Peru - At the crossroads : building a modern state

Report No. 11943-PE Peru At the Crossroads Building a Modern State March 31, 1994 Country Department I Country Operations Division Latin America and the Carribbean Region FOR OFFICIAL USE ONLY Document of the World Bank This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization  CURRENCY EQUIVALENTS (As of December 31, 1993) Currency Unit Nuevo Sol (S/.)' US$1.00 S/.2. 16 S/. 1.00 US$0.46 1 On July 1, 1991, a new monetary unit, the Nuevo Sol (SI.), was introducd at a conversion factor of S/. 1.00 = I/m. 1.00. The so-called Inti Million (I/m.) -- equivalent to one million Intis (I/. 1,000,000) -- was introduced on December 16, 1990 to simplify accounting and as a means of transition between the Inti and the Nuevo Sol.  FOR OFFICIAL USE ONLY LIST OF ACRONYMS APAFAS Asociaci6n de Padres de Familia (Parents Association) BCRP Banco Central de Reserva del Perd (Central Reserve Bank of Peru) CD Certificate of Deposit CEM Country Economic Memorandum CEI Centro de Educaci6n Inicial (Pre-School Centers) CIAS Comisi6n Interministerial de Asuntos Sociales (Interministerial Commission on Social Issues) CMC Cash Management Committee COMDES Comite de Desarrollo Social (Committee on Social Development) COMUNED Concejos Comunales de Educaci6n (Communal Education Committee) CONASEV Comision Nacional Supervisora de Ingresos y Valores (Stock Exchange Supervisory Commission) CP Commercial Paper ENDES Estudio Nacional de Demografia y Salud (Study of Demographics and Family Helth) FONCODES Fondo de Compensaci6n y Desarrollo (National Compensation and Development Fund) GDP Gross Domestic Product GNP Gross National Product ELT Electricity Transmission Towers IDB Inter-American Development Bank IMF International Monetary Fund INEI Instituto Nacional de Estadistica e Informaci6n (National Statistics and Information Institute) INFES Instituto de Infrastructura de Educaci6n y Salud (Health and Education Infrastructure Council) MEF Ministerio de Economia y Finanzas (Ministry of Economy and Finance) MERT Marginal Effective Rate of Tax METC Marginal Effective Tax on Capital METL Marginal Effective Tax on Labor OR Operaciones de Reporte (Report Operations) PAM Partial Adjustment Model PIETBAF Nutritional Support Program for Children PREDES Programa Regional de Desarrollo (Regional Social Development Programs) PRONAA Programa Nacional de Apoyo a la Alimentaci6n (National Nutrition Assistance Program) PRONAME Programa Nacional de Mantenimiento de Equipos (National Maintenance and Equipment Program) PRONOEI Programa Nacional de Educaci6n Inicial (Pre-School Education Program) RER Real Exchange Rate SL Sendero Luminoso (The Shining Path) USES Offices of Educational Services VAR Vector Autoregression System VAT Value Added Tax WDR World Development Report ZONADIS Zones of Integral Health Development This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  PERU AT TNR CROSSROADS BDUILDIG A MODIRtN STATA CON7ENTS INTRODUC7ION AND EXECUTIVE SUMMARY i.......... Introduction ............................... i Economic Policy Before July 1990 ................. i Economic Growth ........................... ii Fiscal Deficits and lnfation ..................... ii Terrorism ............................... iii Economic Policy After July 1990 .................. ii The Stabilization-Structural Reform Program ..... ...... iii Problems in the Transition ........... ........... iv Policy Recommendations ............... ........ v CHAPTER I: ECONOMIC GROWTH 1960190 .... ........ I A. SourcesofGrowth,1960-90 .................. 3 B. Structural Changes in the Economy ............. 4 1. The Public and Private Sectors ........ .... 5 2. Economic Sectors .................... 6 3. Exports ........................... 7 C. Some Macroeconomic Variables .............. . 9 1. Investment ......................... 9 2. The External Sector ................... 11 D. Income Distribution ....................... 12 E. Economic Policies: Impactand Pospect ......... 13 CHAPTER n: THE ECONOMIC COSTS OF TERRORISM . .. 15 A. Intoduction ............................ 15 B. Styhized Facts ................ .............. 15 C. A Framework to Analyze Terrorism ...... ....... 22 1. Terrorism and Net Product ............... 23 2. The Spread of Terrorism and Segrg. 25  3. Effects of Terrorism Across Activities ........ 27 4. Terrorism and Income Distribution ..... ..... 28 D. Policy Implications ...... .................. 29 CHAPTER III: GOVERNMENT EXPENDITURE AND BUDGETARY APPROPRIATIONS ...... .. 30 A. Introduction ............................ 30 B. Public Finances Before July 1990 .............. 31 C. Government Expenditure After July 1990 .... ...... 35 D. What Remains to be Done ................... 38 CHAPTER IV TAXATION, MACROECONOMIC STABILITY AND ECONOMIC GROWTH ...... ...... 39 A. Introduction ............................ 39 B. The Present Tax System .................... 39 C. Potential and Actual VAT Revenues ............. 42 D. Personal Income Tax ...................... 47 E. An Assessment of Taxes on Corporate Income ... .... 51 F. Marginal Effective Taxes ................... 53 CHAPTER V: INFLATION AND STABIUZATION .... .... 57 A. Introduction ............................ 57 B. Trends in Money, Incomes and Prices .... ........ 57 C. Money Supply .......................... 63 D. Demand for Money ....................... 67 E. The Real Exchange Rate .................... 67 CHAPTER V: FINANCIAL MARKETS ................ 72 A. Introduction ............................ 72 B. Trends ............................... 72 C. Areas of Concern ........................ 74 D. Structure ............ .............. 79 E. Recent Performance of Financial Institutions ... ..... 82 F. Conclusions ............................ 88 CHAPTER VII: POVERTYRELEF .................. 90 A. Introduction ............................ 90 B. Living in Poverty ........................ 90  C. Earnings and Education ..................... 92 D. Earnings and Ethnic Origin . ................. 93 E. Social Policies . ......................... 100 F. Education Reforms . ....................... 101 G. Reforms in Health, Diet, and Nutrition ......... .. 103 1. Health . .......................... 103 2. Diet and Nutrition Assistance ........... .. 103 3. Water and Sanitation .................. 104 H. Conclusions ............................ 105 TEXT TABLES Table 1. 1: Determinants of Growth in Per-Capita GDP ...... ............. 3 Table 1.2: Sources of Economic Growth. 4 Table 1.3: Sources of Growth in the Public and Private Sectors, 1980-88. 5 Table 1.4: Sectoral Make-Up of GDP, 1960-90. 6 Table 1.5: Sources of Growth in Manufacturing, 1975-87. 7 Table 1.6: Peruvian Exports, 1970-90. 8 Table 1.7: Investment and Domestic Savings, 1961-89. 9 Table 1.8: Peru's External Debt, 1970-90 .12 Table 2.1: Terrorist Acts by Departments .17 Table 2.2: Loss of Income from Terrorism. 25 Table 3.1: Government Expenditure as a Percentage of GDP .30  -4- Table 3.2: Credit to the Financial Sector and Expansion of the Money Base .......... 34 Table 3.3: Public-Sector Operations ...... ............ 37 Table 4.1: Tax Ratios and Structures ................. 40 Table 4.2: VAT Effectiveness in Latin America ........... 41 Table 4.3: Changes in Revenues by Taxpayer .41 Table 4.4: Classification of GDP According to VAT Status. 45 Table 4.5: Explaining the Gap Between Potential and Actual VAT Revenues .46 Table 4.6: International Comparison of Personal Income Tax .SO Table 4.6A: Peru: Personal Income Tax Exemptions and Credits .50 Table 4.7: Withholding Taxes by Domestic Corporations .51 Table 4.8: Country Comparison of Corporate Taxes .52 Table 4.9: An International Comparison of METCS .54 Table 4.10: Peru: The METC, Tax Incentives and a Tax Reform .55 Table 4.11: An International Comparison of Marginal Effective Tax on Labor .56 Table 5.1: Inflation, Devaluation, and Growth of the Money Base .62 Table 6.1: Monetary Holdings, Broadly Defined. 