Report No. 33501-CL Chile Development Policy Review (In Two Volumes) Volume II: Main Report June 2006 Poverty Reduction and Economic Management Unit Latin America and the Caribbean Region Document of the World Bank Program for International Student Assessment Chile's SocialRsk Management Survey SENCE Subsidies for Employment Creation SIMCE Sistema de Medici6n de la Calidad de la Educaci6n SME Small and Medium Enterprise TFP Total Factor Productivity WB World Bank UK United Kingdom USD United SatesDollar VAT Value Added Tax V l T TechnicalResearch Center of Finland r Vice President: Pamela Cox Country Director: Axel van Trotsenburg Sector Director: ErnestoMay Sector Manager: Mauricio Carrizosa Task Managers: Daniel Oks and James Parks TABLE CONTENTS OF ACKNOWLEDGEMENTS ...........................................................................................................I PREFACE .....................................................................................................................................I1 INTRODUCTION .........................................................................................................................1 Strategic Setting of the DPR........................................................................................................ 1 Matrix 1:Remaining Challenges and Policy Options .................................................................3 1 THE ROAD TO CONVERGENCE FOR CHILE . .............................................................6 Introduction................................................................................................................................. 6 Growth in Chile........................................................................................................................... 7 Slowdown 1998-2003 ...........................................................................................................11 Poverty and Social SectorOutcomes in Chile...........................................................................14 Pro-poor Growth ...................................................................................................................16 Impact of Labor Market on Poverty......................................................................................17 Poverty Inlpact of Government Policies...............................................................................23 Inequalityin Chile................................................................................................................ 2 4 Conclusion:Development challenges ......................................................................................2 7 2 PROMOTINGSUSTAINABLEPRODUCTIVITY GROWTH . ..................................29 Introduction.............................................................................................................................. 2 9 Business Environment and the Dynamism of Firms ................................................................30 Sustaining.theDynamism of Firms during Shocks.................................................................. 33 Reducing the Cost of Reallocating Factors: Labor Flexibility and Access to Credit............35 Facilitatingthe Redeployment of Labor: Job Security, Minimum Wage and Wage Indexation ......................................................................................................................... 37 ImprovingAccess to Credit ...................................................................................................40 Institutionsto Guarantee Competition.......................................................................................44 Conclusions and Recommendations.......................................................................................... 45 Annex 2.1. The Dynamics of Firms-Selected Results............................................................48 Annex 2.2 Investment Climate Assessment ..............................................................................50 Annex 2.3 Insolvency and Creditor Rights ............................................................................... 51 3 INNOVATION . .........................................................................................................................54 Introduction............................................................................................................................... 54 Does Chile have an InnovationProblem? ................................................................................. 54 What type of innovation does Chile lack? ............................................................................ 57 Summingup: The need for a comprehensive view ...............................................................66 InnovationPolicy and Reform in Chile..................................................................................... 67 Addressingthe innovationproblem .....................................................................................6 8 Basic Tasks of the NIS.......................................................................................................... 68 Diagnosis of Chile's NIS and the Need for Reform.............................................................. 69 RationalizingNIS Institutions:a Micro View ...................................................................... 72 Universities........................................................................................................................... 73 StrengtheningUniversity-PrivateSector Collaboration........................................................75 Recent University Developments in Chile............................................................................ 77 Public Research Institutes (PRIs)..........................................................................................79 The Private Sector as a Leader of Innovation....................................................................... 81 Beyond S&TProjects............................................................................................................84 4 EDUCATION . ..........................................................................................................................91 Introduction...............................................................................................................................91 Investmentin Education........................................................................................................ 