pp. 62-65 for a more detailed description of the organization of the CAP. 3/ Initial membership in the CAPs is determined by the Ministries of Agriculture and Labor on the basis of pa,yroll records. In general, full-time workers and others who have worked on the estate over six months prior to its expropriation are granted membership. However, if the Ministry determines that initial membership is below "carrying capacity" of the CAP, the CAP may be rectuired to expand its member- ship. As discussed later in this Chapter, to avoid expanding member- ship many CAPs have reduced labor requirements through mechanization. 41 In small CAPs (less than 500 members) many decisions are taken by the general assembly of all members; in the larger CAPs worker representatives are elected. 51 However, many of the sugar CAPs employ non-members year-round for cutting and other heavy tasks which members prefer not to do. - 315 - Table 99: o suw u cm N . sM AM A c AuzIa CM) late" AUI135N ZUM. N1V1 AIIM Agnvtal 1.Utma*i? farmrom T $d, a Aw t .1" M so V"ta se. cc Arm.2. 15 Le U. ft. *t PimP* 1. St 1J.C.Ba Jaa. is Ut. u_4. rMm. W tpe ma .. *. t iP -w t-s - - S. _. * _ . A ._. 4 ^ _ -. _. .. ' .W.m " r a" b im lava. imPs -Na. ; j 144 14.5 1.1 f4 £ 2444OJ .a )~~~~~~~~~~~~U- "' Ai. _ xe "A tW MM ;1 9 ? , *s _ e w tr; aim6 b1 es ts (r} sau' A*>oaP i .I/ m.8 t _ 'r_ ,, : J ; 1* l e e << *~~~~~~~~~~~~~~~__ .4 a. 4 6 0& seo *1 * @ Is ,.ae .t .&i S. (.1 1.) 0.1) (6 0.2 . 3.0) a .1 (-1 e - s- ci - ,7 Aq.ePS 3.Uim ?mae'W1rn C.aulj^i, - _ _ - I ¶0S. PA m . - - a... --zz**^t - - - *J4~& *1J~ *~4W * * * e Saw~~~~~~mLp1M4 2~~~~~~~~~~~~ - ~~~~~~~~~~UPM9 A3 . I.& 56 0 * 4 =.Mm : - a- . . - . 3.4 1. 30. f_ 1.). s , 3 twa... ta *u_asaiut. _iSt -, - - - o~~1A4 -tc - - --° ^ Pt 21t.st1 Ia.sdag.r1 s - - - - ^ 7 7 - . SA .5 9 . '. ii z.gw~~~M~o wanlc,u." - - - -92 L "3.0 I.) S-.. 4. how"0- L.. . . ^ . ' '&ItlAX:tt tsreltmt~~~~~~~~~~~~m.4 wsssswg ^ ¶92*62. var, mwskj'.. 11. i ZIJ. AZ1 1. i.0 2 A 37 *S' _*t_e. ! . ' 1s 1 9., 4 * 0.3 _ ~ ~0.0 MO 10 X - 4 S A01 19A3 3 14 IM,96Wmaim ma ifS0 (19.41 14 *hli ._ 1.1S"a - _ ~ i (6'.A) (3.31 sia'a afte Am, WIM y . 20 3 m &. zim miams se (3 a.o (14) ..Z6vw" zoe" o J .0 0. W ti piti PeNeael pnW.US woomi. toAR a I5ea is -so -""o _ommB"m l*. LSJ 7w . f~~~~~~~s et . .s0s to=a U 56 ' r CU-§ CU_ _ _7_ O 46~~ ~ ~ 2k.Iwiiru U&A 0 . M 16Xt _sW "MO _W _= _ t See WV Mggot U"_j Wg e twomlmig "M t "%Jm U". AA _ ws vw awe *Z&" imAti. 1_. AS X tY _ dL. t 1a ( _ _MNA_tr1L . CU P) C "Owl exY_ t1 _me- _.4 *_LOW. _aimmer. tam 40M.,m aws. ft_me W Us, ivkw una to. Tommim ,m o . sd to ( _ 9 (1.01. Date L" Al 4bm -ft MaW 40, &VNV t- 14Ot"7 -" fwT. CU 10 -Wi *ta.cres* S',. h:a" Cmms m"(A *13. If N-? not include expjorts of cauted rich and vool articlec. L/ The C=X on zinc aloy* was elicinated in 1975. S;reet Appendix Table 3.16. c. The Role of Industrial Exports in Peru's Long-Term Development 589. Central to Peru's autarkic development thrust and to the skepticism regarding industrial exports is the disillusionment with the export-led growth of previous centuries that centered around enclave developments absorbing most of the benefits, and with little linkages between the export sectors and the rest of the economy. A related argument is that allowing a high share of industrial output to depend on exports would increase the foreign dependence of the industrial sector and of the economy. 590. In fact, as noted in previous chapters, the reason why the distance between the export sectors and the rest of the economy grew, was precisely because most exports were primary commodities with little domestic value added except extraction and more recently some refining. At the same time, the autarkic industrial strategy in a small economy has done little to reduce Peru's dependence on imports and on the vagaries of primary commodit.Les, and has increased its dependence on the attitude of external lenders. Conversely, an export-led industrial strategy could: (i) reduce the vulnerabilit:y of the economy to foreign arbitrary decisions and the capriciousness of the markets of one or two commodities, by diluting the concentration of exports; and (ii) increase the domestic value added and the labor content of exports as well as the linkages between the export sectors and the rest of the economy. Maintaining a high growth rate of the economy in the long run -- without further increasing the economy's dependence on foreign savings, which seems neither feasible nor desirable -- hinges on the brisk expansion of industrial exports. - 430 - Table 130 KMWACTURED zwRTS, 1960.49T5 (Thousafns of US dollars in prices of 19631 Annual Annual. Year Amouat % Change Yeer Amount S Change 1960 2,904 - 1969 7,146 +70.1 1961 2,377 -18.1 11'0 9,614 +314.5 1962 3,455 +45.4 1'm1 11,235 +16.9 1963 3,364 +11.8 1772 20,915 +86.2 i964 2,942 -23.9 1973 29,087 +39.1 1965 3,422 +16.3 1974 33,557 +15.4 1966 3,342 - 2.3 1975 3/ 21.956 -34.6 1967 2,896 -13.3 1968 4,202 445.1 1 Excludes canned and frozen fish. Preliminary. Source: Central Bank Table 131: MANUFACTURED SXPORTS OF rATTNA AMRcAJ AJD CAXRBBEA COUMMIM, 1972-T4. Per X L n u f ac t u r e d E x n o r t s 17 Capita Total Per capita - 1974 (in million $) (in U (in ) 1972 197r 19T2 1974 Brazil - 920 7*9 1921 7.6 18.5 Mei-co 1090 6147 1428 12.0 24.7 Argentina 1520 394 957 L6.5 39.8 Colombia 500 172 398 t.5 17.2 Venezuela 1960 55 13T 5.0 1.8 Tobag 1700 69 109 65.8 101.9 Tobago Guatemala 580 86 n.e. 17.0 M.A. JM SaiLvdor 410 83 n.-. 22.7 n.a. D-inican Rep. 650 59 n.-. 13.9 n.e Cost Rica 840 55 n.a. 30.2 n.s. Coile 830 44 109 4.4 10.5 Nicaragua 670 34 34 17.6 16.7 Jacm.ica 1190 28 35 14.5 17.4 aiti 170 12 25 2.8 5.5 Ecu-dor 480 7 20 1.1 2.9 Per'7 T40 21 34 1.5 2.3 - S1-C rouPs 5-8 minus 68 Sonree: 1BFD, Recent Prends in Manufacturid acd Total Exports from Developing Counmries, June 197T. niaeo., Annuex A-l; IBPD, 19TT World Atlas; TBRD, World Tables 1976 and Table 2.1. - 431 - 591. Another argument against export-led growth of industry, which is also central to the autarkic thesis is that it would result in a pattern of industrial production inconsistent with the consumption priorities of the country. This argument has three basic points: that export demand will be mainly confined to non-essential goods, that all essential goods should be produced domestically, and that to change the present consumption patterns away from "superfluous goods" it is necessary to change the domestic produc- tion system. 592. This report fully endorses the view that all measures available should be taken to make it possible to provide the poor both with an income that allows them to purchase essential goods and with a supply of these goods at the lowest possible prices. However, forcing the productive apparatus to concentrate on production of essential goods regardless of costs is not likely to achieve these objectives because the inefficiencies that would result from such a policy would almost certainly lower the potential growth of employment and of incomes, and would not produce the essential goods at the lowest possible cost. So, in the end the main losers of this autarkic policy are the poor, who were supposed to be the main beneficiaries. The important aspect of the productive system in a country like Peru is that it maximize efficiency of resource use to generate as much possible income and employment, even if this means producing nonessential consumer goods for export. Welfare (consumption) may be higher if export industries generate incomes and the local population consumes imports than with an autarkic model of development. Domestic consumption of these non-essential goods can be restricted with stiff domestic taxation (such as excise taxes), credit restrictions or even prohibitions. 1/ While in some selected cases strategic industries can be set up to produce at costs somewhat above international prices (ir the noneconomic benefits are considered to be higher than the costs), these should be exceptions that receive specific subsidies, instead of orienting the whole of industrial policies towards these industries. 593. Another objection voiced against export-led industrial growth is that it would depend undesirably on cheap labor. It is clear that orne of the main competitive edges of Peru -- or other developing countries with an abundant supply of underemployed labor -- is that its wages are lower than in the already industrialized countries. However, as was discussed in previous chapters, one of the important causes of poverty is that the total number of adequate jobs has not been growing sufficiently rapidly. The expansion of employment opportunities at a faster rate can be achieved only if the output of the economy expands faster and more labor-intensive sectors and technologies are emphasized. The faster expansion of output can be achieved only through a faster expansion of exports and a greater linkage between exports and the rest of the economy. Only industrial exports can achieve this and also stimulate labor-intensive industries. The fact that 1/ To the extent that the demand for nonessential consumer goods by the upper income groups is inelastic to price changes, stiff domestic taxation can have important income redistributive effects. The after- tax incomes of the upper income groups will be lowered and the tax revenues generated can be used to increase the productive potential of the poor and to provide them with essential goods and services. This, of course, assumes that the revenues generated will be spent on improving the lot of the poor. - 432 - by international standards wages are low i.s, to a certain extent, less impor- tant if the new jobs created by industriaL exports provide a substantial absolute improvement in the incomes of the poor. Of course, a necessary complementary condition for the industriaL export expansion to have a signif- icant impact on employment is that the present biases in policies that increase the costs of labor and reduce the costs oE capital be reduced and eventually eliminated. 594. A fourth issue that worries Peruvian planners is whether the condi- tions necessary for a successful industrial export strategy are compatible with the objectives of more domestic control and better distribution of indus- trial assets. General skepticism for the contributions of foreign investors and fear of the pervasive power of multinational corporations, on the one hand, and the difficulties of reconciling direct foreign investment with the Indus- trial Community and the priority given tc, the Social Property Sector, have been important aspects. However, the experience of various countries shows that foreign investors will adapt to the local conditions provided the boundaries within which foreign investmenLt can operate are well-defined and stable, and provided these boundaries do not make it very difficult or impossi- ble for a foreign investor to obtain an adequate return on his investment. Peru itself has seen a large number of foreign multinational corporations adapt to the Industrial Community, others entering into joint ventures wlth the State, and still others forming production agreements with Social Property enterprises. Much could be done by the Gaovernment to ensure "fair-play" by the multinationals. For example, tax agreements could be signed with other countries to cross-check tax returns and receive technical assistance in taxing foreign corporations. But the presence of foreign corporations in the industrial sector would also differ from that in the mining sector. No one of a series of foreign corporation in the industrial sector could have a relative importance and leverage similar to the one the large mining companies used to have in Peru. Finally, although a relation with multi-national companies may be essential for foreign marketing -- and to a lesser extent for technology -- production in various fields could be led by Peruvian entrepreneurs alone. The industries in the fishing sector (boats and nets) and the social property enterprises in the wool and alpaca industries are cases in point. 