75 Table 6.2: Open Exposure .................... 81  -5- Table 6.3: Liquidity Ratios ....................... 83 Table 6.4: Financial Sector Restructuring ...... .. ....... 84 Table 6.5: Measures of Operating Efficiency ....... ...... 86 Table 6.6: Profitability Indicators ....... .. .......... 87 Table 7.1: Poverty Indicators for 1991 ........ ........ 91 Table 7.2: Highest Educational Achievement ...... ...... 94 Table 7.3: Labor Force Participation, Ages 12 to 65 . ....... 95 Table 7.4: Mean Characteristics by Ethnicity ...... ...... 97 Table 7.5: Basic Earnings Functions by Ethnicity ..... ..... 98 Table 7.6: Migration from Place of Birth ...... .. ....... 99 Table 7.7: Enrollments in State Educational, 1985-91 ......... ......... 102 Table 7.8: Teachers Education Degrees ...... .. ........ 102 FIGURES 1.1 Peru's Productivity. 2 1.2 Annual Change in Gross Capital Formation .... ....... 10 2.1 Terrorist Acts and Deaths from Violence (Per Capita) .................... 16 2.2 Electricity Transmission Towers Damaged by Terrorists ......... ...................... 16 2.3 Distribution of Terrorist Actions Over the Decade .18 2.4 Terrorist Action Across Regions Over the Decade .19  - 6- 2.5 Indicators of Regional Presence of Mode o State ............................. 21 2.6 Production Frontier Between Tradeables and Non-Tradeables ............................ 24 3.1 Ratio Total Public Expenditure to GDP (RGC2Y) and Public Sector Deficit to GDP (DEFC2Y) Ratio Tax Collection to GDP (RIGCY) and Ratio Inflation Tax to GDP ........................ 32 3.2 Ratio Public Expenditure to GDP (RGC2Y) and Central Government Expenditure GDP (RGGDP) Ratio Total Public Expenditure to GDP (RGC2Y) and Tax Collection to GDP (RIGCY) .... .......... 33 4.1 Peru: Personal Income Tax Revenues ....... ........ 48 4.2 Peru: Personal Income Tax Revenues and the Real Wage .49 4.3 Peru: Corporate Income Tax Revenues .53 5.1 Real Domestic Money and Real Income Real Domestic and Foreign Money .58 5.2 Estimate of Foreign Money Held in Peru Estimate of Peru-Dollrs .59 5.3 Real Domestic and Foreign Money Liquidity and Income .61 5.4 Sources of Growth of the Money Base ..... .......... 63 5.5 Rate of Growth of Base Money and Devaluation .............................. 64  -7- 5.6 Rate of Devaluation: Actual and Expected Stock Market Index in Dollar and Real Terms ............................ 66 5.7 Index of Purchasing Power Parity Between Peru and USA Real Exchange Rate Between Peru and USA .... ....... 71 6.1 Spread Monthly Interest Rates (Soles) ............... 76 6.2 Spread of Annual Interest Rates (US$) ............... 77 6.3 Lending Rates in Soles ........................ 78 6.4 Lending Rates in Dollar ......... .............. 79 APPENDICES IA Factors Determining Investment ................... 107 IB Sources of Growth Decomposition ...... ........... 113 2A The Economic Impact of Terrorism ................ 115 2B Labor, Capital, and Terrorism: Key Elasticities ......................... 120 4A A Framework for a Decomposition of VAT Revenues ......................... 121 4B Estimating Revenue Functions and Optimal Taxes ......................... 123 4C Measuring Price Distortions with Effective Tax Rates ......................... 127 4D Marginal Effective Tax Rates on Factors ............................... 131 SA The DHSY Model ........................... 134 SB VAR Estimates ............................. 143 SC Estimating the Stock of Dollars in Circulation ............................. 157 6 The Financial System . ........................ 159 7 Institutional Framework of Social Reforms .... ........ 173  -8- STA77S77ICAL APPENDIX ...... .. ................. 174 Table S.1 Peru - Total Population Distribution . . 175 Table S.2 Peru - Demografic Indicators . .176 Table S.3 Peru - Labor Force by Sectors . .177 Table S.3(A) Peru - Labor Force by Sectors .178 Table S.4 Peru - GDP by Sectoral Origin .179 Table S.4(A) Peru - Real GDP Growth Rate .180 Table S.5 Peru - GDP By Expenditure .181 Table S.6 Peru - Nominal GDP By Expenditure .182 Table S.6(A) Peru - Nominal GDP By Expenditure .183 Table S.7 Peru - Real GDP Per Capita .184 Table S.8 Peru - GDP Per Capita, By Department .185 Table S.8(A) Peru - GDP Per Capita, By Department .186 Table S.9 Peru - Summary Balance of Payments .187 Table S.10 Peru - Merchandise Exports by Commodities 188 Table S.ll Peru - Merchandise Exports by Commodities 189 Table S.12 Peru - Merchandise Imports .190 Table S.13 Peru - Merchandise Imports .191 Table S.14 Peru - Export, Import and Terms of Trade Indexes .192 Table S.15 Peru - Total Extemal Debt Outstanding and Disbursed .193 Table S.16 Peru - Service Payments, Disbursements and DOD Public Debt .194 Table S.17 Peru - Non-Financial Public Sector Operations .... 195 Table S.17(A) Peru - Non-Financial Public Sector Operations . . .. 196 Table S.18 Peru - Central Government Operations .197 Table S.18(A) Peru - Central Government Operations .198 Table S. 19 Peru - Non-Financial Public Enterprise Operations .199 Table S. 19(A) Peru - Non-Financial Public Enterprise Operations .200 Table S.20 Peru - Current Revenue of the Central Government By Source .201 Table S.20(A) Peru - Current Revenue of the Central Government By Source .202 Table S.21 Peru - Current Expenditure of the Central Government .203 Table S.21(A) Peru - Current Expenditures of the Central Government .204 Table S.22 Peru - Summary Accounts of the Consolidated Financial System .205  9- Table S.23 Peru - Summary Accounts of the Consolidated Banking System ............. 206 Table S.24 Peru - Summary Accounts of the Central Reserve Bank 207 Table S.25 Peru - Summary Accounts of Banco Nacion 208 Table S.26 Peru - Summary Accounts of Development Banks . . 209 Table S.27 Peru - Change of Consumer Price Index .210 Table S.28 Peru - Exchange Rates ................... 211 BBLIOGRAPHY) .............................. 212  I  PREFACE The government of Peru has met an enormous challenge in resuming economic growth and reducing poverty after three decades of economic stagnation and social instability. The reforms after July 1990 have been extremely successful: terrorism has been greatly subdued, inflation has been drastically reduced, and there has been vigorous economic growth. Nonetheless, Peru is at a crossroad, tax collection is low, and there is lack of institutional mechanisms to control efficiency in government spending. Therefore, building a modern state is crucial in sustaining economic growth and reducing poverty. Edgardo Favaro is the principal author of the Country Economic Memorandum. Edgardo Favaro wrote the Introduction and Executive Summary, chapters II, III, V, and VII. Chapter I draws on a background paper written by Victor Elias (Consultant). Cheikh Kane wrote Chapter IV. Cesar Burga wrote Chapter VI and prepared the Statistical Appendix. Vincent McCullough edited the original manuscript. Gabriela Huffman provided secretarial support. The study benefited from the comments of many colleagues, in particular Karen Cavanaugh, Armeane Choksi, Sebastian Edwards, Janet Entwistle, Heywood Fleisig, Valeriano Garcia, Feliciano Iglesias, Cheikh Kane, Michael Michaely, Saumya Mitra, Jacques Morisset, Izumi Ohno, Demetris Papageorgiou, Rainer Steckhan, Carlos Silva, and Mateen Thobani.   