91 Years of Schoolingand Coverage of Education ................................................................... 94 Education Inequality.............................................................................................................97 Higher Education: Expansion. Quality and Equity .............................................................101 The Quality of Primary and Secondary Education.............................................................. 102 The EducationMarket in Chile ...............................................................................................110 The Role of the State in EnsuringQuality and Equality of Educational Opportunity............113 IncreasingAccountabilityfor Public Resources.................................................................113 Ensuring equity of educationalopportunity........................................................................ 120 Providing Information to Improve the Functioningof the Education Market ....................122 The Evaluationof Education Policies and Programs.......................................................... 123 Summary of Key Policy Messages and Optionsto Raise Education Quality and Equity in Chile................................................................................................................................ 123 Appendix: Example of School QualityInformation Provided to the Public ........................... 125 5 ENHANCINGSOCIAL PROTECTION: STRIKING THE BALANCE BETWEEN . INCOME SECURITY AND ECONOMIC OPPORTUNITY ........................................127 Changes in the Risks to Household Incomes and the Coverage of Social Protection............. 129 The PrincipalRisks to Household Income and Chile's SocialProtectionPolicies.................142 Job Loss. Employment and .theLabor Market .................................................................... 142 The ImpoverishingCost of Health Care and Financial Protectionin Chile........................ 150 The Risks to Income for the Elderly and .theMulti-PillarPension Systemof Chile........... 154 The ChallengesAhead For Social Protection Policiesin Chile .............................................. 161 Covering Job-loss................................................................................................................ 161 Coveringthe Cost of Health Care.......................................................................................165 Income Security for the Elderly.......................................................................................... 166 Options for Policy Improvementsin Chile.............................................................................. 169 6 FINANCING DEVELOPMENTIN CHILE: ISSUES AND POLICY OPTIONS . .........171 Introduction.............................................................................................................................171 Public expenditure:structure,efficiency and long term scenarios........................................172 Public Expenditureand Growth..........................................................................................172 Public Expenditure- Structureand Efficiency................................................................... 174 The Changing Structureof Chile's Public Expenditure...................................................... 174 Chile Ranks High by InternationalStandardsin Terms of Social Spending ......................175 Chile has Improved the Targetingof SocialExpenditures,but there is Scope for Further Improvement...................................................................................................................176 There Is Also Scope for Reducing Inefficienciesin Health and Justice Expenditures ....... 177 Research and Development(R&D) is One Sector Where There Is Scope for Greater Private Sector Involvement.PrimarilyThrough Public-Private Partnerships...............178 Public expenditure.Long-term Scenarios.......................................................................... 179 Public Expenditurewill come under Pressure from Demographic Factors. Greater Demands for Equity and New Public Goods.................................................................. 179 DevelopmentFinancingNeeds of Chile -An InternationalPerspective............................ 182 Financing development-Optionsto improve .theefficiencyand equity of the tax system....185 Chile's Tax System............................................................................................................. 185 Tax Avoidance and Tax Evasion ........................................................................................189 Avoidance and Tax Progressiveness................................................................................... 191 The Differencebetween Corporateand Personal Income Tax Rates.................................. 194 Policy Options to Improve the Efficiency and Equity of the Tax System.......................... 195 Financing development-Fiscal rule, debt. public investmentand internationalrisk diversification...................................................................................................................... 197 Chile's Fiscal Rule -Performance. Shortcomings and PossibleTechnicalImprovenients 198 Longer Term Issues and Options ........................................................................................ 200 Main Conclusions.................................................................................................................... 202 Annex 6.1:Tax Reform Options .............................................................................................204 7 PROSPECTSFOR ACHIEVING CONVERGENCE AND REDUCING . VULNERABILITY ................................................................................................................20 5 Medium-term Outlook............................................................................................................. 205 Risks to Convergence.............................................................................................................. 206 SuboptimalGrowth andforVolatile Growth.......................................................................206 Adverse External Conditions..............................................................................................208 Policy Options for Rapid Growth with Equity .An integrated framework........................215 Agenda for Future Research................................................................................................... 