595. Moreover, opening up the industrial sector to export markets need not result in further industrial concentration nor does it need to be based only on large corporations. As noted above, the autarkic model combined with the fiscal incentives have stimulated industrial concentration in Peru. In an expanded export market there would be room for encouraging the formation of competing enterprises and this would also reduce the risks of excessive influence of one enterprise on Government decision-making. Clearly, if the Government does not take an aggressive role of promotion and anti-trust control, the results of an opening-up to export markets could result (as has occurred in some countries) in heavier concentration of industry, but this is not an inevitable result. At the same time, small scale and artisanal enterprises could find export outlets -- given the unique nature and attrac- tiveness of Peruvian handicrafts in the high income markets -- either as single enterprises or more likely through pooled marketing channels. In this respect, the promotion of trading companies that can contract from a large number of small producers and supply the large volumes required by export markets may be an important mechanism the Government could encourage. - 433 - 596. Finally, since the mid-1960s a growing number of developing countries have been successful in increasing their exports of manufactured goods to developed countries. This should dispel the skepticism regarding Peru's ability to export manufactured goods in view of the trade barriers and/or competitive advantages of the industrialized countries. Besides the now classical success cases of Taiwan, Hong Kong, and Korea, there are more ex- amples in Latin America in large countries (Brazil and Mexico), in medium- sized countries (Colombia), and in smaller countries (Costa Rica and El Salvador). It is true that becuase of the difficulties in adjusting the the internal production structure, various industrialized countries reverted to tariffs and other trade barriers to slow down some of the imports of manufactured goods from developing countries. However, these are short-term measures; the long-term trends are clear: because of lower wages and progress in reducing transportation costs, the competitiveness and importance of de- veloping countries in the world trade of manufactures will continue to in- crease rapidly. Some Peruvian industries have, in fact, been successful in penetrating foreign markets when the exchange rate was not excessively over- valued and in spite of the generally negative impact of industrial policies on exports. 597. Timing is of the essence in determining the prospects of Peru participating on a large scale in the expanding market for manufactured goods from developing countries. Already most Latin American countries are force- fully promoting aggressive industrial export policies. Unless Peru moves quickly to improve efficiency, strengthen the competitiveness of its industrial structure, and establish itself in the world markets, by the mid-1980s it may find itself outcompeted not so much by the industrialized countries of today but by other countries in Latin America. 598. Identifying specific sectors and products in which Peru could be competitive in world markets goes beyond the scope of this report. In general, it can be said that the comparative advantage lies in industries that make intensive use of Peru's most abundant resources: labor, minerals, sorae agri- cultural products (mainly cotton, wools and hides) and metals. Within this general framework, examples of industries that may be successful include assembly industries, textile products and clothing, and metal or textile-based handicrafts. The latter deserve special mention because of their potential to provide employment and incomes to a large number of urban and rural poor, even if, at first view, their foreign exchange generating potential seems small. Peru has a most valuable expertise in these artisanal industries, and there is a rapidly growing demand for these products in other countries. To date, except for sporadic purchases by tourists, the products and their sophistica- tion remain unknown abroad, and domestic producers are not organized to supply high volumes and are uninformed about foreign market specifications and quality standards. An intermediary (or trading company) could be created to provide credit and specifications to small-scale producers and pool the marketing of their output, without disrupting the present ownership structure of the industry. If the handicraft industries succeed in creating prestige abroad,,they would pave the way for a larger-scale participation of Peruvian industry in the rapidly expanding (and very income-elastic) world markets for high-fashion products. - 434 - 599. As noted before, the Andean Market offers important opportunities for some Peruvian exports, although there are various reasons why an export strategy should be directed principally at broader markets. First, the prospects are dim for reaching agreement in. the foreseeable future on basic programs of the Andean Pact (particularly c.n issues such as the sectoral development programs, which have an important impact on the design of a Peruvian industrial exports strategy). Second, the degree of complementarity in industrial production between Andean Pact countries is limited, and with the withdrawal of Chile the size of the market has diminished by about 15%. Finally, some Pact members (particularly Colombia) have already oriented their industrial sectors towards an aggressive penetration of outside markets. Therefore,. it is in Peru's interest to maxiLmize the benefits it can derive from the Andean Market -- which are potentially substantial -- but not concentrate most of its hopes for industrial exports oii it. d. Export-oriented Policies 600. Success in the export markets by Peruvian industry will principally depend on the improvement in general productive efficiency and competitiveness. As discussed above, this calls mainly for a reduction in protection and subsidies to the import-substitution industries. Of equal importance is to maintain a realistic exchange rate policy that reflects changes in prices in Peru and abroad. 601. Were the system's present bias against exports decreased, the specific export subsidies could also be cu.t sharply and the drain on the Treasury reduced. In the past, the CERTEX tax reimbursement system has apparently provided a significant stimulus to manufactured exports, offsetting at least for some products the disincentives created by other industrial policies and the overvaluation of the Sol. The number of products and producers involved in a sustained fashion is still small, however, and it is not clear that all of these really neecd subsidization. Moreover, the administrative complexity of the export regime represents a significant disincentive, particularly to the smaller producers. In the context of a general review of the country's industria:L policies, a comprehensive study should be made of the impact of the CERTEX system to date, with a view towards restructuring and reducing the level of subsidization. It is particularly important that in determining the level of export subsidization necessary to eliminate the bias against exports, domestic inputs be valued at world market prices, or at shadow prices reflecting their real cost to the national economy. The use of domestic prices with the high levels of effective protection currently in place may well result in encouraging industries, both for export expansion and import substitution, that would cause a net loss of both foreign exchange and real income to Peru. In recent years higher CERTEX subsidies have been provided precisely to sectors, such as textiles, enjoying the highest levels of effective protection. Moreover, consideration should also be given to eliminating the differentiation of CERTEX rates among products. It complicates the administrative process and -- particularly with the present formula valuing domestic inputs at domestic prices -- may reward inefficiency rather than encourage competitiveness. - 435 - 602. Besides adjusting macro-policies and tax incentives, an export strategy will require Government intervention in other areas. The first is the infrastructure required to identify foreign markets and their charac- teristics, disseminate the information in Peru and promote exports aggre- sively abroad. There is lack of information on export possibilities (few enterprises have a systematic export policy and a distribution organization), and the Peruvian industries that do try to penetrate export markets often lack contacts with foreign importers and lack experience in finding the correct commercial channels. There is at present an Export Promotion Office in the Ministry of Commerce and an Information Center of the private Exporters' Association. Both institutions may be the seed for aggressive export promo- tion work. But compared to sister agencies in other Latin American countries, the Peruvian institutions are small. Consideration should be given to joining the private and public sector efforts into a larger, well-financed and well- staffed agency. This agency or others in the Government could also assist exporters in improving standards of quality control necessary for a lasting presence in export markets. Finally, the agency should make a special effort, jointly with other Government institutions in assisting small-scale and artisanal industries with an export potential. These small-scale industries face particular problems for penetrating export markets, but they may be an important link to industrialize the urban and rural poor in Peru. 603. A related issue is the promotion abroad of investment possibilities in Peruvian export-oriented industries. Foreign investors may not only bring technology and capital, but may provide the key to important foreign markets at least until the Peruvian industrial structure and foreign marketinig exper- tise are strengthened. Attracting foreign capital should not compromise domestic objectives -- such as share of local participation -- but wLll call for providing credible assurances as to the basic parameters within which foreign investors can operate. Moreover, the number and complexity of admin- istrative controls should be studied to streamline them drastically while enhancing their effectiveness. To this end, the measures taken in early 1977 to centralize into one institution all the Government relations with foreign enterprises was an important step. 604. Another area that requires Government assistance is export financing. The basic financing mechanism for nontraditional exports exists within the Industrial Bank; it draws from its own resources and since 1968 from the Govern- ment Fund for Non-Traditional Exports (FENT). The Industrial Bank's export financing has grown rapidly since 1971 (Table 133), but two-thirds of that has gone to finance fishing boats (which, as shown on Table 132, account for 7.5% of manufacturing exports). Other sectors have received relatively little export financing. At the same time, exporters have to deal with a rather rigid application of pre- and post-shipment rules and delays in obtain- ing credits. To alleviate such problems, FENT's resource volume shou:Ld be increased, the terms of financing improved, and procedures expedited. - 436 - Table 133:, mmu>tma u curs Anwpos O AITZOIA = 196&-WU - -- - (Millions of sols) Ouw Reno,es , 1/ Ya.r _si rhz Boats Other TUT- T-ti 1968 _1.3 - 4.3 1969 31.9 - 3U.9 19T0 - 116.4 - 11S.4 1971 29.3 133.9 - 163.2 19t2 325.t 172.4 - 198.1 1973 1,509.0 227.5 904.0 2,64o0.5 1972 093.1 2o6.1 50. 