INTRODUCTION AND EXECUT SUMMARY Introduction i. In the past 30 years, the Peruvian economy has moved from feast to famine. Throughout the 1950s, when govemment intervention policies were limited (and property rights were well established), there was a steady increase in per-apita income. Such policies were gradually abandoned, on the grounds that they did little to help the poor. In their place, came agrarian reform (including changes in property rights), administered prices and overexpansion of state-owned enterprises. The results for economic growth and poverty relief were disastrous. ii. Policies aimed at social reform worsened and perpetuated poverty. Repeated experiments in social engineering created economic and institutional uncertainty and a prolonged decline in govemance. That, in turn, undermined productivity, fostered anarchy and public turmoil, and, ultimately, made poverty reduction impossible. W. By 1990, per-capita income was below that of 1966; political violence was claiming 3,000 lives each year; tax collections were less than 5 percent of GDP; and prices had increased by a mind-boggling factor of 27 million over three decades. In fact, Peru's economy had all but collapsed. It was the finale on 30 years of misguided policies, economic mismanagement, and, in the past 10 years, rampant and escalating terrorism. iv. For a successful transition from economic stagnation, increased inequality, and anarchy to sustained economic growth and poverty relief, the role of the Peruvian state must be redefined. This inevitably means strengthening government effectiveness to increase the provision of key public goods, while, at the same time, reducing the scope of government intervention in the economy to increase efficiency in the allocation of resources. Economic Policy Before July 1990 v. Economic policies pursued 1950-66 (essentially limited government intervention) were in sharp contrast with those in the next 15 years, a period of extensive intervention. The drastic shift in policy was prompted by changes in both Peruvian society and in political ideas. vi. From 1940 to 1961, Peru's population increased by about 48 percent, while Lima's population tripled, and poverty became visible. The explosive increase in urban population was perceived by successive govemments since 1960 as a prelude for social and political turmoil. Those political views had a common thread: the perception of market mechanisms as forces unable to assure rapid poverty reduction. Each government offered its own idiosyncratic solution to control latent social pressure. Vast experimentation in social engineering led to a steady increase in the scope of government intervention in the economy, drastic changes in the structure of property rights, and a gradual deterioration of governance.  Peru at the Crossroads - ii - Building a Modem State A slowdown in economic growth, growing fiscal deficits and inflation, and mounting terrorism. Economic Growth vii. The contrast in economic performance between 1950-66 and 1967-90 is as sharp as that in economic policy. In 1950-66, economic growth and poverty reduction were sustained. Per-capita GDP increased at 3 percent a year, exports increased by 8.4 percent, productivity increased at 1.7 percent, and the average income of the poorest 40 percent of the population rose by 1.2 percent a year. By contrast, per-capita GDP in 1967-90 declined by 0.7 percent a year, exports grew at only 1.4 percent, productivity fell by 2.9 percent, and the average income of the poorest 40 percent dipped by 0.1 percent." viii. The slowdown in economic growth after 1967 was due to a steady decline in productivity, rather than an absence of investment in human and physical capital. Had productivity remained constant between 1967 and 1990, per-capita income would have increased by 31 percent. Instead, it fell by 16 percent because Government policy effected a shift of resources to economic activities that yielded a negative economic return, and because successive governments overlooked the provision of public goods, thus provoking a drop in economic productivity. Fiscal Deficits and Inflation ix. The ratio of government expenditure to GDP grew steadily from the 1960s until it finally collapsed in 1989. By 1990, it was once again at its 1960s level. Three decades of continuously expanding government intervention in the economy had unexpectedly resulted in the collapse of the public sector, forced by the systematic depletion of every source of government finance. x. The steady increase in government spending since the 1960s was not accompanied by a parallel increase in tax collection. The ratio of tax collection to GDP fell in the 1980s, and collapsed to 5 percent in June 1990. The result was a steady increase in the fiscal deficit and in the government's reliance on the inflation tax as a source of revenue. In fact, in 1988-90, the revenue collected through the inflation tax was equal to that from all other taxes combined. xi. Government spending grew steadily in the 1980s, until it collapsed in 1989. But expenditure on the production of public goods fell steadily throughout the 1980s. The 1/ See Chapter 1 and Richard Webb, The Political Economy of Poverty Equity and Growth: Peru 1948-85. World Bank (processed). Paper presented at the Conference on "The Political Economy of Poverty, Equity and Growth" in Annapolis, July 24-26, 1988.  - iii - Introduction and Executive Summary explanation for this paradox is as simple as it is dramatic: the increase in government spending in the 1980s was caused by the quasi-fiscal deficits run by the development banks and state-owned enterprises. The decline in spending on conventional public programs meant a steady degradation of the civil service and in the provision of public goods. xii. In the 1980s, the state's inability to provide a stable currency (and public order) forced the private sector to develop alternatives. It was able to offset, at relatively low cost, the absence of a stable domestic currency by gradually replacing it with foreign currency. However, the private sector was still unable to devise an efficient solution to terrorism, and, so, both productivity and new investments fell. Terrorism xiii. The spread of terrorism was responsible for a drop in per-capita income of about 8 percent from 1983 to 1990, because it shifted resources away from productive activities, destroyed economic resources, and increased the costs of production and trade. xiv. The spread of terrorism was due mainly to inadequate law enforcement, especially in the Andean highlands. This, in turn, was a side effect of the agrarian reform of the 1960s and 1970s, which dismantled an institutional arrangement that related law enforcement to the structure of land property, but failed to replace it with a more modern system. Economic Policy After July 1990 xv. The government appointed to office in July 1990 faced an economy in tatters. Per-capita income was below that of 1963, prices has increased by a factor of 27 million over three decades, government provision of public goods was dismal, tax collection amounted to only 5 percent of GDP, and political violence was claiming 3,000 lives a year. The new government implemented a huge program of economic reforms. There was a rapid drop in inflation and an across-the-board increase in productivity. Even so, problems persist and must be overcome before a path of sustainable economic growth and poverty reduction can be resumed. The Stabilization-Structural Reform Program xvi. The main objectives of the stabilization-structural reform program introduced after July 1990 were to lower inflation and transform the state-dominated economy into a market-oriented one. The program included stringent fiscal and monetary measures. A cash management committee maintained public expenditure in line with tax collection and credit from the Central Bank to the public sector was eliminated. The tax system was reorganized  Peru at the Crossroads - iv - Building a Modem State and interest rates and foreign-exchange transactions were liberalized. Structural reform also called for liberalization at both national and sectoral levels. It