2 16 STATISTICALANNEX ........................................................................................................... 219 REFERENCES ......................................................................................................................... 237 Figure 1.1: Economic Stabilizationin Chile.................................................................................... 7 Figure 1.2. Capital Flows and Real Exchange Rate. 1980-2004 .....................................................9 Figure 1.3:The Contributionof Investment. Domestic and External Demand to GDP. 1997-2004 ................................................................................................................... 12 Figure 1.4.Chile Real GDP (Percentagedeviation from trend).................................................... 12 Figure 1.5. Poverty Headcount Ratio: LAC countries. Late 1990s.early 2000s........................... 15 Figure 1.6.Persons Active. Employed and Unemployed. 1990-2004 ........................................ 17 Figure 1.7. Employment Growth and Unemploymentin Chile. 1990-2004 ............................... 18 Figure 1.8.Proportion of Low SkilledWorkers by Sectorin Chile. 1990and 2003..................... 19 Figure 1.9. Distribution of Indigent and Poor by Sectorin Chile. 1990and 2003........................19 Figure 1.10. Chile is One of the Most Unequal Countries in the World .......................................25 Figure 2.1. Number of Firms by Size. and GDP............................................................................31 Figure 2.2. Employment:Employeesby Size of Fimi .................................................................. 31 Figure 2.3. MicroeconomicFlexibility in Chile ............................................................................37 Figure 2.4. Minimum Wages relative to the Wages of Unskilled Workers...................................39 Figure 2.5. Bank Credit to Private Sector: Stocks and Flows........................................................ 43 Figure 2.6. Bank Commercial........................................................................................................ 44 Figure 2.7. Bank Commercial Loans............................................................................................. 44 Figure 3.1: "High tech" success stories:Deviations from predicted values for 4 indicators of technological advancement...................................................................................................59 Figure 3.2. The Latin model .......................................................................................................... 60 Figure 3.3:Advancedresource-abundantcountries....................................................................... 61 Figure 3.4: ManufacturingEstablishmentsDeclaring that they have Made Innovations in aspects of production and marketing................................................................................................. 63 Figure 3.5:Enterpriseswith successfulinnovation ....................................................................... 63 Figure 3.6. Innovation effort.........................................................................................................6 4 Figure 3.7. Total innovation expendituresand total turnover........................................................ 64 Figure 3.8. In-house Expenditure on R&Dyas a share of total turnover .......................................64 Figure 3.9: Basic components of a Core NIS and "extended" policy framework for innovation..67 Figure 3.10. Present organizationof support for innovation.........................................................70 Figure 3.11:Possible alternative structureof support for innovation............................................71 Figure 3.12. Publicationsper million people relative to GDP per capita. 1999............................74 Figure 3.13: Scientificarticles per 100FTE researchers. 2000.....................................................74 Figure 3.14. Research citations. 1990and 1999............................................................................ 75 Figure 3.15. Financial Structureof Tech Centers in Europe .........................................................88 Figure 4.1:Total EducationExpendituresrelative to GDP. 2000 ................................................9 2 Figure 4.2. Monthly Salariesof Full-Time Teachers (44 houriweek) ........................................94 Figure 4.3: Years of Schoolingby Income Decile. Adults Aged 25 years or More. 1992-2002 ..95 Figure 4.4. Enrollment in Primary and Secondary Education. 1982-2003....................................95 Figure 4.5. Dropout Rates in Primary and SecondaryEducation. 1991-2002 ..............................96 Figure 4.6. Enrollmentby Type of School. 1981-2002 ................................................................96 Figure 4.7. Enrollment by Age. Chile and Selected Other Countries........................................97 Figure4.8. Pre-schoolEnrollment by Income Quintile.................................................................98 Figure 4.9. Secondary Education Enrollment by Income Quintile................................................ 98 Figure 4.10. Higlier Education Enrollment by Income Quintile....................................................98 Figure4.11:PISA 2000 Average Mathematics Scoresby Country ............................................ 103 Figure 4.12. PISA 2000 Average Reading Scores. by Country................................................... 104 Figure 4.13. PISA 2000 Math Scores and GDP per capita. by Country................................. 104 Figure 4.14. PISA 2000 Math Scores and Expenditures per Pupil. by Country.......................... 105 Figure 4.15: Distribution of Studentsby Level of Performance accordingto PISA 2000 reading scores. by Country.................................................................................................. 106 Figure 4.16. Eighth-Grade SIMCE Scoresand Vulnerability Index. 2000 ................................. 107 Figure 5.1:Rate of SustainedAchievement of Minimum Conditions. and Relapse ................... 141 Figure 5.2. Minimum Cost of Labor in Chile and OECD Countries. 2002*............................... 143 Figure 5.3. Changes in the Minimum Cost of Labor in Chile. 1993-2004.................................. 144 Figure 5.4. Unemploymentand Duration of Job Searchin Chile. 1964-2004.......................... 145 Figure 5.5. Coverage and Targetingof Public Employment Programs in Chile......................... 