1,239.2 AJ 7m for Uotdlticnal iports. -oure: Iadutrilal Bank of Peru 605. Industrial exports would be given a further impetus by allowing the duty-free importing of inputs for export production and by creating duty-free areas for industries exclusively oriented to export markets. Mexico and other countries have been successful in inducing the establishment of export indus- tries with this type of incentive. As long as they generate a net gain in foreign exchange (valued at world prices) and employ a large number of workers, these export-oriented industries leave a positive impact on the domestic economy. 606. Finally, an important measure to stimulate industrial exports, particularly by small-scale producers, is the streamlining of the complex export control procedures. Some type of export licensing may be warranted to ensure that the foreign exchange is surrendered to the Central Bank, but the present procedures and the number of institutions required to authorize them should be drastically simplified and reduced. Registering in one transaction with one institution should be sufficient to achieve the objective. It is more questionable what purpose does requiring a prior enrollment in the Registry of Exporters serve and whether it is necessary to obtain an export license for every export contract. Except in a few selected items of real strategic domestic need, consideration should be given to eliminating the export prohibitions procedures altogether. - 437 - H. CONCLUSIONS 607. Peruvian industry has been growing rapidly for at least three decades. However, this growth has been achieved at a high cost to the rest of the economy, employment generation has been small, and the sector remains plagued by problems of protected inefficiency, monopolistic markets, excessive capital intensity and heavy dependence on foreign exchange and savings gener- ated by other sectors or borrowed from abroad. The Government's own planning efforts in 1968-70 cogently analyzed these defects of the industrialization strategy, but the policies followed since then have probably exacerbated the problems. 608. The Peruvian Government is aware of these problems; the Government has been intensely reviewing its industrial policies for more than a year. The modifications announced thus far have centered on stabilizing labor- management relations and reducing uncertainties regarding the future role of the private sector. The underlying economic strategy, however -- the pursuit of greater independence through import-substitution -- and the mix of policies for achieving it have not changed. 609. A major concern arises from the tendency of Peruvian policy-makers to equate economic independence with autarky. No country of Peru's size and resource endowment can achieve autarky without great economic cost and sacri- fice. Early growth in various sectors will quickly run up against the limits of the domestic market and/or the supply of local resources, and small-scale will hold production costs above international prices. 1/ The improvement of real incomes is thus limited, and the transfer of resources to industry whether through subsidies or higher prices -- represents a drain away from other sectors, limiting the growth of output and employment. Independence thus conceived is, therefore, in conflict with the objective of more rapid economic growth and is likely to prevent a reduction in poverty. Moreover, unless it is willing to accept virtual stagnation, Peru's industrial sector will "depend" on the foreign exchange with which to buy the needed intermediate and capital goods from abroad. Meanwhile, the policies necessary to support an inefficient industrial sector also tend to inhibit the growth of exports from that sector. The result is that as industrialization proceeds in such a framework, and as incomes and employment become more dependent on the level and growth of industrial output, the economy as a whole becomes more, not less, dependent on the vagaries of international commodities markets and on the availability of capital abroad to finance the current account shortfall of the balance of payments. Greater economic independence is a worthwhile objective of industrial policy, but it will not come through the pursuit of industrial autarky, but rather through the diversification of Peru's export base and the expansion of industrial exports. If a more export-oriented industrial strategy had been followed in the past, the growth of output and employment would have probably been higher and at a lower cost to the rest of 1/ The Andean Pact offers, in theory, a partial solution to this problem through its sectoral development programs. Little progress has been made in practice, however, as Peru and other members have had diffi- culty resolving their conflicting desires for greater national autarky and regional economies of scale. - 438 - the economy. In any event, it is doubtfuL that even the growth rates of the past can be maintained in the future, unless important changes are made to the industrial strategy. 610. An export-oriented strategy will call for important changes in the macropolicies of the industrial sector. In the first place, it is crucial to ensure that the exchange rate does not become overvalued. At the same time, the present system of industrial incentives is both unnecessarily costly and unsatisfactory in pursuing the Government's development objectives. The high level of protection assures satisfactory profits at normal levels of aggregate demand, regardless of productive efficiency and regardless of the level of priority accorded their product or the location of their plant. Although the tax and other incentives greatly enhance the profitability of investment, there is little net differentLation among the three priority categories or among locations. While the incentives may have raised the rate of investment overall, available investment data suggest that the differential provided has not been sufficLent to overcome the uncertain- ties of investing in sectors other than where profits were earned. At the same time, the loss of tax revenues to the Treasury has contributed impor- tantly to the Government's budget difficuLties. There is, moreover, reason to believe that the substantial subsidization of capital provided by the incentive system has resulted in the creation of idle capacity in some indus- trial subsectors while generally biasing production processes in favor of capital intensity to the detriment of employment. At the same time, the major tax incentive to reinvestment probably reduces the mobility of capital, favors firms and investors in the higher. tax brackets, and contributes to the concentration-of market power. 611. In sum, the incentive system coLld be simplified, its administrative costs substantially cut, and the incentives themselves reduced, particularly in view of recent Government actions to reduce private investor uncertainties. We specifically recommend: (i) that the profits distribution to labor be taxed as ordinary income; (ii) that tax incentives to investment be sharply reduced in size and converted to a tax credit applying to investment from all sources (not only profit reinvestments); and (iii) that incentives be equalized across industrial subsectors, with exceptions strictly limited to cases of clear social interest. Most tariff exceptions should also be reduced or eliminated in the context of a general reduction of trade barriers. Exceptions might be limited to imported inputs for export products. 612. Peru's import regime consisting of a complex combination of tariffs, exchange controls, absolute import prohibitions, and exceptions to each, results in an extremely high level of protection, one of the highest in Latin America. This strongly biases the economy in favor of production for the local market and represents a serious disincentive to the development of manu- factured exports. Moreover, some of the highest levels of effective protection are provided to the traditional consumer goods industries and so raise and assure profit levels in those industries (at the expense of consumers) as to make largely ineffective the incentives for priority industries, for nontradi- tional exports, and for location outside Lima-Callao. In addition to retarding the growth of exports, the high level of protection raises costs in the short run and dampens the incentive to improve product quality and productive - 439 - efficiency, thereby reducing the real incomes of Peruvians generally. Finally, the preferential treatment accorded capital goods imports reinforces the bias of other incentive policies in the direction of promoting capital-intensive production processes. 613. In addition, a successful industrial export drive will require more Government action to: (a) promote export industries domestically and export products abroad, (b) increase the availability of export financing and reduce the unnecessary administrative procedures for export licensing, and (c) give special attention to the export potential of the labor intensive small-scale industries. 614. Besides orienting towards exports, industrial policies should also allow the small-scale industries to play a larger role in the development process. The net effect of the present policies is a strong bias against the artisanal and small-scale enterprise, yet these offer more hope for an effi- cient reduction of poverty. 615. A shift in the orientation of sector policies as large as t:he one suggested in this report should be implemented gradually in order to minimize the short-term disruptions on the installed industrial capacity caused by more competition, less subsidies and the time needed to develop foreign markets. Nonetheless, the longer it is waited before initiating the change of policies, the more difficult and costly the transition will be. Pari passu with the changes in macro and sector policies outlined before, a strategy of gradual change -- to avoid further industrial concentration and temporary contraction of investment and employment -- will require that the Government action includes: (a) launching a campaign to gain the confidence and support of the industrialists on the necessity of changing the thrust of the sector policies in order to achieve fast industrial growth in the long run; (b) deve:Loping efficient mechanisms to provide services (technical and managerial assistance, export promotion, etc.) and credit, particularly to domestic-based and small- scale industries; (c) applying more forcefully antitrust legislation to avoid the takeover of an excessive number of firms undergoing difficulties by single financial groups; (d) expanding informal training programs to facilitate the shift of labor to the new dynamic industries; and (e) reducing the burden of government administrative requirements on industry. - 440 - TECHNICAL NOTE I: DETERMINAMNS OF INVESTMENT AND SAVINGS 1. Introduction 616. Two types of regression models,, usually found in the literature, were applied to explore the determinants of investment and savings in Peru. The first type of model makes the absolute level of the investment or savings variable to be predicted function of the income and the rate of change in prices which correspond most closely with its, and other variables: S = a + b Y + c (dP/P) + d OTHEHR (a.1) where S = savings Y = income (dP/P) = rate of change of prices OTHER = other variables The second type of model attempts to predict the percentage ratio of the variable to income as a function of per capita real income, the rate of inflation and other variables. (S/Y) 100 - a + b (_/p) + c (dP/P) + d OTHER (a.2) PO where (S/Y) 100 - percentage ratio of saving to income (Y/P) 5 real income POP = population and (dP/P) and OTHER are the same as in equation (a.1). The two models were applied to gross domestic investment (GDI) and gross national saving (GNB). In each case, variations were played on the variable for the rate of change in prices, which was defined as the "expected" rather than the actual rate of inflation. The expected rate of inflation is calculated as an unweighted five or six-rear moving average of the past and present rates of inflation. No other explanatory variables were used except for two alternative dummies to test whetlher there has been a structural change since 1969. - 441 - 2. Gross Domestic Investment 617. The two models described above were applied with relatively good results to the explanation of GDI--Table A.1 contains the data used and Table A.2 shows results from regressions based on equations of the form (a.1). 