strengthened the role of market mechanisms in the allocation of resources and, implicitly, defined a subsidiary role for the Peruvian state in the economy. The reform program resulted in a drastic drop in inflation (from 7,600 percent in 1990 to 139 percent in 1991, and 57 percent in 1992) and an increase in productivity. Problems in the Transition xvii. Resuming economic growth and reducing poverty after 30 years of economic stagnation and social instability takes time. It requires (sometimes gradual) changes in the roles of the state and the private sector. Reducing the scope of government intervention in the economy and increasing government effectiveness will not be achieved by decree, although that may be a starting point. Similarly, reshuffling resources from less productive to more productive uses will not be effected by law, but trade reform may be the beginning. xviii. Government effectiveness has greatly increased since 1990, particularly in restoring public order and in economic stability, implementing poverty reduction programs, and tax collection. Even so, public order is still weak, stabilization is fragile, and tax collection low. Because Peru's fledgling institutions are still weak, results still depend too much on isolated individuals. The civil service has been ravaged by years of negligent policies, and the cadres needed to increase government effectiveness often do not exist. xix. Lack of institutional mechanisms to control efficiency in public spending and still-low tax collection are constraining the production of public goods and government effectiveness, and causing low economic productivity. Supporting the provision of public goods depends ultimately on tax revenues, and therefore on the design and enforcement of an effective tax system. The current difficulties in tax collection result from problems in the present system's design and enforcement. Increasing expenditure on public programs of questionable social return will not increase government effectiveness. An efficient use of public sector resources, therefore, depends on implementing institutional mechanisms that will reduce discretion and increase accountability. xx. Two decades of state intervention have also severely scarred the private sector. For years, private sector management wasted resources trying to adapt to intricate govemment regulations, rather than devoting their energies to surviving in a competitive economy. In many Peruvian firms, the managerial skills required for the new market conditions may not be available and this may affect the speed of economic recovery. Moreover, it may take time to build up the human capital needed to sustain private-sector development.  - v - Introduction and Executive Summary xxi. The banking sector in particular is likely to encounter difficulties in the new economic environment. Its loan portfolio has a high percentage of non-performing loans. Its network of branches and its work force are overextended. These characteristics of the banking sector result in a large spread between loan and deposit interest rates. This spread will be reduced as new financial institutions enter the market and as new financial instruments substitute bank credit. In the meantime, Peru's partial financial isolation from the rest of the world helps to perpetuate a high cost of credit, and therefore inhibits faster adjustment of the economy. xxii. The recent fall in inflation has brought about a drop in the real exchange rate. Economic theory predicts that this will be temporary but does not specify the length of the transition. Increasing credibility in stabilization will certainly reduce the length of time necessary for the transition to a new, higher real exchange rate. Eliminating barriers that make the short- and long-run supply of non-tradeable goods less than infinitely elastic will also have the same effect. In the meantime, there is little the government can do to eliminate this phenomenon beyond creating a fiscal surplus. Maintaining the stabilization program's targets is crucial, however, even in the face of pressure from interest groups to intervene in foreign-currency market. Policy Recommendations xxiii. Increasing government effectiveness means defining which public goods will be produced and in what amounts, establishing mechanisms to control efficiency in public spending, increasing tax collection, building an efficient civil service, and establishing an efficient regulatory framework. xxiv. The priorities in the provision of public goods are: Reestablishing public order. This is the government project that will yield the highest social rate of return in Peru. Improving security conditions means not only allocating resources to law enforcement in areas of turmoil, but also establishing incentives for those citizens to contribute to the provision of local security. A clear definition of property rights on land and other assets may have more lasting effects on security than an expensive defense program. Maintaining economic stability. This will require maintaining a tight monetary policy, reducing the fiscal deficit, completing the reintegration of Peru in the world financial market, and reducing uncertainty about the stability of the tax regime. Securing property rights means providing an effective judiciary, and avoiding intervention in private contracting. For instance,  Peru at the Crossroads - vi - Building a Modern State implementing land titling and establishing a market for water rights may have a large impact on productivity and, at the same time, help to preserve the natural environment. * Efficienr poverty reduction includes the design of a framework that will provide primary health and education services, and programs targeting those in extreme poverty. Initiatives that attempt to reduce illiteracy and infant mortality, and improve primary health standards will also help. * A regulatory role in some economic areas. This will require a framework to regulate natural monopolies and to compensate losses resulting from externalities, whenever these losses are higher than the costs required to regulate and to monitor enforcement. For instance, it is necessary to regulate telecommunications, electricity generation and distribution, fisheries, hydrocarbons, and construction and maintenance of infrastructure. xxv. Strengthening budgeting, accounting, and control systems will help in the evaluation of spending and will reduce the likelihood that future administrations might allocate resources according to their own agenda. Increasing tax collection is necessary to support the production of public goods. This means broadening the tax base through eliminating exemptions and enforcing tax laws. Building an efficient civil service is necessary to assure that government objectives are achieved at minimum cost. xxvi. Defining a subsidiary role for the state in the economy includes continuing privatizations, strengthening the role of markets, eliminating distortions in the labor and capital markets, and continuing to open the economy to competition with the rest of the world.  CHAPTER I ECONOMIC GROWTH 1960-90 1.1 When Peru's economy collapsed in 1990, it was the finale on 30 years of misguided policies, economic mismanagement and, in the past 10 years, rampant and escalating terrorism. It was not always so. In the 1950s, there was a sustained increase in per-capita income, thanks to limited government intervention and free trade. These policies were then abandoned on the grounds that they did little to help the poor. They were replaced by agrarian reform, changes in property rights, administered prices, and over-expansion of state-owned enterprises. By the 1980s, the results for both economic growth and poverty relief were disastrous. Per-capita GDP, which had grown at a steady 2.6 percent a year in 1960-75, began to decline. It fell by 2.2 percent annually from 1975 to 1991, with the fall becoming sharper in 1987 (see Figure 1.1). External, as well as domestic, factors contributed to the slide in Peru's economy. Between 1975 and 1982, the oil price shock had the biggest impact, while from 1983 to 1990, political unrest and terrorism took its toil on investment and productivity. 