149 Figure 5.6: Out-of-PocketHealth Expenditureas a Percentage of Total Health Expenditures ..151 Figure 5.7: Percentage of Households Falling Below the Poverty Line Due to the Cost of Health Care. by Household Income Quintile ...................................................................... 152 Figure 5.8: Average Replacement Rates from Mandatory Earnings-RelatedPension Plans. Chile and SelectedOECD Countries.................................................................................. 156 Figure 5.9: Minimum Pensions Guaranteed by the Mandatory Pension System. Chile and Selected OECD Countries................................................................................................... 157 Figure 5.10: Earnings Subjectto Mandatory Contributionsin Chile and SelectedOECD and Latin American Countries................................................................................................... 169 Figure 6.1: Share of Expenditures in GDP .................................................................................. 174 Figure 6.2. Real Expenditures: 1990-2003 ............................................................................... 175 Figure 6.3. Total Expendituresand SocialExpenditures for SelectedCountries........................ 175 Figure 6.4. Expendituresin Health and Education for SelectedCountries ................................. 176 Figure 6.5: Targeting of Social Expenditures Shareof Public Subsidies Received. by Income Quintile. 1990.2000 and 2003............................................................................................ 177 Figure 6.6. Expendituresin Research and Development in selected Countries .......................... 178 Figure 6.7. Age Distributionof Population in Chile................................................................... 179 Figure 6.8. International Comparisons-Distributionof Public Expenditure............................. 182 Figure 6.9. GDP growth in selected comparatorcountries.......................................................... 183 Figure 6.10. Total Public Expenditures of Selected Countries(index at time t=100) ................. 184 Figure 6.11:Total Public Expenditures of Selected Countries.................................................... 184 Figure 6.12. Tax Ratios and GDP per capita ............................................................................... 186 Figure 6.13. Revenuesby Tax ..................................................................................................... 186 Figure 6.14. Foreign Direct Investment under D.L. 600 ............................................................. 187 Figure 6.15. SIIBudget ............................................................................................................... 188 Figure 6.16. VAT EvasionRate .................................................................................................. 190 Figure 6.17. Average Tax Rates Gap Between Second CategoryTax and Surtax......................191 Figure 6.18. Tax Progressiveness (SII. 2003).............................................................................. 192 Figure 6.19. Average EffectiveTax Rate. 2004 .......................................................................... 192 Figure 6.20: Average EffectiveTax Rates (with and without credits. deductionsand exemptions)......................................................................................................................... 193 Figure 7.1:Alternative Possible Growthpaths of GDP per capita in Chile. 2004-2030 ............207 Figure 7.2. GDP Growth and External Conditions. 1971-2003 .................................................208 Figure 7.3: SovereignRisk: EMBI Spreads2000 -2005 Overall Spreadsand Correlation with Latin Composite................................................................................................................2 1 0 Figure 7.4. Real Copper Prices and GDP growth 1960-2004 .....................................................211 Figure 7.5. Chile: Response to a Copper Price Shock.................................................................212 Table 1.1.Chile: GDP GrowthIncreased Sharply after 1984......................................................... 8 Table 1.2. Chile and Comparator Countries: Annual Real GDP GrowthRates............................ 11 Table 1.3:Measures of TFP Growth ............................................................................................. 14 Table 1.4.Poverty Headcount: Chile. 1990-2003 ......................................................................... 15 Table 1.5:Pro-poor Growth and Progressive GrowthMeasures, Chile. 1990-2003 .................... 16 Table 1.6. Share of Adults in the Labor Force. Chile. 1990-2003 ................................................ 17 Table 1.7.EmploymentDistributionby Economic Sector. 1990-2004 ........................................ 20 Table 1.8. Growthby Sector. 1990-2004 .................................................................................... 2 0 Table 1.9.Reduction in Poverty and Indigenceby Sector. 1990-2003 ........................................21 Table 1.10.Wages. Hours and Labor Income in Chile by Level of Education. 1990-2003 .........21 Table 1.11:Proportion of Employed. Unemployed and Inactive over 15who are Poor or Indigent. 1990-2003 ............................................................................................................2 2 Table 1.12.Years of Education for Adults aged 25-65 in Chile. 1990-2003 ................................ 24 Table 1.13:Distribution of Household Income per capita in Chile. 1990-2003 ...........................26 Table 1.14.Income Distributionby Decile in Chile. 1990-2003 ..................................................26 Table 1.15.Chile: Convergenceof GDP per capita with Spain .................................................... 27 Table 2.1: Chilean Firms by Size. Sales. Exports and Employmenta............................................30 Table 2.2: Cost of Starting and Closing a Firm. Job Security and Cost of Hiring and Firing Workers................................................................................................................................3 5 Table 2.3: Creditor Rights. Rule of Law. Days needed to Enforce a Contract. Days to Collect a Bounced Check. EffectiveCreditor Rights. and Volatilityof Credit.................................36 Table 2.4. Size of Financial Markets in SelectedCountries. 