618. The application of the first type of model was made in nominal terms for two reasons. First, the major interest is explaining why the external resource gap widened, since this is a constraint to the long-term growth of the GDI/GDP ratio, whether measured in nominal or real terms (the two are equal except for differences in the rate of change in prices of investment goods relative to GDP prices). Secondly, previous studies have been tmable to establish a satisfactory relationship of real investment to output in Peru. 1/ 619. The first regression made use of the actual rate of inflation defined as the rate of change in the consumer price index. The coefficient of this naive inflation variable did not perform well, but that of the dummy vrariable to distinguish 1969-75 from the earlier years was highly statisticallyr signifi- cant, and in the expected direction, i.e., indicating that GDP has been on average since 1969 above the level which might have been forecast on the basis of the earlier years. In view of the lack of significance, the naive infla- tion variable was dropped in the second and third regressions where it was tested whether any of the two dummy variables would remain statistically significant. 620. Further tests on a more sophisticated version of the inflation variable were conducted for the whole period for which data were available (1956-75 when the expected rate of inflation was a 6-year moving average, and 1955-75 when it was a 5-year moving average). The statistical significance of the inflation coefficient improved remarkably and was consistently of the 1/ See E. Thorbecke and A. Condos, "Macro Economic Growth and Development Models of the Peruvian Economy" in Adelman and Thorbecke, The Theory and and Design of Economic Development (The Johns Hopkins Press, 1966, pp. 188-189); N. Roemer, The Dynamic Role of Exports in Economic Development (Unpublished dissertation, MIT, 1968, pp. 419-431). Both references are quoted by Edilberto L. Segura, Forecasting Financial Resources and Needs for Peruvian Economic Growth (Unpublished M.A. dissertation, Columbia University, March 1970, pp. 25-26). Segura follows J. Vanek, Estimating Foreign Resources Needs for Economic Development (McGraw-Hill Book Co., 1967, pp. 6-10) in disaggregating data by investment components in the hope of finding a better way of predicting the aggregate investment volume. However, his procedures are disappointing because he used a time variable in the disaggregated regressions, which is risky for predictive purposes, especially when there are structural changes of the magnitude of the most recent years (his equations underestimated substantially the actual growth in investment in 1970-1974). - 442 - predicted (negative) sign; 1/ the best results were obtained when expected inflation was defined in terms of a 6-year average. Finally, equations were also run for a shorter sample period (1959-75) to verify whether exclusion of the earlier period would make any noticeable difference; it- turned out that the estimates were gene;ally consistent with those of the longer period. However, although the R of the regression with the 6-year moving average rate of inflation and the dummy for 1959-75 was also the highest, the "t-ratio" of the inflation coefficient was not statistically significant. 621. Table A.3 compares the actual and predicted values of GDS from six selected regressions in Table A.2. As expected, regression (1), which did not include a dummy variable, underestimated substantially GDS in 1974-75. But even the equations with dummy variables did not pick up well the sharp rise in 1975. Thq predicted values from equation (4)--the optimal regression in terms of the R and the "t-statistics" of the regression coefficients-- tracked reasonably well the path of GDI between 1956 and 1975, except for a few years, among them 1975 which had an error of 6%, or about twice as much as the standard error of the regression. 622. Model (a.2) was also applied orn GDI. As expected, the fittings were much worse than those of model (a.l). Only the best estimated equation is reported below: (GDI/GDP)100 = 41.01329 - 0.00Cl66 (G/ GDP) - 1.25773 (dP/P)e (6.20544) (1.62430) POP (2.39423) R = 0.46077; SEE = 2.46715; DW - 0.83942 Period of estimation: 1960-75. 623. This equation explains about half of the variation in the GDI/GDP ratio; even though no dummy variable was included to capture its exceptional rise since 1968. The equation conforms basically with a priori expectations. The effect of rising real per capita GDP is to lower the proportion of domestic investment to GDP, although the coefficient is not statistically significant at the 95% confidence level. A higher e:pected rate of inflation is also associated with a lower investment-to-GDIP ratio; the expected rate of inflation performed best when defined as a six-year moving average of the actual rate of inflation, thereby adding further consist:ent evidence on the inflationary expectations which may be formed in Peru. 1/ For theoretical justification of the expected (negative) sign of this coefficient, see R.I. McKinnon, Money and Capital in Economic Development (Washington, D.C.: The Brookings Institution, 1973). Table A.1. Peru: Data for Regressions on Gross Domestic Investment Gross Gross Gross domestic Rate of domestic domestic Deflator product change inv-stnent product of CDP per capita of Expected rate of (in (in GD- 100 ('970 Popula. at 1970 consumer change of millions millions GDP 1.0) (in prices price consumer price index of snles) of soles) millions) (GDP/PGDP) index (dP/Pe GDT CDP "CDP POP POP (dP/P) (n=4.h1.0) (n5,h=1.0) 1950 2,721 15,860 17.1563682 n.a. 8.2166 n.a. n.a. n.a. n.a. 1951 4,537 19,929 22.7658187 n.a. 8.3885 n.a. 10.1729400 n.a. n.a. 1952 5,455 21,162 25.7773367 n.a. 8.5640 n.a. 6.9252078 n.a. n.a. 1I53 5,957 22,852 26.0677402 n.a. 8.7432 n.a. 9.1321244 n.a. n.a. 1954 5,192 26,728 19.4253213 n.a. 8.9261 n.a. 5.3808111 n.a. n.a. 1955 6,658 29,610 22.4855647 n.a. 9.1128 n.a. 4.6743007 7.2570768 n.a. 1956 8,617 33,007 26.1065835 n.a. 9.3034 n.a. 5.4878049 6.3200498 6.9621981 1957 10,135 36,114 28.0639087 n.a. 9.4981 n.a. 7.3954437 6.4140970 6.4992821 1953 10,199 40,302 25.3064364 n.a. 9.6968 n.a. 7.9468102 6.1770341 6.6695492 r 1959 8,947 47,353 18.8942622 n.a. 9.8996 n.a. 12.6851444 7.6379008 7.2617192 w 1`60 12,292 56,909 21.5993055 0.399 10.1067 14,112.3287 8.6413326 8.4313072 7.8051394 1961 14,223 63,885 22.2634421 0.4;4 10.3190 14,954.1229 5.9774796 8.5292421 8.022-3359 i962 16,725 73,376 22.7935565 0.435 10.6175 15,887.0223 6.5898045 8.3681143 8.2060025 1963 16,391 80,519 20.3566860 0.460 10.9247 16,022.5273 6.0445387 7.9876600 7.9808517 1964 18,111 96,741 18.7211213 0.518 11.2407 16,614.5069 9.8500000 7.4206311 8.2980500 1965 21,370 114,902 13.5984578 0.584 11.5659 17,011.2140 16.3586709 8.9640987 8.9103044 1966 27,499 136,783 20.1041065 0.650 11.9005 17,682.9028 8.8796745 9.5445377 8.9500280 1967 31,017 156,863 19.7733054 0.730 12.2448 17,548.7408 9.7650356 10.1795839 9.5812874 1968 25,783 185,782 13.8780937 0.863 12.5991 17,086.5080 19.1084053 12.7923573 11.6677208 1959 27,863 208,997 13.3317703 0.932 12.9636 17,298.1046 6.2434735 12.0710520 11.7008766 1970 31,049 240,666 12.9012823 1.000 13.3386 18,042.8231 5.0126739 9.8018526 10.8946556 1971 39,636 264,437 14.9888253 1.045 13.7245 18,437.8127 6.8078818 9.3874940 9.3028574 1972 41,834 294,683 14.1962719 1.100 14.1220 18,969.9502 7.1856840 8.8716237 9.0205257 1973 51,182 355,850 14.3330266 1.261 14.5320 19,418.9836 9.5180723 6.9535571 8.9793651 1974 77,416 448,602 17.2571678 1.491 14.9530 20,121.2626 16.8670438 9.0782712 8.6058049 1975 106,756 573,464 18.6159898 1.830 15.3870 20,365.7832 23.6443100 12.8045984 11.5059443 .~ ~~U . 190195 1950. Sources: CGI and GDP series from Banco Central de Reserva del Peru, Cuentas Nacionales del Peru, 1950-1965; 1950- 1967; and 1960-1973. CDP deflator and population series from IBRD data file. Rate of change of consumer price index and expected rates of change computed fronm average index publishied by Oficina Nacional de Estadfstica y Censos, Inlices de Prerios del Consumidor. (Tlhe first ex)ected rate series is a 5-year moving average of the actual rate of ciiange; the soconld is a 6-year l-oving average Table A.2. Perts: Regreecione on Oroee Domestic Invectment (GDI) GDI . * + b I + c (dP/P) d OTHER Period of Conctant GDP (dP/P) (dP/P)a DWm DUN I (dP/P) w SEE D.W. ertication n-5.h-1.0 n0.4,h1.0 (19o0-15) (1969-75) 1951-75 -16,194.94 0.21257 17,5214.88 -242.32 0.97995 3,424.e0 1.55451 (4.28798) (12.970341) (4.04i84) (0.90037) 1951-15 (1) 1,51*540 o.16151 -0.95866 4,911.84 0.6T026 (1.10577) (23.61326) 1951-75 (2) -14,599.22 0.20025 14,612.39 0.98012 1.63t4B (4.39631) (22.30892) (5 .08a) 1955-75 -5,474. 7 0.216QT -1,613.11 16,616.72 0.g6f7T 4,362.85 1.33223 (1.01954) {12.340720) (2.18243) (3.4T756) 1955-75 -11,135.66 0.2090o -619.56 438.8 O.9759 3,696.24 1.89275 (2.23388) (18.10412) (1.18167) 0.9151 3928 1956-75 (3) 4,66o.41 0.22209 -2,808.11 15,186.32 0.97830 3,652.45 1.5L253 (o.8BU41) (16.76346) (3.86441) (4.45085) 1956-75 (4) -3,913.81 0.21006 -1,402.77 14,325.31 OT98I36 3,474.56 1.91209 (0.62305) (20.36938) (2.17442) (4.85496) 1959-75 -13,281.75 0.19634 11,488.62 0.95830 5,165.45 0.89011 (2.20794.) (14.64899) (2.44083) 1959-75 (5) -16,690.T4 D.20638 15,01T.13 0.97555 3,955-09 1.84211 (4.01011) (1H.11512) (4-41680) 1959-75 (6) 4,821.28 0.22220 -2,828.70 15,812.21 0.97438 4,048.31 1.51091 (0.646o5) (15.1oCf,2) (3.32933) (4.01413) 1959-75 -7,041.00 0.21656 1,1432.02 16,270.56 0.96175 4,946.57 1.29046 (0.99210) (10.3Lo66) (1.50545) (2.95061) 1959-75 -6,435.0o 0.21081 -1,116.52 14,531.149 O.9T743| 3,799.16 1.95128 (0.8012*) (18.157106) (1.47241) (4.47732) 1959-75 -14,356.71 0.20864 -321.96 15,327.64 .99746 14,066.16 1.93266 (2.25634) (16.19189) (0.49564) (4.31309) Notes; Firs.a in pa-Lentheses belov the coefricients are their respective t-statiette; I2 l. tbe adjusted h SE2 to the standa" error of eatimate of the cquetionj D.W. la tht Durbin-Watson atatiatic. Regressions Indicatedl wI A nwusbera (1) through (6) in parenthesee after their respective periods vere ueed to obtain predicted values of CD., an reported is Table A.3. - 445 - TjblIe A.3. Peru: Actu;il ;-'. !'LI.L:ted V;.1-.S . of Gross Di;.:ic T:C ((DI) Actual (1) (2, Predicted Actual ~(1) (2) (3) (4)()() 1950 2,721 n.a. n.a. U.a. n.a. n.a. n.a. 1951 4,537 4,764 4,004 n.a. n.a. n.a. a.a. 1952 5,455 4,963 4,251 n.a. n.a. n.a. n.at. 1953 5,957 5,236 4,589 n.a. U.a. n.a. n.a. 1954 5,192 5,862 5,365 n.a. n.a. n.a. n.a. 1955 6,658 6,328 5,942 n.a. n.a. n.a. n.a. 1956 3,6.7 6,A76 6,623 8,227 7,578 n.a. n.a. 1957 10,135 7,376 7,245 10,227 8,)S n1.n1. . n.a. 1958 10,199 8,055 8,084 10,669 9,521 n.a. n.a. 1959 8,947 9,194 9,495 10,572 10,172 8,100 , l0, C1 1960 12,292 10,737 11,409 11,168 11,417 10,072 11 ,4C, 1961 14,223 11,864 12,806 12,107 12,577 11,512 1?.,1,5 1962 16,725 13,397 14,707 13,700 14,313 13,470 13,725 1963 16,391 14,550 16,137 15,918 16,130 14,945 15,949 1964 18,111 17,170 19,385 18,630 19,092 10,293 18,657 1965 21,370 20,103 23,022 20,944 22,043 22,041 20,'60 1966 27,499 23,637 27,404 25,692 26,559 26,557 25,710 1967 31,017 26,881 31,425 28,379 29,921 30,701 28,386 1968 25,783 31,551 37,216 28,943 33,069 36,669 28,910 1969 27.863 35,301 27,252) 34,006 23,574 26,442 33,974 1970 31,049 40,41t 33,594 27,517 31,357 32,978 27,479 1971 39,636 44,255 38,354 37,266 38,583 37,884 37,264 1972 41,834 49,140 44,410 44,776 45,332 44,126 44,783 1973 51,182 59,019 56,659 58,477 58,239 56,750 58,491 1974 77,416 74,000 75,232 80,125 78,24 75,893 80,157 1975 106,756 94,166 100,236 99,712 100,406 101,662 99,69° Notes: Actual values from Table A.1. Predicted values from regressions indicaited (1) through (6) in Table A.2. 