1.2 Over the past 10 years, Peru's economic performance is among the worst in Latin America, with only Nicaragua registering the same steep declines in GDP seen in Peru in 1983 and 1987-90. Although the rest of Latin America also saw per-capita GDP fall in the past 10 years (but total GDP rise by 12 percent), Peru's per-capita GDP fell by three times as much.   Peru at the Crossroads - 2 - Building a Modem State PERU' S PRODUCTIVITY 1400 1300- 1200 1100- 1000 f -,,,/'' 900 ,--" 800 ,, 700 -, . . . . . , . , . . , . . . . . . . , . 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 - Per-capita GDP Per-capita GNP Flgure 1.1 Source: Webb and Baca (1991). 1.3 GDP can be broken down into its components - labor, capital, and technology. So, therefore, can changes in per-capita GDP (see Table 1.1). In 1960-75, when Peru's economy was growing, an increase in capital-labor ratios accounted for 51 percent of GDP growth; the rest came from increased productivity (41 percent), and changes in labor force participation (8 percent). In 1975-90, when per-capita GDP declined by 3.4 percent a year, the fall in the capital-labor ratio accounted for 26 percent of the decline, and labor force participation increased. Hence most of the decline came from lower productivity.  - 3- Economic Growth 1960-90 Table 1.1: DETERMINANTS OF GROWTH IN PER-CAPITA GDP % Change In Per-Capita Labor's Capital/ GDP Share Labor Ratio Technology 1960-75 2.59 0.21 2.02 1.05 1975-90 -2.20 0.69 -0.88 -2.32 Notes: The value of 0.65 was used for B. Sources: Elias (1992); Webb and Baca (1991). 1.4 All Latin American (and most industrialized) countries had severe economic problems in the 1980's (see Elias, 1992). Peru was affected by much the same phenomena as neighboring countries and trading partners, but they were more strongly felt in Peru. They include slower world growth; natural disasters (such as El Nino); high external debt; Dutch disease; accelerating inflation; political unrest; disparity between domestic prices and international prices; adverse international capital movements; and a relatively closed economy. A. Sources of Growth, 1960-90 1.5 Changes in the composition of GDP inputs capture structural change in the economy and GDP can be disaggregated to quantitative and qualitative changes in factors of production (see Table 1.2 and Appendix 1B). For labor, the educational level of the Peruvian work force changed, as did its sectoral composition. In the case of capital, there were changes in its composition between the public and private sectors.  Peru at the Crossroads - 4 - Building a Modem State Table 1.2: SOURCES OF ECONOMIC GROWTH % Change In Labor Capital Total GDP Quantity Quality Quantity Quality Technology Inputs 1960-75 5.53 3.01 1.92 5.09 0.00 0.50 5.03 1975-80 0.12 3.17 2.26 2.23 0.10 -3.30 3.41 1980-85 -0.55 3.10 1.48 1.75 0.10 -3.36 2.81 1985-90 -0.73 3.10 n.d. 1.81 0.10 -3.06 2.33 Sources: Elias (1992); Webb and Baca (1991). 1.6 Growth of capital contributed substantially to Peruvian growth in the 1960-75 (almost 60 percent), but then slowed to a rate below that of labor. The decline in Peru's ratio of investment to GDP (at current prices) was three times as greater as elsewhere in Latin America; despite that, its capital stock grew in both the first and second halves of the 1980s. Technology which made a significant contribution to Latin American growth (at least until 1980) had virtually no impact in Peru. The improved quality of labor contributed substantially to the growth in labor and GDP throughout the past 30 years. The decline in GDP growth in 1975-90 appears to stem from the behavior of capital (not labor) and from the inefficient use of resources. Although capital formation declined sharply, labor inputs increased at a fast pace in both quantity and quality. According to the behavior of the labor and capital inputs, Peru's economy ought to have grown at about 3 percent a year in the 1980's. B. Structural Changes in the Economy 1.7 To identify the causes of Peru's poor growth in the second half of the 1980's, a deeper look is needed at the changes in the composition of the economy. These changes include those between the public and private sectors, among economic sectors, and between the domestic and external sectors.  - 5 - Economic Growth 1960-90 Table 1.3: SOURCES OF GROWTH IN THE PUBLIC AND PRIVATE SECTORS, 1980-88 %6 Change In Output Labor Capital Productivity Public 1.13 6.23 1.11 -1.77 Private 1.30 1.80 2.50 -0.96 Sources: Elias (1992); Webb and Baca (1991). 1. The Public and Private Sectors 1.8 Although output growth was more or less the same in both the private and public sectors in the 1980's, the use of inputs differed greatly (Table 1.3). The public sector had high growth in labor, but substantially lower capital growth than the private sector; and the fall in total productivity of factors in the public sector was almost twice as high as in the private. 1.9 In the 1980's, the public sector decreased its share of both consumption (down from 14 percent to 9 percent) and investment (from 30 to 16 percent). Its share of capital investment declined from 30 to 26 percent. The change in the composition of the capital stock (Table 1.2) had a positive effect on capital growth of 0.10 percent annually. This was not enough, however, to offset the greater productivity decline in the public sector. 1.10 The composition of public-sector employment varied. The share of government enterprises in total employment fell from 19.2 percent in 1975 to 16.6 percent in 1980, before increasing to 19.4 percent in 1990. In other words, in the 1980's, the government enterprise sector saw more employment growth than did the public sector generally. Because it has a smaller share of total employment than total capital investment, the public sector is more capital intensive than the private.  Peru at the Crossroads - 6 - Building a Modem State 2. Economic Sectors Table 1.4: SECTORAL MAKE-UP OF GDP, 1960-90 (at constant 1986 prices) % Change 1960 1975 1980 1990 1975-90 1980-90 Agriculture 16.2 10.4 9.4 11.8 0.96 1.68 Fishing 0.8 0.4 0.5 1.1 6.80 8.25 Mining 3.0 2.0 3.7 2.9 2.66 -3.05 Manufacturing 25.1 25.3 25.1 22.7 -0.62 -1.66 Construction 8.3 9.6 8.3 8.4 -0.78 -0.50 Services 46.6 52.3 53.0 53.1 0.24 -0.62 Total 100.0 100.0 100.0 100.0 0.12 -0.64 .Sources: Elias (1992); Webb and Baca (1991). 1.11 Agriculture and fishing escaped the general recession of the 1980's. Although growth in mining was positive for 1975-90 (thanks to strong growth in 1975-80), mining and manufacturing were the most severely affected. Some of the effects of changes in the sectoral composition of GDP can be evaluated by considering changes in the quality of labor (see Appendix 1B). For the whole of the 1980's, this change is estimated at an almost negligible 0.06 percent. (Verifying this involves estimating the labor quality effect, which might show the effect of the reallocation of resources.) But, to further explore the role of sectoral changes, let's look at the sources of growth in the manufacturing sector in 1975-87 (Table 1.5).  - 7 - Economic Growth 1960-90 Table 1.5: SOURCES OF GROWTH IN MANUFACTURING, 1975-87 % Change In Output Labor Capital Intermediate Inputs Productivity Aggregate value -0.16 1.41 -1.01 0.003 Production 1.46 1.41 -1.01 2.76 0.020 Notes: In the aggregate value version, the analysis used a B of 0.65, and in the production version, intermediate inputs accounted for a 0.55 share. Sources: Elias (1992); Webb and Baca (1991). 