2003................................................41 Table 2.5. Size of Financial Markets in SelectedCountries. 2003................................................43 Table 3.1:Aggregate R&D investmentwhen OECD sector investment rates are applied to the Chilean economic structure (1995-99 average) .................................................................... 56 Table4.1:Total Expenditure on Education relative to GDP. 1990-2002 .....................................91 Table 4.2. Central GovernmentExpenditure per Student.............................................................. 93 Table 4.3:Enrollment by Level of Education. Region. and Income Quintile ............................. 100 Table 4.4. Enrollment in Higher Education by ............................................................................ 101 Table4.5. Average SIMCE Test Scores. 1992-2004 .................................................................. 103 Table4.6. Distribution of PISA 2000 Mathematics Scores. by Country................................. 106 Table4.7. Average SIMCETest Scoresby SchoolType............................................................ 109 Table4.8. SIMCE 8th Grade Test Results. 2004......................................................................... 120 Table 5.1:Main Indicators of Risk to Earnings Capacity and Household Income...................... 130 Table 5.2: Change in the Main Indicators of Risk to Earning Capacity and Household Income 131 Table 5.3: Coverage of Chile's Principal Social ProtectionPrograms and Interventions........... 135 Table 5.4: Change in Coverage of Chile's Principal Social Protection Programs and Interventions........................................................................................................................ 136 Table 5.5:ProgressAchieved by ChileSolidario........................................................................ 140 Table 5.6. Public Eniployment Programs in Chile ..................................................................... 148 Table 7.1 Chile: Projected Macroeconomic Indicators. 2005-2008 ............................................ 205 Table 7.2 Chile: Projected Year of Convergence with Spain......................................................206 Table 7.3. Structureof employment:Chile. 1990-2003 .............................................................217 Box 2.1: Chile.Job Security. Minimum Wage. and Wage Indexatiion........................................3 8 Box 3.1:An alternativeview of the TFP gap: microeconomic flexibilityand the costs of institutionaladjustment......................................................................................................... 58 Box 3.2. Discovery and Deepening............................................................................................... 72 Box 3.3. The Importanceof Collaborationbetween Universities and -thePrivate Sector............. 76 Box 3.4. Policies to link research and industry-The Mobility of Human Capital.......................79 Box 3.5. Market Failure and the Rationale for PRIs .....................................................................80 Box 3.6. Box: Clusteringin Finland..............................................................................................87 Box 3.7. Mobilizing SMEs:Technology Centersin Spain............................................................89 Box 4.1:How Finland Recruits and MotivatesTeachers ............................................................ 115 Box 4.2: Components of an Effective System for Attracting. Retaining and Motivating Highly QualifiedTeachers.................................................................................................. 118 Box 5.1:Risks to Income throughoutlife.................................................................................... 132 Box 5.2. Chile Solidario: ConfrontingIndigence........................................................................ 138 Box 5.3.Shiftingthe Focus of Employment Assistance:Recent Trends in OECD .....................164 Box 5.4: : Spain's ShiftFrom Bismarck to Beveridge:De-linlungHealth Coverage from Employnient Status............................................................................................................. 166 Box 6.1:Tax Avoidance through Incorporation.......................................................................... 191 Box 6.2. Reforming the Personal Income Tax -An Example..................................................... 194 Box 6.3. The Arithmetic of Public Investmenta...........................................................................202 ACKNOWLEDGEMENTS The Development Policy Review (DPR) was co-managed by Daniel Oks (Lead Economist, LCSPR) and James Parks (Sector Leader and Lead Economist, LCSPR). Emily Sinnott (LCC7A) put together the main report based on contributionsfrom a team of Bank staff and consultants followingmissions held during the second half of 2004. It draws upon discussions with a wide range of individuals in Government, think tanks, academia, consultants, and international organizations. It also draws upon material presented by experts at two workshops held in Santiago de Chile during November- December 2004. A list of the team members who wrote or contributed to the various chapters in Volume 2 followsbelow: Chapter 1: Development Challenges.Emily Sinnott (LCC7A) Chapter 2: Promoting SustainableProductivity Growth. Sara Calvo (LCSPE) Background papers: Alejandro Micco and'~1exGaletovic (consultants). Chapter 3: Innovation. William Maloney (LCRCE), Kristian Thorn (LCSHE) and Andres Zahler (consultant) Chapter 4:' Education. Emiliana Vegas (LCSHE) Chapter 5: SocialProtection. Truman Packard (LCSHD) Chapter 6: Public Finance Issues in Development.Daniel Oks (LCSPR) Staff Alvaro Vivanco (LCSPE) Background papers: Alvaro Vivanco (LCSPE), Claudio Agostini, Francesco Giavazzi, Roberto Perotti (consultants) Chapter 7: Risks to Convergence. Emily Sinnott (LCC7A) and James Parks (LCSPR) Background papers: Alvaro Vivanco (LCSPE),Francesco Giavazzi (consultant) Editor: Frank Earwaker The team would like to thank various Government officials including the Finance Minister-Mr. Nicolas Eyzaguirre-and the Budget Director-Mr. Mario Marcel-for their cooperation and assistance in identifying the agenda and obtaining the information. Sara Calvo (LCSPE) and Mr. Luis Escobar (Director at Ministry of Finance) led the organization of the workshops in Santiago. Valuable suggestions and support were received from Axel van Trotsenburg (LCC7C Director), Jesko Hentschel (Sector Leader LCSHD), Carter Brandon (Sector