624. Table A.5 shows the results of estimating regressions of the form (a.1). These are not generally as good nor as consistent as those for GDI. The performance of the inflation variable is particularly puzzling. Contrary to what was found in the case of GDI, the current rate of inflation ,defined ft*l AA. Fermi Data for sr mpaole on Gros National 5BrAPo Oross rros prers arora" national national national national lep eule ratest saving product product Dcflator product late at hanSeo of (in (in (a n41110ns of Population per capita chan. OW dtl ter mllions miflLons ef oleo (NP fin in solos at CI? of soles) of molee) oa orf 193) (1963-1.0) millions) of 1963 dofl ert Gila GNP (GNP/towp) (Pa POP {cr/wlPcw W^ oW'- 1950 2,909 15,51T 18.6749695 38,956 0-39986138 8.2166 1,741.13380 D. s. '. ; 1951 4,563 19,746 23.1081777 43,f0 0.45882511 8.3885 5,130.35701 14.7 t1 n.c, o. 1952 4,755 21,021 22.6202369 416,2 5 0.47467540 8.56h0 5,171.06492 3-4.56 n.m. na. 1953 4,871 22,673 21.4837031 45,247 0-5019D9400 8.7432 5,175.10151 5.5656133 n.c. a. 1954i 4,914 26,302 18.602899 .49,613 0.53014331 8.9261 5,558.19451 5.7971786 n.c. n.e. 1955 5,055 28,947 11.4629495 52,065 0.55597810 9.1128 5,713.39215 4.8731720 6.88T3091 n.e. 1956 6,662 32,335 20.5712521 54,433 o.59488602 9.3034 5,851.51665 6.9981020 5.33T7181 6.9o51746 1957 7,380 35,535 20.7602566 541968 o.646467o1 g.9.901 5,T7O.26272 U.6707393 6.3W0bIl 5.093221b 1958 7,349 39,549 18.5820122 56 819 0.69605238 9.6968 5,859.506R2 7.6702042 . 6.8018792 6.5958349 1959 7,925 46,260 17.1314310 58,834. 0.78628001 9.8996 5,943.06841 12.9627695 8.2349974 1-8286943 1950 12,791 55,518 23-0393746 646,175 0.86510323 10.1067 6,349.7h1819 10.0248241 9.2653278 8.5333019 1961 13,998 62,294 22.4708640 69,411 0. 8746582 10.3190 6,726,52389 3.71.08930 8.6138860 8.3445887 1962 15,754 71,7c0 21.9121060 75,836 0.94546i26 10.6175 7,142.54768 5.347na49 7.9193151 8.0695525 1963 1i,225 78,710 18.0726718 78,710 l.00000000 10.9247 7,20.77.450 5.7684798 7.5689703 1.58358426 196'. 18,544 94,994 19.5212329 84,093 1.12956313 U.2407 7,181i.56254 12.9563129 7.5676789 8.4668607 1965 18,133 113,000 16.0469027 88,146 1.28196401 11.5659 7,621.19679 13.4920202 8.2611181 8.5550691 1966 22,575 134,016 16.8450036 93,166 1.43615595 11.9005 7,830.42729 12.1838000 9.9196996 8.9148985 1961 23,011 152,763 15.0632025 94,711 1.61293831 12.24ll8 7,734.79355 12.1532273 11.3107680 10.3169542 1968 241,72 181,242 13.6679136 95,353 1.90074775 12.5991 7,568.23900 17.83T972 13.7258315 12.3996062 1969 29,o66 204,059 11.2439197 99,556 2.0h969063 12.9636 1,679.65689 7.8360150 12.1017119 22.7441954 1910 39,317 237,316 16.5673617 108,590 2.185143144 13.3386 8,141.03h29 6.6225025 11.3278684 11.6885601. 1971 39,551 262,500 15.o670h76 115,690 2.268)9473 13.12h5 8,129.45098 3.8236518 9.6558388 1o.0o7t656 1912 41,435 292,240 1.4.1781.150 122,697 2.38180233 14.1220 8,688.35859 4.9716995 8.2195332 8.87511.89 1973 45,253 352,583 12.83470816 129,928 2.71367988 14.5320 8,940.82026 13.9338834 7.y3r5501o4 9.1719249 1914 50,270 445,943 11.2727411 13e,200 3.22619450 16.9530 9,242.29252 18.9084431 9.6520360 9.31.93659 1975 D.e. 569,757 n.e. 143,728 3.96413361 15.3810 9,3160.87216 22.85051411 12.8976304 11.8517824 Sources. OHS, GNP and real (IMP frco Banco Central de Reserva del Pert, Cuentns Nlecionales del Peru, 1950-1965; 1950-19671 nnd 12 0 3 Updates for 1973-74 were provided by the National Income Accounts section of the Banco Central de Baserva del Peri. Data for 1975 were not available at the time the regressions were performed. 0IS for 1975 is estimated at t3,370 mtilion solee (aee Table 8.3 In Btatistical Appendir). Population data froc IBRD data files. Table A.5. Regressions on Groos National Saying GNs - a + b GNP +. c 1d/P)GP + g OTHE Period ( dP/F'P) ej of Conotant GNP DUN DUN 1 (d'P/P)uur F, SEE DW eetimation (n5.h-1)_ 1951-74 (.) 6 196.55 0.12405 -272.798 0.96414 2,Tho0.oo 0.89926 (C.80832) (23.y81914) (2.02903) 1951-74 6,743.60 0.12180 -602.822 -252.331 0.96240 2,805.61 o.8EP124 (1.95362) (8.5836T) (0.17153) (1.38504) 1951-T4 7,094.84 O.12081 -933.219 -246.697 0.96250 2,801.92 0.80292 (2.087o6) (9.81702) (0.28666) (1.49600) 1951-74 4 188.41, 0.12010 o.95906 2,921.68 0.17T36 (1.74908) (23.23215) 1951-74 8,516.61 0.17OTB -3,790.99 0.96075 2,866.30 0.77044 (2.63323) (10.60612) (1.39T20) 1951-14 8,421.05 0.10922 -3,622.36 0.96029 2,883.33 o.56704 (2.49352) (U.13126) (1.29692) 1951-74 (2) 4;188.45 0.12010 0.95906 2,927.68 o.11T36 (14.74190) (23.23215) 1956-74 -241.798 0.10831 759.235 0.96002 2,788.32 1.15620 (0.07074) (16.61L504) (1.80075) 1956-74 6,223.40 0.09271 664.961 -4,T43.79 0.96543 2,592.80 1.32T44 (1.32588) (8.9g9840) (1.68226) (1.87193) - 448 - as the rate of change of the GNP deflator) had the expected sign and was statistically significant, while the expected rate of inflation (defined as a six-year moving average of the rate of change of the GNP deflator) produced the opposite (positive) sign that was expected. Moreover, no form of the dummy variable was found to be statistically significant. These disappointing results lead to the conclusion that GNP alone may be the best predictor of GNS. The predicted values from the regression based on this hypothesis and an alternative regression with the current rate of inflation are reported in Table A.6. Both equations overestimate GNS in 1974, suggesting that they would have overestimated 1975 even more because there was an absolute decline in GNS last year. 625. Regressions based on model (a.2) performed in a more consistent manner and displayed invariably the expected signs of the coefficients of the explanatory variables (Table A.7). The regressions suggest that there has been a long-run decline in the proportion of GNS to GNP as real GNP per capita has risen. They also show a likely negative effect from inflation, but the coefficients of this variable were not statistically significant and the expected rate of inflation performed worse than the actual rate of inflation. Use of the dummy variables gave only slLght indication that there might have been a deterioration in the GNS/GNP ratLo after 1969 or after 1970; the coefficients of the dummy variables werea not statistically significant but were all of the expected sign. If the :regressions had covered the 1975 year, it is likely that the significance of the dummy variables would have increased because of the decline in GNS in that year. - 449 - Table A.6. Peru: Actual and Predicted Values of GNS Actual Predicted Actual (1) ~~~~~(2) 1951 4,563 4,623 6,560 1952 4,755 7,'862 6,713 1953 4,871 7,491 6,9.L1 1954 4,914 7,878 7,347 2955 5,055 8,458 7,6655 1956 6,662 8,305 8,078 1957 7,380 8,239 8,456 1958 7,349 9,010 8,938 1959 7,925 8,399 9,744 1960 12,791 10,349 10,856 1961 13,998 12,904 11,670 1962 15,754 13,632 12,799 1963 14,225 14,387 13,641 1964 18,544 14,446 15,597 1965 189133 16,534 17,759 1966 22,575 19,498 20,283 1967 23,011 21,832 22,535 1968 24,772 23,812 25,955 1969 29,066 29,373 28,695 1970 39,317 33,830 32,689 1971 '39,551 37,718 35,714 1972 ''"41,435 41,094 39,285 1973 a 45,253 46,135 46,532 1974 50,270 56,359 57,745 Notes: Actual values from Table A.4. Predicted values from tegressions indicated (1) and (2) in Table A.5, respectively.. Table A.T. Peru% Regressions on the Ratio of Orona National Saving to Oross National Product (GNS/ONP) 100 a + b(CP/pflOR) I c (dP/P)C1p + g OnTHE OPo Period GNP/P%;p (dP/P),ep (dP/P) (dP/P) -2 Of Consatnt FoP IGhp DUN DIM I GNP Hj SEE IDW eatimnation Ont5,b1.0) (n-4,h 1.0) 1956-197T4 35.93071 -0.00207 -0.37875 0.56335 2.26913 1.24504;. (9.55W7) (3.35639) (1.05596) 1956-979 j53.53636 -0.0017T ° .1.2_ 2 o.53962 2.32996 1.29889 (4.Vf,C201 (1.(354118) (i.106163) (0.41878) 1956-1974. 32.13648 -0.00171 -0.34231 1.19064 o.54877 2.30669 1.291t69 (4.8217) (2.10604) (0.92926) (o.69509) 1955-1974 33.0435s. o0.00194 -0.27252 0.49584 2.37343 1.13647 (9.18521) (3.48198) (0.94771) 1955-1,fl14 28.16748 -0.00132 -0.29234 1.9O646 0.50252 2.35766 1.23162 (14.41444) (1.68242) (1.02150) (1.10822) 1951-1974 33.17169 -0.00199 -0.1.4774 o0.61702 2.17080 1.21101 (13.185'15) (5.283fo) (1.1.U98) 1951-1974 26.85838 -0.00128 2.51.595 -0.20754 o.64795 2.08131 1.40600 (6.03156) (2.30093) (1.6U663) (1.95056) - 451 - TECHNICAL NOTE II: INFLATION, REAL INTEREST RATES AN]D THE DEMAND FOR MONEY IN PERU 1/ 1. General Model of the Demand for Real Monetary Balances and the Effect of Inflation 626. A general model to analyze the relationship between monetary aggregates and prices is provided by a demand for money function of the Cagan type. It is postulated that the desire to hold monetary assets depends posi- tively on real income, as in the simple quantity theory, and on the interest rate on these assets as well, and negatively on inflation. While the interest rate is assumed to be known to asset holders in advance, the rate of inflation that will prevail over the relevant horizon for decision making is variable and can only be guessed. It is then assumed that inflationary expectations are formed on the basis of past observed inflation rates over a finite number of periods (n). Each period is assigned a weight according to a weighting scheme determined by a geometrically declining series with a parameter h as the base (O < h < 1) and integer exponents which increase as the period moves away into the past (i = 0, 1, ...n -1), and all weights are normalized so that they add up to one. The parameters n and h are chosen empirically so as to maximize. 2. The Demand for Money Function 627. The first type of demand for money function used here is of the Cagan type: 2/ In (M/P) - a + b ln Y + c (dP/P) + u (b.1) where M money stock P - consumer price index Y = real gross national product (dP/) expected rate of change of P u = disturbance term 1/ For an earlier study based on the same function, see J.R. Patron Basurco, Mario Dejo Soto, Reynaldo Susano Lucero and Jose Rodriguez Portocarrero, "La Demanda de Dinero en el Peru," Simposio sobre el mercado de capitales en el Peru (Comision Nacional de Valores, Lima, 1972). 2/ See Philip Cagan, "The Monetary Dynamics of Hyperinflation" in Studies in the Quantity Theory of Money, edited by Milton Friedman (Chicago: University of Chicago Press, 1956). - 452 - 628. In equation (b.1), the variable (dP/P) is not observable. However, we may specify an expectation generation function based on-the current and lagged values of the inflation rate with coefficients declining geometrically. If the number of periods relevant to the determination of the expected rate of inflation is infinite, the series of coefficients might be determined by the following expression i Ci = (1 - h) h , i = 0 ... , 0 < h < 1 so that imn c =1 n i=O 629. In practical applications, the choice of infinite periods creates serious problems of estimation. Although there are techniques to reduce the regression equation to an estimable form such as those of Koyck and Nerlove, the reduced form so obtained includes invariably lagged values of the dependent variable, thereby making consistent estimaation very difficult in the presence of other econometric complications. To avoid this problem, the method followed by a large number of studies has consisted of specifying the expected rate of inflation in terms of a finite number of lags. The method is to choose the combination of values of n (nymber of lags) and h (weighting parameter) so as to maximize the adjusted R. of the regression. 