1.12 Productivity growth in Peruvian manufacturing (using either the aggregate or gross value of production) was zero, despite a big increase in the share of intermediate inputs in production. Although part of the decline in capital formation is attributable to manufacturing (at least up to 1987), that does not explain the downward trend in total productivity in 1975-90. 1.13 From 1975 to 1987, the average size of firms in manufacturing fell by at least 10 percent, with smaller firms becoming less capital intensive. The industrial make-up of the manufacturing sector also changed slightly. Most affected by recession were metals and machinery, basic metals, and paper and printing; the least affected were food, textiles, and leather. 3. Exports 1.14 From 1970 to 1990, Peruvian exports underwent a big change in composition (Table 1.6). Nontraditional exports increased significantly and mining products more or less maintained their share. Oil exports increased to about 20 percent of the total between 1980 and 1985, before falling back. Fish meal, which accounted for a large chunk of exports in 1960, thus began to recover after losing its leading role in 1975-85.  Peru at the Crossroads - 8 - Building a Modem State Table 1.6: PERUVIAN EXPORTS, 1970-90 (percentage of total exports) 1970 1975 1980 1985 1990 Fish meal 29.3 12.6 5.0 4.0 10.4 Copper, lead, zinc 35.2 33.5 34.3 31.8 40.5 Oil 0.7 3.1 20.2 19.6 8.6 Nontraditional exports 3.3 7.2 21.6 24.0 29.7 Source: Webb and Baca (1991).  - 9 - Economic Growth 1960-90 C. Some Macroeconomic Variables 1. Investment Table 1.7: INVESTMENT AND DOMESTIC SAVINGS, 1961-89 Investment Rate Domestic Savings Rate (gross investment (domestic savings to GDP) to GDP) 1961-67 21.0 19.8 1968-69 14.7 15.6 1970-75 18.4 17.4 1976-78 20.4 15.6 1979-81 23.4 26.5 1982-85 24.0 20.9 1986-87 21.9 16.4 1988-89 15.3 12.1 1990 16.0 15.6 1991 14.0 13.7 1992 14.6 13.1 Sources: Neyra, Hendrick, and Rossini (1990) and IFS. 1.15 From 1960 to 1971, the relationship between investment and domestic savings rates was close but, thereafter, was highly unstable.1' Gross domestic fixed investment is erratic, too (Figure 1.2). Total investment, however, is only one factor. The uses to which it is put are also important. So, what factors influenced its effectiveness, both on output and on the capital stock? 1/ The correlation coefficient was only 0.50 throughout the period, which according to the Feldstein-Horioka test of intemational capital mobility, does not indicate whether there was greater or less capital mobility in Peru.  Peru at the Crossroads - 10 - Building a Modem State ANNUAL CHANGE IN GROSS CAPITAL FORMATION 24 10- 0 -10 -20 -301 60 62 64 66 6B 70 72 74 76 78 80 82 84 86 88 90 Figure 1.2 Sources: Webb and Baca (1991). 1.16 The composition of gross fixed investment changed markedly in 1960-90. Machinery and equipment accounted for 40.2 percent of the total in 1960, rising to 55.1 percent in 1975. Thereafter, it fell to 46.8 percent in 1980 and to a very low 26.8 percent in 1990. Some economists have found a higher positive correlation between investment in equipment and machinery (especially if transportation equipment is excluded) and growth, than between total investment and growth (De Long and Summers, 1991). This suggests that besides the decline in the overall rate of capital formation in Peru, there was also a big fall in the formation of that component which could be most important for growth (see Appendix 1A).  - 11 - Economic Growth 1960-90 1.17 Why did the composition change? One factor is political violence but changes in the relative prices of capital goods may also have played a part. The price of capital goods in relation to Peru's per-capita income was one of the highest in the world (De Long and Summers, 1991). This seems to indicate that Peru needs a higher rate of investment than other countries to obtain the same amount of capital goods. This situation continued until 1986, when the relative price of capital goods began to decline, a trend that continued into 1990. 2. The External Sector 1.18 Two approaches can be used to quantify the external source of economic growth. The first (applied by Denison) is to evaluate the effect of increasing returns to scale. The second (by Maddison) proposes a measurement of the external sector "input" and a parameter for measuring its impact. The "input" is measured by taling the average growth of exports and imports multiplied by the average shares of exports and imports in GDP. The value of the parameter is selected conceptually, depending on whether the period at issue is one in which foreign trade would have had a greater, or lesser, impact on the country's economic activity. Foreign trade 'input" measures in part the openness of the economy, and also indicates the increase in the market size. Other indicators of openness are levels of protection - that is, tariffs and nontariff barriers. 1.19 Studies by De Gregorio (199la) and Easterly (1990) indicate that Peru was one of the least open countries - that is, most oriented to the domestic market. Exports and imports as a share of Peru's GDP climbed from 1950 to 1967, after which the share began to decline (though with major fluctuations), a pattern which mirrors changes in investment and in the investment-domestic savings ratio. 1.20 The external sector affects growth through international movements in inputs and capital. The mobility of inputs is easily incorporated into the methodology of growth sources but capital mobility is not. Its impact is analyzed using another conceptual framework. 1.21 Labor input (quantity and quality) already takes into account international migration, although, perhaps, the rate of change in the quantity in the past decade is overestimated. International migration from Peru increased substantially, from an annual 0.54 percent in 1980-85, to 1.21 percent a year in 1985-90. 1.22 Some studies indicate that Peru experienced the same problems as most Latin American countries when there were large capital inflows, including a fall in the real exchange rate (Calvo, Leiderman, and Reinhart 1992a and 1992b). Many countries were part of what became known as the debt crisis. Peru was one of them. And several studies by the IMF indicated greater problems with productivity and growth in countries which had greatly increased external debt.  Peru at the Crossroads - 12 - Building a Modem State Table 1.8: PERU'S EXTERNAL DEBT, 1970-90 1970 1975 1980 1985 1990 External Debt (millions of US$) 3.681 6.257 9.595 13.721 17.347 External Debt/Export Ratio 3.56 4.70 2.45 4.61 5.29 External Debt/GDP Ratio 0.51 0.38 0.46 0.76 0.40 Sources: Webb and Baca (1991). D. Income Distribution 1.23 In 1960-75 (the growth period), Peru's real unit wages grew at 1.58 percent a year, much faster than the growth in total factor productivity (0.50 percent). Growth in the quality of labor input was 1.92 percent, resulting in a real increase in the price of input services of around 0.672 percent a year. This makes a total of 1.172 percent a year, still less than the growth in wages and indicating redistribution in favor of labor. 1.24 During the recession (1975 to 1990), real unit wages fell by 9.21 percent a year and by 10.93 percent in the 1980's)'. This is much faster than the decrease in total productivity (3.3 percent annually), a redistribution away from labor. The wage share of GDP increased from 35.4 to 37.3 percent in 1960-75, but plunged to 17.1 percent (24.9 percent in constant prices) in 1990. 1.25 Real income in the informal sector almost doubled from 1976 to 1987 but fell sharply in 1987-90 - by more than over 50 percent (December to December) and 36 percent using annual averages. This brought real income almost back to 1976 values. If much of this income is wages, the net effect should be a fall in income distribution less severe than the drop in the labor input. 