Leader, LCSES). The report also benefited from the able assistanceof Maritza Bojorge (LCC7A) and Maria Estrella (LCC7A). PREFACE The Chile Development Policy Review (DPR) covers economic developments in the country through early 2005. The Summary Report was sent to the previous Government in September 2005 and the Main Report (volume 2) was submitted to the new Government in March 2006. Following consultationswith the Government in June 2006, it was agreed to proceed with circulation of the report to the Board in grey cover in the present form. In addition, it was agreed to organize a seminar in Chile to disseminatethe report. The DPR complements the Policy Notes presented to the new administration in March 2006. Since the DPR was communicated to the authorities, the new Government, which took office in March 2006, has launched its development strategy. The new strategy reiterates Chile's commitment to the pursuit of prudent macroeconomic management, fbrther development of market institutions, innovation and strengthening of social policies and education. The new Government has strengthened the focus on enhancing equality of opportunitythrough education,pension reform and child development. It has appointed a Pension Commission, an Education Commission and a Child Development Commission, all charged with preparing policy proposals in consultation with stakeholders, including civil society. The Government has also moved to create an Innovation Council with representatives of 5 ministries and non-government stakeholders. All these areas - social security, education and innovation - are at the core of the DPR and the new Government's policies are fully consistent with the options presented in the report. The other key areas covered by the report - investment climate, labor markets, taxation and public finances - are also important elements of the new Governmelit's strategy. Therefore, the themes and policy options discussed in the DPR remain relevant for Chile today. Since the new Government has taken office, the main economic developments that have emerged are the further accumulation of fiscal surpluses stemming from copper-related revenues and growing demands for social expenditure. The Government has committed to continued full compliance with the fiscal rule and has prepared proposals to invest copper surpluses (in a pension fund that will cover State pension liabilities, in central bank recapitalization and in a Fund for Economic and Social Stability). Managing the copper surpluses and assessing the effectiveness and fiscal impact of new public spending initiatives will be key challenges for the Government. INTRODUCTION STRATEGICSETTING THE DPR OF 1. Chile has become a benchmark for reform in the LAC region. From 1990to 2004 it was the fastest growing economy in Latin America with an average annual growth rate of 4.1 percent per capita. At the same time, there has been a dramatic reduction in poverty, with poverty levels having more than halved since 1990.Much of this successis attributable to the comprehensive reform process carried out over the past three decades. 2. There is consensus in Chile on the need for further reforms to foster a more knowledge-based economy. That, in turn will ensure sustained high growth so that Chile may converge towards the advanced economies.At the same time, Chile remains a small, open economy, vulnerable to external shocks. The slowdown in growth between 1998 and 2003 underscored the need to further reduce vulnerability and increase the speed of adjustment to shocks. 3. Considerable gains have been made in the fight againstpoverty since the return to democracy in 1990. Yet, almost one in five of the population is still classified as poor. The degree of income inequality remains one of the highest in the world. Inresponse to these challenges, the Government seeks to reduce poverty yet f~lrtherand increase equalityof opportunity. 4. This development policy review (DPR) seeks to (i) present a comprehensive overview of development challenges, and (ii) address selected issues of importance for the attainment of high growth with equity. In particular, the areas of private sector developmeiit; innovation; education; social protection, and development financing receive detailed consideration. 5. The DPR is intended to contribute to the dialogue on policy and institutional priorities. It has been prepared at the request of the Chilean authorities. The focal areas of the report emerged from consultationswith the Govemment; including a preparatory brainstorming session with the Ministry of Finance and two workshops held in Chile during November and December 2004. 6. The report aims to present a clear picture of Chile's developmeiit challenges and to suggest policy options to address them. It concentrates on a few selected areas which the Govemment, the Bank and many stakeholders believe are crucial to the country's agenda for growth and social development. Other areas that are also key to sustained growth-like trade, financial markets and infi-astructure-in which Chile has already made important advances have been deliberately left out of the report. The report also attempts to identify important gaps in knowledge--e.g. determinants of informality-for future research. The DPR will serve to guide the World Bank's future policy dialogue in Chile. It covers many key areas of the World Bank's program in Chile and will be a key analyticalinput into the forthcoming country assistance strategy(CAS). 7. The main messages that emerge from this DPR have been set out in Volume I which also contains an Executive Summary. Volume I1presents the detailed analysis and recommendations for seven selected areas of special importance for the development agenda of Chile in the coming years. The following matrix summarizes key policy options for the achievement of high and equitable growth in Chile. MATRIX1:REMAININGCHALLENGES POLICYOPTIONS AND resources across firms component and increasing the unemployment component thus keeping neutral the change; and (ii) evaluate the feasibility of tying the level of minimum wages to observable variables such us unemployment. Evaluate alternative wage indexation options with the aim of making at least one wage component flexible. A first order of business is transparency of data in the private and public sector. Credit issues. Consider F S S A recommendations to increase efficiency and make the financial sector resilient to shocks. Consider recommendations of World Bank study on small and medium enterprises. Evaluate the feasibility of