1/ The formula for the expected rate of inflation is: n-1i £E hi (dP/P)t i (dP/p) i0 (b.2) n-I i , O h 1/ In this case, this is equivalent to maximizing the t-statistic of the coefficient of the expected rate of change in prices. - 453 - 630. This formula ensures that the sum of the weights adds up to one. The determination of the best values of n and h (i.e., those that maximize the R and t-statistics) is an empirical question. For h = 1 all the values of the actual and past rates of inflation that enter into-the determination of the expected rate of inflation are given equal weights; for values of h approaching zero, the weights decline very rapidly and become insignificant for earlier periods. 631. A straightforward generalization of Cagan's approach may consist of substituting the expected real rate of interest for the rate of inflation in the demand for money function. The formula used to compute the expected real rate of interest is as follows: Re = rt - (dP/P)e (b.3) R __ t b3 t - _ _ 1 + (dP/P)et where e R= expected real rate of interest in period t r effective nominal rate of interest in period t )e = epce (dP/P) expected rate of inflation in period t defined as in equation (b.2). If the actual rate of inflation is substituted for the expected rate in equation (b.3), the equation may be used to compute the actual real rate of interest. 1/ 1/ This formula has been used by the staff of the Asesoria de la Gerencia General del Banco Central de Reserva del Peru to compute series on the real rates of interest on a variety of financial savings instruments for the period 1970-75. That this is the correct way of computing t:he real rate of interest, rather than the usual formula (r - (dP/P) ), may be shown as follows. Assume that a capital vt with a real value oi V = vt/P at time t is invested at an effective nominal yield of rt. At the eni of the period, the nominal value of the capital is vt+1 = vt(1 + rt) and its real value is V = t + r -t Therefore R = vt+1 - Vt = rt (dP/P)t t+1 t ( t V 1 + d Pt(1 + (dP/P)t) * 1 + 2'dP/P)t Table B.I. Peru: Data for Demand for Money Regressions GNP in /2. .pi .1 _7 .7/ Billions of Yi /P M2/ P 4 constant soles (1950-1958) (1963 - 100) (In billions of constant 1963 soles) (1) (2) (3) (4) (5) (6) (6) 1959 58.834 7.568,5' 1l.4615/ 124794WI 13.208=, .7684 .3932 (1950) 1960 64.175 7.431k-' 11.493-' 12.669- 13.079- .8348 .4332 (1951) 1961 69.411 8.07 4k- 12.690/-' 13 887A/ 14.320 -' .8847 .4632 (1952) 1962 75.836 8.411-n 13.48 8& 14.890 -- 15.32 .9430 .5055 (1953) 1963 78.710 8.712 / 14.2876/ 15.96 6/ 16.4626/ 1.0000 .5327 (1954) 1964 84.098 9.340 15.271 17.274 17.852 1,0985 .5576 (1955) 1965 88.146 9.333 15.334 17.423 18.405 1.2782 .5882 (1956) 1966 93.186 10.011 16.616 19.189 20.162 1,3917 .6317 (1957) 1967 94.711 9.853 15.994 18.759 20.065 1.5276 .6819 (1958) 1968 95.353 9.747 15.101 17.469 18.856 1.8195 1969 99.556 10.311 15.715 17.767 19.415 1.9331 1970 103.590 13.845 20.155 21.019 23.606 2.0300 1971 115.690 15.428 22.005 22.126 26.450 2.1682 1972 122.697 16.471 23.435 23.577 28.654 2.3240 1973 129.928 18.014 25.432 25.583 31.510 2.5452 1974 138.200 19.141 25.931 26.121 32.406 .2.9745 1975 143.728* 19.207 25.441 25.632 31.757 3.6778 ( - ~~~~~~~~~~~~(continued) Sources: All data were prepared for this exercise by the Asesorfa de la Gerencia General of the Banco Central de Reserva del Per6t. Their specific sources. were: (1) Banco Central de Reserva del PerG, Cuentas 1Naclonales del Peru; (2), (3), (4) and (5, Banco Central de Reserva del Peru, Desarrollo Monetario Crediticio Anexo EstadIstico; W6 Oi:Lzi-c?a Nac'conal de Estadcstica y Censos, Indice de Precios del i8au-niTdor 1Y/i3 - ioU - W%Tffe_nc-y annd-emand deposits held by the public in real terms (i.e., divided by the consumer price index) - monEhly average except wherever noted. a 2J/ 1I plus savings and time deposits in real terms - monthly average except wherever noted. 3/ M2 plus deposits in foreign currency in real terms - monthly average except wnerever noted. 4J Total liquidity of the puilce in r-al terLs - monrhly average except wherever noted. '/ Year-end observation. 6/ Average of tuo observatiors at the beginning and end of year. 7/ Consumer price index - average of monthly Observations 1963 1.00. - 455 - 632. As in the Cagan equation, the method of estimation may conslst of choosing the best combination of n and h values which maximizes the R and the t-statistics of the coefficient of the expected real-rate of interest. Another alternative, which has been used here, is to use the series on the expected real rate of inflation obtained from the previous set of regressions in order to construct a series on the expected real rate of interest and then re-run the regression with the new variable. This second method has the advantage that it avoids repeating the tedious exercise of scanning for the best values of n and h, but it may not lead to the optimal values of these parameters. The reason for our use of this second approach, as explained below, is that the almost nil variation in the effective nominal interest yields during the period covered by the regression did not justify the effort. The Data and Special Characteristics of the Period 633. Table B.1 presents all the data used in the estimation of equation (b.1) and subject to (b.2) or (b.2) and (b.3) for the period 1959-1975. (Back data for the price variable, which is represented by the consumer price index, was necessary to construct the expected rate of inflation.) Four definitions of money were used in the regressions: money as the sum of currency and demand deposits held by the public (M1); M1 plus savings and time deposits (M2); M2 plus foreign currency deposits; and M4, the broadest defini- tion of money including all other quasi-monetary liabilities of the banking system such as certificates of deposit and other special deposits. This latter definition is often used under the name of "total liquidity of the economy" by the Central Bank. Money and price data are averages of monthly observations, except for monetary data for the early years; for the first year of the regressions only the December figures were available, thereby tending to overestimate the true yearly average because of the normal growth in the series. (This has caused a significant error in the estimates for that year which is well described by the equations, as is shown later.) 634. Data on interest rates corresponds to the nominal and real effective yields on the four monetary aggregates used in the regressions. Aliost all the variation in the data on nominal effective yields is due to shifts in the composition of the aggregates because there was no change in interest rate policy throughout the whole period, except for the elimination of the 1/2 percent paid on demand deposits until 1964. The real effective yields are constructed from substitution of the actual rate of inflation in equation (b.3). Although they are not used in the regression, they are reported for comparison with the estimated expected real yields. Unfortunately, the lack of variation in the interest rate structure makes very difficult the estima- tion of the effect of changes in nominal yields as distinct from inflation. 635. The period of estimation was characterized by similar condiitions in the application of monetary policy instruments except for the capital repatriation measures taken in 1970 which caused a large rise in monetary balances, particularly in M . For this reason it was considered appropriate to test for the hypothesis that these measures had caused a significant alter- ation in the demand for money after 1970 and to correct for the shii-t if the statistical test showed that it had been significant. The Chow tesi: was used - 456 - to compare the whole sample period with the subperiod 1959-69 at the 5 and 10 percent significance levels. The Chow test is provided by the following statistics: 1/ £- (Yi -Y_i + (Y2 _ 2)J (n -k) n 2 m (y 1 l)iL where F is distributed as an F-distribution with (n - k) and m degrees of freedom and Y1 - actual values of dependent variable in period 1959-69 Y2 - actual values of dependent variable in period 1970-75 Y, = estimated values in peri.od 1959-69 from regression for 1 the period 1959-69 'Y1 = estimated values in period 1959-69 from regression for the period 1959-75 Y = estimated values in peri.od 1970-75 from regression for the period 1959-75 n - number of total observations in regression for 1959-69 (11) k - number of variables (3) m - number of additional oboervations in regression for 1959-75 (6) 1/ See J. Johnston, Econometric Methotds (New York: McGraw-Hill, 1963), p. 138, based on Gregory C. Chow, "Tests of Equality Between Sets of Coefficients in Two Linear Regressions," Econometrica, Vol. 28, No. 3 (July 1960), pp. 591-605. - 457 - Straightforward Estimates for the Period 1959-1975 636. Equation (b.1) subject to specification (b.2) was applied to2the four definitions of money for the period 1959-75. The values of the R and t-statistics for the variable (dP/p)e for different pairs of values of the parameters n and h are reported in Tables B.2 and B.3. It was not necessary to run equations for all pairs of values of h and n because we have good a priori information on the plausible range of these parameters from earlier studies. 1/ Table B.4 shows the best estimated equations for the period 1959-75. 637. The equations demonstrate clearly the high correlation which exists in Peru between real money, however defined, and real income, and also show that a higher expected rate of inflation is associated with a lower demand for real balances. As expected, the effect of expected inflation is somewhat less discernible when a narrow definition of money is used because this money is more associated with direct transactions. 638. It was found that the pair of values n = 4 or 5 and h = 1.0 produced the best estimated equations. This suggests that the period of formation of price expectations is of about 4 or 5 years and that people attribute equal weight to all these years. It follows therefore that our series for the expected inflation rate consist simply of 5 or 6-year moving averages (includ- ing the current year) of the rate of inflation. 2/ This useful finding has been exploited to test whether the expected rate of inflation had any signif- icant effect on savings aggregates. 639. Although the estimates are consistently robust, the DurbinL-Watson statistics show signs of significant autocorrelation of the residuals, except for the M definition which is within the range of doubt. This suggests the the possibility that significant variables have been omitted from the regres- sions and is in accordance with our a priori knowledge that the capital repatriation measures of 1970 caused a marked shift in monetary asset holdings. It is therefore necessary to proceed to test for this hypothesis and to correct the estimated equation as appropriate. Test for Structural Change in 1970 640. To test the hypothesis that there was a structural change in real asset holdings after 1970 as a result of the capital repatriation measures 1/ For Peru, see Jorge Rizo Patron Basurco, et.al., which found (n - 5, h - 0.9) as the optimal pair of values. 