1.26 There are fewer statistics on the personal distribution of income, and they are not particularly comparable. In 1961, the poorest 20 percent of the population had 2.4 percent of total income, and the richest 20 percent had 66.8 percent (Webb and Fernandez Baca, 1991). In 1985, the poorest 20 percent had 4.4 percent and the richest 20 percent had 2/ Other studies estimate a 5.67 percent decline for the 1980's (see Altimir, 1992).  - 13 - Economic Growth 1960-90 51.9 percent. In other words, income distribution improved somewhat between 1961 and 1985. 1.27 Critical poverty indicators, covering those with the lowest personal income, improved between 1985 and 1987. The poverty index declined from 100 to 76. There was, however, a sharp deterioration to 155.5 in 1990, taking 1985 as the base year (Webb and Fernandez Baca, 1991). E. Economic Policies: Impact and Prospects 1.28 Since the early 1960's, Peru first saw a period of economic growth (in the first 15 years), followed by a recession in the past 15 years. Although Peru was exposed to phenomena similar to those experienced by its neighbors and trading partners, its economic performances in the past 15 years was one of the poorest. This suggests that internal factors were more important than external factors. During the first half of the recession, the major factor was higher oil prices, which affected the whole of the economy; in the second half, it was political unrest, which reduced net investment, as well as its effectiveness, even in the informal sector. 1.29 Inflation rose after 1980. Real bank credit to the private sector grew during the first half of the 1980's, and then began to decline rapidly; real interest rates were frequently negative. Both could have had an adverse impact on capital formation and on its efficient use. 1.30 Although Peru's public sector is not large compared with some other Latin American countries, employment in the sector grew fast in the 1980's. That was partly responsible for the decline in total factor productivity, although the composition of the public sector (administration and enterprises) did not change substantially. Judging from the composition of central government tax revenue, however, tax policy changed, which had some effect on investment, although not as much as expected. These effects, however, could have more of an impact in the 1990's. The external sector also influenced investment, through the debt crisis; and, ratios of extemal debt to exports and to GDP indicate that this could continue to be an important factor. 1.31 Opening up the Peruvian economy was limited by commercial policies; the external sector made a substantial contribution to Peru's growth in 1960-75, but that waned during the recession. This pattern must be reversed. Changes in the external sector also help to explain the shift in composition of investment, particularly the fall in the equipment and machinery component in the 1980's. 1.32 The lowest income earners may have improved their position in the personal distribution of income in the 1960-75, but that seems to have been reversed in the past five  Peru at the Crossroads - 14 - Building a Modem State years. Moreover, urban employment, which increased in the first half of the 1980's, fell in the past five years. There were, however, no major regional changes. 1.33 During the recession, Peru's resource base (labor and capital) increased by 67 percent, but that was not reflected in an increase in GDP. Much of these resources were not used efficiently. Improvements in efficiency in a new expansionary phase of, say, five years might lead to annual growth of 5.9 percent, but that source of growth would then be exhausted. 1.34 The quality of Peru's labor increased significantly over the last three decades. Thus, if there is economic expansion, the impact of labor quality could be strong. The relative fall in the prices of capital goods over the past five years is another favorable sign. An improvement in capital formation and policies leading to a better use of resources will go a long way toward restoring Peru's economic performance.   - 15 - The Economic Costs of Terrorism CHAPTER II THE ECONOMIC COSTS OF TERRORISM A. Introduction 2.1 Most analysts of Peru agree that terrorism has been a huge barrier to economic growth. Moreover, it has been estimated that the wealth loss from terrorism is US$25 billion. Yet there are no studies explaining why or how terrorism affects growth.1' 2.2 So, what are the consequences of terrorism for the allocation of resources, economic growth, and poverty? And what are the implications of the relative scarcity of resources and budgetary restrictions for the terrorists, the public, and the state of Peru? This economic approach to the analysis of terrorism does not ignore the insights provided by other studies and approaches. Rather, it highlights issues that have not been studied in a systematic way and suggests an agenda for the analysis of these issues. B. Stylized Facts 2.3 The number of terrorist acts per capita increased rapidly until 1986 and remained roughly constant until 1991. Deaths per capita, on the other hand, increased substantially after 1987. The differences reveal mostly an increase in repression during the decade. The number of electricity transmission towers damaged or destroyed by terrorists in the 1980s' followed a similar pattern to all terrorist acts until 1985, but grew at a faster rate after. 1/ Most of the work is descriptive or anecdotal. An exception to this critique is McCormick (1991) and (1992)).  Peru at the Crossroads -16 - Building a Modern State Tarrist AMU and Deisa Frm V_hmc (per Capka) io . " 0.10 IA . *~~~~~~~~..... . so SI U2 i4sO s;7N U Temorbt AU Vhok. Deat Figure 2.1 Uestaicity Trnmnaission Tor. Damxapd by Twrwints 300 250 200. SO SI 82 03 B5 is 87 I' I690 Filgure 2.2  - 17 - The Economic Costs of Terrorism 2.4 Terrorist acts varied both across regions and over time but the variability (or dispersion) declined over the decade (see Figure 2.3). Until 1984, most terrorist activity was concentrated in Ayacucho; since then, activity has spread to the rest of the country and the relative importance of Ayacucho has declined (see Figure 2.4). After 1986, there was a significant increase in terrorist activity in Lima and the surrounding departments.' Table 2.1: TERRORIST ACTS BY DEPARTMENTS (per 100 people) 1980 1991 1982 1983 1984 1985 1986 1987 1988 1989 1990 1 Loreto 1.2 2.1 0.6 0.5 0.2 3.1 0.0 1.8 0.5 2.8 1.8 2 Huaco 0.0 0.4 0.6 2.4 14.5 15.1 9.7 17.2 20.1 17.5 16.4 3 Juina 3.5 6.0 4.6 7.3 8.8 17.8 16.9 23.7 31.3 55.9 34.5 4 Pasco 1.8 6.9 8.0 17.2 35.5 50.9 42.8 48.7 36.4 49.2 20.9 S Arequipa 2.1 3.0 3.2 4.0 5.2 4.3 11.8 9.5 9.2 5.9 5.8 6 Auash 1.2 3.4 4.4 2.7 1.7 7.4 5.7 12.2 10.0 24.8 18.5 7 FPra 0.1 0.3 0.0 0.3 2.2 0.6 1.3 1.9 1.9 2.9 2.5 U Tunbes 0.0 0.0 0.0 0.0 0.8 1.6 0.8 8.5 10.6 4.4 2.1 9 Apurunac 2.6 4.9 15.2 3.4 6.5 4.8 38.5 29.4 38.1 18.4 6.7 10 Cusco 0.9 8.5 3.2 3.5 3.2 7.2 7.1 4.8 3.5 2.9 5.2 11 Madre de Dios 0.0 0.0 2.7 0.0 2.5 0.0 2.3 4.5 0.0 0.0 6.1 12 Ayacucho 9.3 28.5 60.6 85.3 120.2 65.7 63.6 71.8 68.6 66.3 53.8 13 Huanavlsca 2.5 1.9 10.6 21.2 49.5 44.0 29.8 18.2 29.6 42.7 33.0 14 Ica 0.0 3.1 4.2 1.5 4.9 3.3 2.9 4.6 12.7 16.6 9.1 13 Moquegua 1.0 13.6 1.9 0.0 0.9 0.0 0.0 0.0 0.8 0.8 0.0 16 PUDo 0.3 2.6 1.3 2.7 6.2 6.6 28.5 6.4 5.2 10.2 10.3 17 Taca 2.1 13.6 11.1 1.9 0.0 2.3 2.3 3.8 3.7 11.7 4.4 18 Amazoum 0.0 2.2 0.7 0.7 0.3 0.0 0.6 0.0 0.3 1.2 0.6 19 Carjaarc 0.5 2.4 3.0 1.8 1.3 1.8 2.1 8.3 5.6 9.6 4.3 20 Laubayeque 1.7 2.1 3.5 2.0 2.8 8.7 8.2 5.1 8.9 6.0 8.1 21 La Labtad 0.6 1.7 2.9 2.2 9.0 14.0 11.3 10.9 7.3 8.6 15.1 22 SanMartin 0.0 0.0 0.0 1.1 2.9 2.1 5.7 15.6 19.2 17.2 24.1 23 Ucayasi 5.8 1.1 0.0 0.5 6.6 0.5 0.0 12.2 12.3 47.6 14.3 24 limay Callao 0.7 3.5 3.2 4.5 5.0 9.8 13.5 11.0 7.9 9.6 12.1 Source: Peru en Nuneros 1991, Cuanto. _/ Figure 2.3 portrays the distribution of terrorist actions per capita over the 1980s; XO, X2, X4, X6 X8 and X1O portray the histogram corresponding to 1980, 1982, 1984, 1986 and 1990 respectively. 