establishing MATRIX1:REMAININGCHALLENGES POLICY OPTIONS (CONT.) AND Strengthencapacity in Science and Options include measures to: (i) review the S&T dimensions of the National Innovation System-universities, public and Technology private research institutions and supporting institutions; (ii) define the mission of national universities and align policies to promote excellence and to foster collaboration with the private sector; (iii) evaluate ongoing programs for competitive funding and centers of excellence and strengthenthem if merited; (iv) strengthen incentives for outward orientation; and (v) provide better information on the certified quality of universities and on the achievements of their graduates in the labor market. PRIs would benefit from a global and individual evaluation to identify their current role in the NIS and to address market failures in the specificsectors they participatein. Address Educational and Human Consider measures to increase the number of PhDs and evaluate the extent to which their specialization is oriented towards Capital Shortages science and engineering. Evaluate current programs that foster collaborationbetween researchers and firms and facilitate linkages with private sectorthrough internships, mentoring, etc. EDUCATION Supervise and support schools to Under-performing schools would benefit from increased attention and supervision, while high-performing schools could achieve the expected results receive increased autonomy and decision-making authority over financial and human resources. Ensure equality of educational Consider introductionof a student subsidy differentiated by student income level. For quality and equitypurposes, schools opportunity that receive the differentialsubsidywould be subjectto eligibilityrequirements, would required to limit or eliminate unfair student selectionpractices, and would be held accountable for how the additional resources are employed. Improve the information available The State has an important role to play in ensuring that accurate, comparable information among schoolsis made publicly to parents and civil society on available in ways that are easy to understand and are timely. school quality Education policy and program Systematic evaluations of education policies and programs would enhance the quality of current initiatives as well as evaluation future reforms. In the project design stage, steps could be taken to collect informationon a comparator group as well as on the experimentalgroup that is affectedby the program. Attract and retain qualified Consider multi-year agreement with teachers to secure participation. teachers Evaluate improved incentives through changes in the salary structure. SOCIAL PROTECTION Social assistance Completea rigorous evaluation of ChileSolidario. Examine more proactively the role of OMIL injob search assistance, starting with the beneficiaries of Chile Solidario, and drawing from experiences of the UK and US. Income risks from health costs Evaluatepossible furthermoves toward financing the Plan Auge from general taxation. MAT^ 1:REMAINING CHALLENGES POLICY OPTIONS(CONT.) AND FINANCING DEVELOPMENT Identify fiscal impact of medium- Comprehensively evaluate the fiscal impact of the current and projected medium-and long-term demand for social and long-term development needs transfers, including contingent liabilities. Fine-tuning the tax system to Tax system: improve efficiency and equity Consider conducting a survey of tax compliance. Review the case for VAT exemptions and specialrates. Considereliminating high-distortionary taxes such as the stamp tax. Ensure that market prices reflect externalities by levying a specialtax on diesel cars. Consider reducing the (35percent) additional tax on dividends. Considerreducing the number of brackets and marginal tax rates. Consider eliminating some credits, deduction and exemptions to improve equity and simplifythe tax system. Improve implementation of fiscal Consider extension of the role of automatic stabilizers, improvementsin the consistencybetween forecasts of GDP and rule copper prices, and reduction of the variability of fiscal policy in any one fiscalyear. Long-term financial arrangements Considerestablishing a target for Governmentdebtrather than one derived fromthe fiscal deficit. Internationalcapital markets look at the stock of debt outstanding rather than the flow. Assess optionsin international capital markets-like equity swaps-to hedge risks that have hurt Chile's budget and economyin the past. 1. THE ROAD TO CONVERGENCEFOR CHILE 1.1 Chile's growth has been impressive: per capita income has almost doubled since 1990.' Whereas Chile had a per capita income equivalent to only 27 percent of incomes prevailing in high income OECD countries in 1990,that ratio rose to 35 percent by 2003. Hence, the gap is closing, although there is still a long way to go before convergencewith the OECD is complete. Good economic performance has been the result of a combination of factors including a strong institutional framework, structural reforms since the 1970s, supportive fiscal and monetary policies, trade liberalization and favorable external conditions. Economic stability today is characterized by single-digit inflation, a structural budget surplus, a strong financial sector, a low level of public debt and sovereign spreads equal to those enjoyed by advanced economies. 1.2 Strong growth, coupled with well-targeted social policies, has led to a sharp drop in poverty. The official poverty headcount has more than halved since 1990, falling from 38.6 percent in 1990to 18.8percent in 2003. However, income inequality still remains a problem. Inequality, as measured by the Gini coefficient, is 0.56 which is among the highest in the world. While the Government has successfully established mechanisms to cushion the impact of shocks on incomes and on human capital,those households close to the poverty line still remain vulnerable. 1.3 The exogenous shocks provoked initially by the Asian and Russian crises and thereafter by the slowdown in the US economy had a much less serious impact on output and growth in Chile over 1998-2003than did comparable shocks in the past. That points to greater economic resilience on the part of the economy which, in turn, is due to strong economic management and purposeful policy reform. Nevertheless, as a small open economy still dependent on primary commodities, Chile remains vulnerable to the business cycle as evidenced by the recent downturn. Hence, there is a need to further build on the record of success, particularlywith regard to speedy adjustment. In addition, the pursuit of sustained high growth calls for a continued focus on competitivenessin the world economy,into which Chile has so successfullypropelled itself. 