2/ This is easily shown by substituting h = 1.0 in equation (b.2). - 458 - -2 ~~~2) Table B.2. Peru: R (Adjusted R ) for the tfmand for lIoney Function: In ( /P)>= a + b ln Y + c (dP/P)e (1959-1975) 3 4 5 6 7 8 Currency znd demand deposits (M1) 0.65 0.91541 0.91785 0.91833 0.91748 0.91691 0.91664 0.7 0.91671 0.92018 0.92106 0.91971 0.91867 0.91816 0.75 0.91806 0.92278 0.92431 0.92224 0.92044 0.91953 0.8 0.91940 0.92559 0.S2801 0.92497 0.92203 0.92048 0.85 0.92071 0.92847 0.9S3194 0.92765 0.92314 0.92064 0.9 0.92193 0.93123 0. 93558 0.92967 0.92331 0.91947 0.95 0.92304 0.93365 0.93816 0.92998 0.92191 0.91641 1.0 0.92400 0.93552 13889 0.92753 0.91849 0.91155 Currency and demancL, saving and time deposits (M2) 0.65 0.96058 0.96388 0.96425 0.96330 0.96266 0.96237 0.7 0.96221 0.96674 0.96745 0.96600 0.96488 0.96436 0.75 0.96382 0.96979 0.98723 0.96890 0.96705 0.96615 0.8 0.96537 0.97291 0.97472 0.97180 0.96893 0.96745 0.85 0.96680 0.97589 0.97820 0.97427 0.97014 0.96783 0.9 0.96805 0.97850 0.98076 0.98920 0.970G9 0.96661 0.95 0.96910 0.98049 0.98152 0.97406 0.96800 0.96293 1.0 0.96991 I0.981671 0.97986 0.96913 0.96322 0.95630 Currency and all deposits, including foreign currencv deposits (M3) 0.65 0.97525 0.97690 0.97683 0.97641 0.97621 0.97604 0.7 0.97596 0.97827 0.97824 0.97759 0.97723 0.97691 0.75 0.97669 0.97982 0.97985 0.97889 0.97828 0.97771 0.8 0.97743 0.98149 0.98161 0.98024 0.97930 0.97833 0.85 0.97815 0.98322 0.98336 0.98149 0.98013 0.97856 0.9 0.97883 0.98490 0.98482 0.98227 0.i8054 0.97810 0.95 0.97946 0.98639 0.98557 0.98202 0.98017 0.97658 1.0 0.98000 | 0.987_5g 0.98529 0.98027 0.97865 0.97384 Total liqu:L_-iy of the public (i4) 0.65 0.97302 0.97460 0.97478 0.97426 0.97395 0.97372 0.7 0.97390 0.97614 0.97653 0.97569 0.97513 0.97469 0.75 0.97482 0.97789 0.97860 0.97732 0.97635 0.97555 0.8 0.97575 0.97980 0.98095 0.97910 0.97750 0.97613 0.85 0.97668 0.98178 0.98345 0.98084 0.97841 0.97617 0.9 0.97757 0.98370 0.98576 0.98216 0.97878 0.97530 0.95 0.97840 0.955-3 0.8737 0.98239 0.97£20 0.97321 1.0 0.99083 0.93681 LL..8779 0.98084 0.97629 0.97003 Note;: n is tble tr'hcE, laO :, vti' tD conn'r-lct tL, ½ e:.-pcct.: rate of chan-c' in pri.ces; h is th e c cicffe 1:fnt hic C''1 mVtdC' the -r-=i4ght Ls sL of the laggcd the. of the o rz;tc o. ch.nt in priczs. Sec equ:t:.nn ( ) in the text for £urrlc-r cl-:r r -.;on. - 459 - Table B.3. Peru: t-statistics for the Expected Rate of Change of Prices in Lltc Demand for Money Function: ln(i i/P) ='a + b ln Y + c(dP/P)e (1959-1975) nX 3 4 5 6 7 8 h Currency and demand deposits (Mi) 0.65 1.26733. 1.43890 1.47103 1.41357 1.37390 1.35526 0.7 1.36043 1.59345 1.65053 1.56275 1.49375 1.45976 0.75 1.45270 1.75977 1.85497 1.72520 1.60998 1.55073 0.8 1.54237 1.93436 2.08267 1.89624 1.71213 1.61307 ,0.85 1.62745 2.11081 2.32287 2.06099 1.78236 1.62324 0.9 1.70591 2.27947 2.54814 2.18406 1.79254 1.54667 0.95 1.77583 2.42819 2.71067 2.20294 1.70460 1.33903 1.0 1.83565 2.54462 12.75734 2.05340 1.48213 0.96219 Currency and demand, saving and time deposits (M) 0.65 2.54743 v.89206 2.93148 2.82975 2.76228 2.73244 0.7 2.71584 3>20730 3.28855 3.12375 3.00016 2.94367 0.75 2.88608 3.56933 3.72189 3.46002 3.24246 3.14061 0.8 3.05383 3.97674 4.23682 3.82697 3.46350 3.28933 0.85 3.21369 4.41660 4.80171 4.17060 3.61237 3.33305 0.9 3.35947 4.85590 5.28956 4.34747 3.60648 3.19228 0.95 3.48488 5.23606 5.45165 4.14099 3.35370 2.79103 1.0 3.58438. 15.48327 5.11069 3.48852 2.82186 2.11753 Currency and all deposits, including foreign currency deposits(M3) 0.65 1.90903 2.21589 2.20272 2.12570 2.08803 2.05666 0.7 2.04119 2.47083 2.46353 2.34277 2.27575 2.21687 0.75 2.17662 2.76466 2.77033 2.58601 2.47199 2.36601 0.8 2.31291 3.09888 3.12265 2.84760 2.66402 2.48115 0.85 2.44704 3.46967 3.50070 3.09773 2.82508 2.52405 0.9 2.57550 3.86365 3.84292 3.26083 2.90584 2.43792 0.95 2.69463 4.25331 4.03399 3.20938 2.83336 2.15647 1.0 2.80092 14.596411 3.96147 2.85247 2.541.20 1.641J.1 Total liquidity of the public (M4) 0.65 1.79088 2.06717 2.09981 2.00907 1.95483 1.91469 0.7 1.94582 2.33615 2.40348 2.25801 2.15974 2.08360 0.75 2.10580 2.64511 2.77274 2.54375 2.37180 2.23323 0.8 2.26839 2.99642 3.22093 2.86458 2.57553 2.33412 0.85 2.43044 3.38754 3.74893 3.19769 0.98402 2.340038 0.9 2.58808 3.80641 4.31278 3.46678 2.80695 2.18959 0.95 2.73700 4.22615 4.77126 3.51516 2.70138 1.82309 1.0 1.87283 4.60294 L°_92031 3.19845 2.36227 1.21963 Note: n is the number of lags uszed to construct the excpectod rate of C}rŽ in prices; h is the coefficIent which provides the weighting systen of t'Ž lagged valuer of the actual r'te of change in p-;ices. See equatiol ( ) in the text for further clarification. - 460 Table B.4. Peru: Demand for Money Reg"essions In (Ifi/P) a + b In Y + c (dP/P)e (1959-1975) (n = 5 , h = 1.0) In (M = - 3.327325 -+ 1.362643 ln Y - 0.049604 (dP/P) e (9.095136) (14.269066) (2.757340) R = 0.938894 ; SEE = C0.082814 ; DW = 0.619437 (n = 4 , h =1.0) in (JP) = - 1.86725; + l.C98295 In Y - 0.031660 (dP/P)e (11.552045) (28. 706502) (5.483271) 2 ~ 0.981671 ; SEE 0.037480 ; DW = 1.185496 (n 4 , h = 1.0) 3-ii f'3rI) - 1.101354 0.921272 In Y - 0.018543 (d/P) e (9.755946) (34.462155) (4.596418) -i£ 0.937568 ; SE= 0.026188; DW = 1.642739 ln (MI,/P) - 2.30t04,2 + 1.240517 In Y1 - 0.035780 (dP/P)e (15J.986'42) (32.OQl(5-39) (4.902033) 0.- 77rC( ; SEE - 0_1300 ; D, 1.388278 -461- Table B.5. Peru; V (Ad justcd R )- for the Dema-nd for M4oney Function: ln(M iIP) = a + b in y + c( dP/p)e (1959-1969) 3 4 5 6 78 Currency and demand deposits (M 1) 0.65 0.95573 0.95576 0.95566 0.95557 0.95554 0.95552 0.7 0.95588 0.95594 0.95577 0.95561 0.95555 0.95552 0.75 0.95609 0.95618 0.95591 0.95565 0.95555 0.95549 .0.8 0.95635 0.95618 0.95609 0.95567 0.95551 0.95545 0.85 0.95667 . 0.95684 0.95628 0.95566 0.95546 0.95541 0.9 0.95704 0.95724 0.95648 0.95561 0.95541 0.95544 0.95 0.95746 0.95766 0.95664 0.95553 0.9554.4 0.95570 1.0 0.95792 10.958051 0.95674 0.95544 0.95574 0.95637 Currency and demand, saving and time deposits (M2 0.65 0.92358 0.93044 0.93027 0.92759 0.92629) 0.92590 0.7 0.92692 0.93654 0.93669 0.93236 0.93006 0.92927 0.75 0.93037 0.94329 0.94410 0.93745 0.93362 0.93214 0.8 0.93383 0.95040 0.95232 0.94262 0.93658 0.93401 0.85 0.93720 0.95747 0.96088 0.94748 0.93844 0.93430 0.9 0.94038 0.96403 0.96895 0.95138 0.93855 0.93246 0.95 -0.94331 0.96959 0.97536 0.95342 0.93613 0.92810 1.0 0.94592 0.97382 _0.978901 0.95250 0.93049 0.92138 Currency and all deposits, including foreign currency deposits(M) 0.65 0.90394 0.90872 0.90881 0.90716 0.90625; 0.90585 0.7 0.90584 0.91305 0.91347 0.91056 0.9088] 0.90801 0.75 0.90794 0.91834 0.91944 0.91460 0.91151. 0.90994 0.8 0.91021 0.92450 0.92684 0.91920 0.91402 0.91121 0.85 0.91258 0.93132 0.93557 .0.92417 0.91593 0.91133 0.9 0.91499 0.93843 0.94514 0.920910 0.91674 0.90988 0.95 0.91736 0.94539 0.95463 0.93330 0.91580 0.90673 1.0 0.91964 0.95175 10.962781 0.93576 0.91262 0.90246 Total. liquidity of the public (H4) 0.65 0.93673 0.93752 0.93753 0.93719 0.93701 0.93693 0.7 0.93700 0.93855 0.93863 0.93787 0.93746 0.93725 0.75 0.93740 0.94009 0.94038 0.93885' 0.93801 0.93756 0.8 0.93795 0.94221 0.94293 0.94017 ,0.93857 0.93773 0.85 0.93864 0.94491 0.94641 0.94178 0.93902 0.93762 0.9 0.93944 0.94808 0.95078 0.943-59 0.93918 0.93723 0.95 0.94035 0.95156 0.95578 0.94537 0.93887 0.93669 1.0 0.94133 .0.95514 ~o9c~ 0.94677 0.93806 0.93660 Note: n- is the n't-ber of lags used to construct the expected rate of change in prices; ii iz the coefficient whlich provides the weighting system' of the lagged values of the. actmial rate of chang;e in priccs. See equation() in the text f'or further clarification. - 462 - Table B.6. Peru: t-sLatistics for the ExpcctEd Rate of Cl.ange of Prices in ti.e Demand for Money Function: ln(Mi/P) - a + b ln Y + c(dP/p)e (1959-1969) ; 3 4 5 6 7 8 Currency and demand deposits (M.) Ui.65 0.24206 0.25435 S0.21427 0.17346 0.15422 0.14330 0.7 0.29445 0.31209 0.25735 0.19451 0.16258 0.14305 0.75 0.35278 0.37651 0.30440 0.21114 0.1602S 0.12727 0.8 0.41597 0.44567 0.35345 0.21975 0.14166 0.03305 0.85 0.48279 0.51688 0.40157 0.21598 0.09992 0.O203. 0.9 0.55194 0.58696 0.44458 0.19486 0.027114 O.3552 0.9' 0.62215 0.o5265 3.47741 0.15127 O.E532 .2752- 1.e ',. E9222. 0.71109" 03.69658 0.08116 0.2'L3r 0.42i3 Currency ari, dema;d, saving and' timne deposits (M,) 0.65 1.61140 1.90322 1.5010 6 1.7342, 1.72793 1.71L3C 0.7 1.7553 2. 182(0 .I89:- 1.99292 1.89165 1.85733 0.75 1.903.: 2.5057C 2.347%. 2.22387 2.04917 1.98323 0.8 2.05634 2.683' 57 2.47244 2.18384 2.06660 0.85 2.21230 3.32280 3.E6Z:L3 2.72365 2.27051 2.08GO5 0.9 2.36302 3.80927 4.252.35 2.94217 2.27560 1.99769 0.95 2.50;06 4.31640 4.98649 3.063/44 2.16317 1.80643 1.0 2.64094 4.7 850 j T5.112 3.00796 1.91050 1.51656 Currency &nA all deposits, inrcluding foreign currency deposits(M) c~~~~~~ 3 0.65 0.78339 1.03189 1.03630 0.95549 0.90871 0.88O,3a 0.7 0.88732 1.23145 1.24996 1.11879 1.03721 0.99770 0.75 0.99424 1.46029 1.50724 1.29949 1.16231 1.08976 0.8 1.10261 1.72034 1.81960 1.49676 1.27400 1.14£44 0.85 1.21066 2.01186 2.200411 1.70628 1.35748 1.15409 0.9 1.31662 2.33220 2.661,24 1.91594 1.39204 1.08698 0.95 1.41872 2.67417 3.199!;. 2.09897 1.35194 0.93384 1.0 1.51542 3.02477 3 77 2.20921 1.21228 0.69542 Total liquidity of the public (N4) 0.65 0.15501 0.35482 0.35695 0.28798 0.24642 0.22417 0.7 0.24177 0.51181 0.52250 0.41381 0.34333 0.30203 0.75 0.33295 0.68920 0.718C09 0.55074 0.43554 0.36257 0.8 0.42727 0.88662 0.9471.0 0.69708 0.51507 0.39248 0.85 0.52335 1.10249 1.21426 0.84935 0.57108 0.37596 0.9 0.61976 1.33375 1.52176 1.00057 0.58940 0.219c7;} 0.95 0.71513 1.57558 1.86517 1.3.3781 0.55270 0.141' 1.0 0.80823 1.82152 12.230 51 1.24006 0.44309 O.On%n- Note: n is the number of lags used to c.or..truct the e-:pected rate of ch; Ice in prices; h is the coefficicent which providcs the weightin- :yftnm of thu lagged values of the actual rate of change in. prices. See equation ( ) in t. text for further clarification. - 463 - Table B.7. Peru: Demand for Money Regressions (1959-1969) (n= 4 , h = 1) In (1./P) = - 0.71367 + 0.66961 ny - 0.00419 (dP/P )e (3.16697) (11.30910) (0.71109) R = 0.96644 ; SEE = 0.02340 ; DW = 2.36727 (n =5 h = 1) ln (M21P) = - 1.27094 + 0.969-31 lay - 0.03732 (dP/F) (6.1164i7) (16.80705R) (5.51152) R2 = 0.98&3.2 ; SEE = 001e;4 ; DE? = 2.665-84 (n- 5 , h = 1) ln (IT /) = 1.602C,0 + 1.07474 l-y - 0.03857 (dP/p)e 3 (5.1Ce%'§'' (12.34J1J.,7) (3.77227) = 0.17022 ; SL1 O.Q2ci1 ; D; -- 1.60655 (n - 5 sh 1) ln (M414) = - 1.. .7554 + 1.0., JuMy 0.02571 (dPID) (4. S7) (11. GZ57'1) (2.23005) - - O .&3 S ; , -F~ 0.02.3 5 , :D~7 1.3'9444 Toit1 D eral Aetl ont Zsutate4 V3mg. of the Dem for Money (In Natural lo&arthM.) l,h;1j Z%tht-lj ted ,nt 'h Actuml _ Estiated - e5hi tn259h-1 (n-,b12 5p"4h*1 ,.Ti,h n.,hXl n",1J tw5h1 11259-5 z15-9 1959- 52 (f99-91959-75) (19S9,5-69 1S-'W _15 1959 2.02392a83 1.86485477 1.98327954 1.856528143 1.98281T66 2.41369Z4997 2.36616322 2.41081444 2:37229TTs 2-10713313 1960 2.005660144 1.95630417 2.03912196 1.94261M80U 2.03761951 2.44173815 2.e43C'7935 2.46019015 2.bh41655 2.471681338 1961 2.083(q902 2.0520o468 2.o9o6783 2.04099521 2.08978811 2.s540S1428 2.51951903 2.535704.19 2.52917499 2. 4-39LQ 1962 2.12954037 2.16392583 2.11913170 2.1610610T 2.31'974235 2.60165211 2.62165064 2.61081972 2.62339183 2.61855752 1963 2.161.70139 2.22578062 2.111.96117 2.22193855 2.1762P'362 2.65935003 2.674174904 2.66169852 2.67603136 2.66301478 1964 2.23430625 2.30027041 2.2179z2o0 2.32662872 2.22295553 2.725955(3 2.7Gs121.05 2.73522159 2.73813939 2.115358313 1905 z.2jy5>c6) 2.3339W033 z-.4679(31 2.33(b6t61 2.24,797005 2.73t07259 2.7681Q011 2.73791994 2.16506537 2.73SO0& 1966 2.3036(Vi49 2.40175689 2.233132960 2.388114957 2.28217121 2.81036609 2.01080112 2.77278h00 2.82815150 2.79049401 1)61 2.28T71598 2.39856311 2.99202(39 2.38MO403 2.2909s00t 2.7(221365 2.0o086o007 2.7(1110142 2.81931219 2.10261205 1968 2.27695955 2.304271301 2.201741.10 2.3102322 2.261.56o26 2.71476097 2.73330452 2.710681,63 2.73610241 2.71136191 1969 2.33321129 2.36110535 2.31646573 2.390167104 2.31646503 2.75461567 2.80351513 2.76T53351 2,7N,319111 2.75193525 1970 2.62192416 2.51975509 2.57836636 3.003h.5240 2.97075379 2.9191,8324 1911 2.7361&io04 2.60501764 2.6742371/4 3.09126970 3.053h3257 3.06102103 1912 2.80160126 2.7r915093 2.7616230r 3.151,23063 3.L34131.859 3.1410369). 