3/ Figure 2.4 portrays the number of terrorist actions per capita per department. The horizontal axis depicts Peru's 24 political departments according to the order presented in Table 2.1. The vertical axis portrays terrorist actions per capita. XO, X2, X4, X6, X8, X1O represent 1980, 1982, 1984, 1986, 1988, and 1990 respectively.  Peru at the Crossroads - 18 - Building a Modem State Distribution of Terrorist Actions Over the Decade Histogr ams EM Au~ ~ u 'a L U ma ~ ~ ~ ~ ~ ~ M l 2.3 I U n U 7 D U 11* i 3 4 W *_ L * L.a is ma 4L * 'a Egure ~ ur 2.3  0 3 a E3 C) t,t ,t a a| 0 C.) C) S '.4 ..IMI~~~~~~~~~~~a ............. ,o om [ i[ .IIoa S -t '.4 La) q we >~~ [: 2 : =2:  Peru at the Crossroads - 20 - Building a Modern State 2.5 The upward trend in terrorist activity does not necessarily imply that terrorist organizations have steadily increased their power over the past 10 years. An increase in terrorism will have different implications, depending on the importance of the area in which it takes place. Nor does the high initial incidence of terrorist activity in the highlands (one of the poorest regions in Peru) indicate a simple direct link between poverty and terrorism. Poverty has been part of highlands society for centuries; terrorism is a recent phenomenon. Moreover, extremely poor areas, like Loreto or Tumbes, are relatively calm. Rather, this pattern suggests that the absence of adequate law enforcement is the main cause of terrorism. That, in turn, is partly explained by the rough terrain, and consequent high costs of enforcing the law in the highlands. 2.6 Before land reform, there was an institutional arrangement ("gamonalismo") for law enforcement. The gamonal system was like a feudal regime. It implied protection in exchange for services between landowners and peasants. Despite the abuse of peasants, this regime compensated for the absence of modern institutions in these areas. With land reform came the destruction of the gamonal regime. That left a power vacuum, which was exploited by terrorist organizations. 2.7 The speed and scope of penetration of terrorist activity in the highlands in the 1980's mirrored the Peruvian state's inability to provide basic services, such as security and administration of justice, in these regions. There is no information available about spending on these specific services for each department, but it was probably similar to expenditure on other public services. Thus the observed correlation between indicators of public health and nutrition services and terrorism may approximate that between local security and terrorism (see Figure 2.5). Even allowing for any shortcomings in this analysis, the scatter diagrams are consistent with the conjecture that terrorism has secured deeper roots in areas where the presence of the Peruvian state is weak.4'-' 41 Statistical correlation should not be interpreted as evidence of causality; however, economic theory allows to build hypotheses about causality which can be empirically tested and rejected. The graphs highlight a negative correlation between presence of the state and terrorism. The model presented in the next section and in Appendix 2A state that lower law enforcement results in higher terrorism because it reduces the expected costs for those participating in terrorism. I/ Other indicators of the presence of a modern state were studied. In all cases, the correlation between the number of terrorist actions and these indicators was the expected: negative with the number of physicians per inhabitant; negative with the number of nurses per inhabitant; negative with the number of hospital beds per inhabitant; and, negative with social security coverage.  - 21 - The Economic Costs of Terrorism Indicators of Regional Presence of Modern St ate Percentage or Illiterate in Total and Terrorist Actions per Inhabitant 0.25- 0205 + 0.15- + 0.10 0.105 0.00 . 0 25 50 75 XB Death Rate Among Noew Born and Terrorist Acts per Inhabitant 0.125- 0.100- lm4 + . 0.075- .F+ + 0.050 + 0 10 20 30 40 50 60 X10 F%ure 2.5  Peru at the Crossroads - 22 - Building a Modem State 2.8 That weak profile was not confined to highlands areas rife with terrorism. It was in the towns, too. After the reforms of the 1960's and 1970's, there was high migration from rural to urban areas. The population of Lima and its surroundings grew by 5.5 percent a year in 1961-72, and by 4.7 percent in 1961-81. Low economic growth aggravated the difficulties of urban Peru in absorbing the flow. Rural migrants swelled the ranks of the urban poor, and encouraged the expansion of the informal sector. Shanty towns migrants created their own rules and means of subsistence. They did not have the same access to public services as other residents and did not pay taxes. The result: decay in the quality of life in Lima and other urban areas and an increase in crime. 2.9 The growth in urban crime reveals deep changes in the society. Rural migrants did not abide by the rules of urban residents, and the absence of common precepts reinforced segregation. Crime spread from shanty towns to the rest of Lima, although all urban residents were not affected equally. Those living in high-income districts and large firms substituted private for public-sector security services, while residents of low-income districts and small firms had to adapt to lower security. 2.10 There are no systematic data at a micro level showing the economic importance of terrorism. The payment of bribes to terrorists, however, especially in the agricultural and mining sectors, was widespread. Nationwide there are 429 formal and specialized security-services firms, employing roughly 20,000 people and chalking up sales of US$30 million a year. 2.11 There are no systematic data either about the share of security services in total production costs across sectors and locations, but some firms spend substantial sums on security. For example, in 1991 one palm tree plantation spent 4 percent of the value of sales (6.1 percent of value added) on security services and a storage firm in Lima spent 1 percent of the value of sales (1.4 percent of value added). A food producer in Lima had security costs equivalent to 0.4 percent of the value of sales, or 2.7 percent of value added. The consequences of terrorism for individuals or firms have differed depending on location, type of activity, and revenue. C. A Framework to Analyze Terrorism 2.12 Most studies of political violence in Peru spend much time and effort on the ideological origins of terrorist organizations, and comparing their actions with other terrorist organizations. Yet a simple fact is usually overlooked: if no resources are allocated to the production of terrorist acts (or to their repression) the existence of terrorism (or its repression) is inconceivable. Relative scarcity of resources restricts behavior.  - 23 - The Economic Costs of Terrorism 2.13 Whatever the ideology, basic organizational principles will induce terrorist organizations to economize on resources that are relatively scarce (and hence expensive) and try to maximize results, probably in terms of power, from terrorist actions. Budget restrictions and relatively scarce resources also affect law enforcement. In general, no society will find it optimal to allocate resources for law enforcement sufficient to drive crime to zero. It is simply too expensive to eliminate all crime.