1973 2.0911h923 2.8592211o 2.90516&891 3.23600022 3.257s.5489 3.20956113 1974 2.95183263 2.96185555 2.914188319 3.25543916 3.25081475 3.29691.l81 19-75 2.955sz1480 2.8T7144062 2.81.14390 3.23636205 3.10358700 3.21559854 (continued) Jiote: The estimated values were obtained from the regression In (M1JP) * n * b l Y + e ,/P) f t v o (jP/P) whleh etmtcd the' of the regreasion fr thc perlod 1959-69 or for 1959.75. Tho first sct of estimLted vuluce correepondl to the values of n end h which moaxflzed the in the regresslon for 1959-75g the second set correspondB to the n ond h aalueh which maxximied the J In the 1959-69 period. (For en explanation of the teaning of the parameters n ,ond h see p. e. shese estimated values were used to test the hypotheois of a change In th,e regression in 1970. - 465 - taken in that year, the demand for money regressions were run again for the period 1959-69 and the predicted values were compared statistically with those for 1959-75. 641. Tables B.5, B.6 and B.7 present the results of estimating equation (b.1) subject to (b.2) fo2 the period 1959-69. The estimated equations show some improvement in the R 's, except for the M13 definition. The n parameter shifts from 4 to 5 years or vice versa for the first three definitions of money, but h is equal to 1.0 in all cases as in the regressions for 1959-75. The Durbin-Watson statistics for the M1 and M2 definitions improve signifi- cantly so that there is no evidence of autocorrelation in these equations; this adds further evidence that the omission of the effects of the capital repatriation measures were a determinant factor in the serial correlation observed in the 1959-75 regressions. 642. Table 8 presents the actual and estimated values from the regressions for the whole sample period (1959-75) and the period 1959-69. As a result of the distortion of the 1970 measures, the estimated values for the period 1959-75 for M1 and M2 tend generally to overestimate those for the period preceding the measures and underestimate the years after 1970. The effect of the measures on M4 and M4 is not clearly discernible. Table B.9. Peru: F-Values for the Chow Test for Sample Periods 1959-69 and 1959-75 (Using best regressions for 1959-75) M (n = 5, h - 1) F - 27.00228 x2 (n = 4s,h = 1) F - 6.02793 3 (n = 4, h = 1) F - 0.17532 M4 (n = 5, h = 1) F = 1.19723 (Using best regressions for 1959-69) Hi (n = 4, h = 1) F - 29.50226 M (n = 5, h = 1) F - 8.69802 M (n = 5, h = 1) F - 0.93016 14 (n = 5A h = Id F - 1.19728 Sor.rc': Computad from T,b½ S using formula (b.4). - 466 - 643. Table 9 contains the F-statistics for the Chow test computed from Table 8. The F-tests were applied both to the comparison of the best estimated regressions for 1959-75 with their 1959-69 counterparts (i.e., with the same values of the parameters n and h) and to the comparison of-the best estimated regressions for 1959-69 with their 1959-75 counterparts. The latter tests, although less appealing, were conducted to see whether reversing the procedure would make any difference. When compared with the critical value of 4.15 for F (8, 6), the values of the F-statistics indicate that the hypothesis of a structural change in 1970 is acceptable for the definitions M1 and M at the 95 percent confidence level, but is not cLCceptable for M and M4. This indi- cates that the capital repatriation measures were'significant in altering the composition of the broadest definitions of money but not the total amount. Sample Period Regressions Corrected for Plossible Structural Shift 644. The results of the Chow test obtained above justify the use of a dummy variable to distinguish the periods 1959-69 and 1970-75 at least for the Ml and M2 definitions of money. Instead of running the entire set of regressions again for the model including the dummy variable, it was decided to retain the values of the parameters n and h which maximized the R in the model without the dummy variable. As shown earlier, the optimal values of n and h were fairly consistent regardless of the sample period, so that we could not expect very different results for these parameters. 645. The results of estimation of the model for the sample period 1959-75 including the dummy variable are presented in Table B.10. As expected, they show that the coefficient of the dummy variable is highly statistically significant for MI and M2 and less so for M14; for M3 the coefficient of the dummZ2variable is not statistically sign:Lficant and has the incorrect sign. The R of the regressions increases markedly for M1 and M2 and more moderately for M4 while it decreases slightly for M13. These results are all consistent with our a priori expectations and the sl:atistical tests conducted above. In the light of these results, it is concluted that the use of the dummy variable to take into account the capital repatriation measures of 1970 introduces a useful improvement in the predictive ability of the regressions for M1V M2 and M4P but not for M13. - 467 - 646. Because of the possibility that the 1970 measures may have affected real asset holdings only temporarily, additional tests were conducted to see whether the use of other variations of the dummy variable or other sample periods altered sigificantly the strong conclusions obtained above. Table B.11 presents the R and coefficients of the dummy variable for regressions on the demand for real money using different definitions of the dummy variable to distinguish 1970-71, 1970-72, etc., from the rest of the sample. The best results were obtained when the dummy variable distinguished all the years after 1970 from the preceding ones for M , M2 and M4; for M3 the best results were those without any dummy variable. Tese results add further evidence of a significant structural change in real asset holdings, except for M3, after 1970. 647. The results from moving regressions for sample periods ranging from 1959-69, 1959-60, etc., until 1969-75 with the use of the dummy variable are presented in Table B.12. Because these regressions refer to different sample periods, they are not strictly comparable but serve to illustrate some useful flcts. The fittings are best when the longest period is used. (However, the R for Ml and 14 was maximized for 1969-74 instead of 1969-75.) The coeffi- cients of the dummy variables were not significantly altered by adding or drop- ping a few more years. For M1 the t-statistics of the dummy variable is maximum when the 1959-72 period is chosen, indicating perhaps that the effects of the 1970 measures were particularly concentrated in the first two to three years. On the other hand, the coefficient of the dummy variable in regressions for M2 and M4 acquire maximum significance when the entire period is considered. 648. Summing up, the various tests carried out in this section tend to favor our acceptance of regressions contained in Table B.10 for Mi. M2 and M4 as the best explanations for the sample period 1959-75. - 468 Table B.10. Peru: Demand for Money Regressions L (Mi/P) - a + b ln Y + c (dP/P)e + d Di!:/ (1959-1975) Cn 3 5, h 1.0) ln (M/P) = - 1.05614 + 0.84347 inL Y - 0.02006(dP/P) e- 0.28071 DUM (3.36989) (11.37582) (2.42751) (8.27955) R2 - 0.98951; SEE 0.0343:L ; DW = 1.29047 (n - 4 , h = 1.0) ln (M2/P) = - 0.91879 + 0.88638 irn Y - 0.02155(dP/P)e - 0.12193 DUM (5.07650) (21.61853) (6.10419) (5.97035) R - 0.99473 ; SEE 5 0.02011 ; DW = 1.92290 (n = 4 , h= 1.0) ln (M3/P) = - 1.18702 + 0.94030 ln Y - 0.01945(dP/P)a + 0.01095 DUN (4.88195) (17.07103) (4.10148) (0.3t>905) 12 - 0.98677 ; SEE 0.02701 ; DW - 1.75425 (n = 5 * h = 1.0) in (M44/P) - - 1.68104 + 1.09891 In Y - 0.02772(dP/P) - 0.07656 DOY (6.70255) (18.52027) (4.19198) (2.82185) R2 0.99185 ; SEE = 0.02746 ; DW - 1.53069 1/ DUM is defined as 1 for 1959-69 and 0 for 1970-75. - 469 - Table B.11. Peru: Selected Statistics from Demand for Mloney Regressions With and Without Dummy Variables: ln (Mi/P) - a + b in Y + c (dP/P)e + d DUMMY (1959-1975)-' DUM DUM 1 DUK 2 DUM 3 DUN 4 DUM 5 Withoul -i2 dummy 0.9 89511 0.95273 0.94839 0.94475 0.94482 0.94363 0.93889) 9 I0.99473 0.98383 0.98289 0.98417 0.98327 0.98142 0.98167: N3 0.98677 0.98694 0.98668 0.98661 0.98665 0.98678 [0:-89564 H44 0O.9918 0.98958 0.99050 0.98843 0.98822 0.98818 0.98779( Coefficient of the duoy variable Ml -0.28071 -0.13279 -0.09689 -0.03470 -0.09734 -0.12729 (8.27955) (2.25798) (1.89065) (1.57573) (1.58255) (1.47525) X2 -0.12193 -0.05432 -0.03826 -0.04311 -0.04252 -0.03539 L(5.97035)1 (1.69319) (1.41336) (1.79268) (1.52951.) (0.90158) H3 0.01095 0.01407 0.00514 -0.00090 -0.00380 -0.011:37 (0.39905) (0.57527) (0.25379) (0.04788) (0.1803E) (0.404:71) N4 -0.07656 -0.04624 -0.04457 -0.02973 -0.03164 -0.04329 1(2.82185) (1.84478) (2.23299) (1.33145) (1.22644) (1.20676) 1/ DUHL is defined as I' for4j-i-69 and 0 for 1970-75. DUM 1 is defined as 0 f f ;970-74 and 1 otherwLse. DUM 2 is defined as 0 for'1970-73 and 1 othervise. DUM 3 is defined as 0 for 1970-72 and 1 otherwise. DUN 4 is defined as 0 for 1970-71 and 1. otherwise. - 470 Table B.12. Peru: Selected Statistlcs from Demand for Money Regressions: in (14k/P) a + b n Y + c (dP/P) + d DU: for Various Sample Periods Peid of Definiestia. 1959-69 1959-71 1959-72 1959-73 1959-74 1959-75 K-2 Hi 0.96644 0.98563 0.98890 0.98870 0.99025 0.98951 0.98312 0.98847 0.99198 0.99404 j99526j 0.99473 143 0.97022 0.96861 0.97581 0.98103 0.98477 0.98677 144 0.96869 0.97935 0.98535 0.98333 0.99035 | .99185 Coefficient of the dunmxy variable ;5 - -0.27144 -0.27704 -0.27761 -0.27275 -0.28071 (10.4027'3) ( 5 (9.12354) (8.81017) (8.27955) - -0.10928 -0.1102L8 -0.11212 -0.112 2n -0._o, 119 3 (4.94025) (5.3:L776) (5.63553) (5.84:501) __5__7___ H3 - 0.01470 0. 0:1719 0.01391 0.01'398 0.01095 (0.43736) (0.45449) (0.47568) (0.39905) 44 -0.07219 -0.07353 -0.07383 -0.07674 -0.07656 (2.32747) (2.52614) (2.53233) (2.69459) 1/DUM is defined as 1 for 1959-69 and 0 for 1970-75. - 471 - Table B.13. Peru: Demand for Money Regressions e~~~~~~~~~~~~_ . In (O4 M ) a + b y + c e (1959-1975) (n - 5, h = 1.0) ln (1/P) - 3.48336 + 1.40225 ln Y + 5.82849 (9.39236) (14.04654) (2.95811) R2 - 0.94198 SEE = 0.08070, DVt 0.63S''6 (n = /:, IL -.= ) -ln (M4JP) = - 2.5Q;689 + 1.13061- In Y + 3.390r)2 ne 2 (10.96Qe. ) (24.s3°}4$44) (4.62VqOt 2 O < o0.q7716 ; Si*, -7 0 .0 4i4 O li ._ 4, h 1.0) in (I./P) - 1A25883 0 94S'0 -ia Y 4 2.14i:1 Ee (10.4t !524) (32A21hZ8,% (4.607S- i 12 = 0.S87GO ; Sr. 1- 0.02C - rD s.7 * r 0 (u = 5 ' h 1.0) 1D (/P) - - 2.34055 - 1.22335 In Y + 3.5,474 E' (14.83661) 031.51556) (4.55;8010, 12 0 0.98665 ; SEE 0.03514 ; Dr 1.26C9-7 - 472 Table B.14. Peru: Demand for Money Regressions ln (M./P) a + b Y + c Ri + d DUM (1959-1975' (n = 5, h = 1.0) ln (N1!/P) - 1.15200 + 0.86669 ln Y + 2.38888 R - 0.27667 DUM (3.57596) (11.24057) (2.59788) (8.24299) 72 = O.93996 ; SEE = 0.03356 ; DW = 1.28067 (n = 4, h = 1.0) ln (M /P) = - 0.97128 + 0.88970 ln Y i- 2.35953 Re _ 0.13527 DUM 2 (5.43052) (22.30293) (6.36641) 2 (7.18170) -V = 0.99505 ; SEE = 0.01948 ; DIP = 1.80361 (n = 4, h = 1.0) in C13/P) = - 1.23107 + 0.94241 ln Y + 2.11643 R - 0.00335 D.UN (4.93660) (16.93010) (4.07246) (0.12858) = 0.98637 ; SEE = O.0?712 ; DW = 1.68010 (n 5, l 1. 0) In C4/P -- 1.67/i72 + 1.07737 ln Y + 2.76886 _ 0.08135 DUN (6.50184) (18.56222) (4.00956) 4 (2.96687) = 0.991/;3 SEE = 0.02016 ; DW' = 1.50087