1.4 The chapter begins by outlining the development performance of Chile focusing, in particular, on the period since democracy was restored in 1990. Section A examines the growth record, followed by Section Bywhich analyzes poverty and social sector issues. The chapter concludes with section C which gives an overview of the developmentchallenges that Chile faces to reach its goal of high growthwith equity. 'InPPP terms, 2004 income was approximately 1.8times the 1990level. 6 Figure 1.1: Economic Stabilizationin Chile (i) Price Stability:Inflation has been reduced from 27.3 percent in 1990to 2.4 percent in 2004 - h S 30 E 25 \ u .ti 20 - 15 - 8 10 \ ci 5 4 0 - , , , , , I I 1 I I I I I ci a" 1990 1993 1996 1999 2002 Source: InstitutoNacional de Estadisticas (INE), Chile. (ii) Low public debt: Central Governmentdebt fell from 46 percent of GDP in 1990to 11percent in 2004 I I 1990 1993 1996 1999 2002 External debt 0 debt -Central Domestic government debt I I Source: Ministry of Finance (DIPRES), Chile; and Bank staff estimates (for 2004). (iii)Sovereign spreads are low I I Source: JP Morgan, 2005. 1.5 For most of the twentieth century the Chilean economy grew only slowly. From 1900 to 1984 growth in GDP per capita averaged just 0.9 percent a year. Moreover, the economy was subject to a series of strong fluctuationsin the first part of the 20th century. The most critical of these were the crisis in the nitrate industry in the 1910s, and the mining income collapse precipitated by the Great Depression in the 1930s. More recently, substantial contractioiis in the economy were caused by the economic and political turmoil of 1972-73, followed by the oil and copper crisis of 1975, and the debt and banking crisis of 1982. Table 1.1: Chile: GDP Growth Increased Sharply after 1984 Growth GDP GDP per capita 1811-1878 2.7 0.8 Note: * The rate of growth of per capita GDP from 1932to 1950was 3.9 percent. From 1930to 1932,Chile's GDP per capita decreased by close to 50 percent as mining income from copper and nitrates declined due to the Great Depression. Source: De Gregorio (2004) and World Bank staff estimates (for 2004). 1.6 Growth in income per capita increased sharply after 1984, averaging some 4.1 percent between 1985 and 2004 (Table 1.1). The post-1984 economic expansion occurred 10 years after the military regime introduced a reform program in 1973 transforming Chile into a market-based economy. By 1982, most of the substantial reforms had been put in place, reducing enormously the involvement of the State in the economy. Labor relations were decentralized, most interventions by the State with respect to prices and purchasing were eliminated, and Government expenditures were cut substantially. A wide-ranging privatization of State enterprises took place. Financial sector development was encouraged by privatization, a cut in reserve requirements and the removal of controls on foreign financing of the domestic banking sector. In 1981, the pay-as-you go pension system was replaced with a mandatory savings schemewith individual accounts. The regulatory framework governing hnd managers encouraged domestic investment of the pension hnds, greatly contributing to the development of the domestic financial sector. The economy was also opened up to internationaltrade and investment.Most non- tariff barriers to trade were eliminated, tariffs reduced and capital controls relaxed. Hence, trade liberalizationtook place in Chile ten years before any other Latin American ~0untl-y.~ 1.7 Production, employment and incomes responded very positively to the economic liberalization of the seventies and may have continued on a path of sustained growth if the Government had not changed track. However, in an attempt to bring under control a persistent and irksome inflation, the Government abandoned its flexible exchange rate policy in 1979 and pegged the peso to the dollar while, at the same time, opening up the Mexico was the next big reformer in the region. 8 capital account. The result was a massive inflow of capital attracted by perceived high rates of return and then, two years later, an equally abrupt capital flight when it became clear that the level of foreign debt was unsustainable and an increasingly overvalued peso was leading to unsustainable trade deficits. In 1982, insolvency was no longer avoidable for the domestic banking sector. The Government stepped in to take control of the largest private commercial banks and to supply them with liquidity. It also renegotiated US$4.7 billion of external debt. 1.8 Following the financial crisis, an adjustment program was put in place in 1985. Structural reforms focused on rebuilding the financial sector, which had been devastated by the crisis of 1982-83. Banks and firms that had gone bankrupt during the crisis were recapitalized and privatized. Steps were taken to improve the previously weak regulation of domestic banks which had contributed to the crisis. Macroeconomic policy was tightened to secure economic stabilization. The public finances were protected from volatile copper prices through the establishment of a copper stabilization fund in 1987. External debt was reduced through the extensive use of debt conversion schemes and a flexible exchange rate was adopted. These policies contributed to the restoration of export competitiveness, with a depreciation of the peso by over 50 percent in real terms from 1982to 1988. Figure 1.2: Capital Flows and Real ExchangeRate, 1980-2004 -8 0 - Non-FDICapital Flows - US$bn, left I)cC Real ExchangeRate Index -- Source: Banco Central de Chile and World Bank staff estimates. 1.9 When a democratically elected Govemment assumed office in 1990 it pledged a continued commitment to the market-based economy, but with a renewed emphasis on social programs for poverty reduction. Increased social expenditures were funded through a combination of expenditure reallocation, increased taxes and foreign aid. The Government's commitment to fiscal responsibilitywas demonstrated by sustainedbudget surpluses. Private sector concessions were negotiated to finance and build much-needed infrastructure. The Central Bank, which was made independent of Government in 1989, instituted an inflation-targetingregime which led to a gradual reduction in inflation from the high rates of the late 1980s and early 1990s. An export-oriented growth policy was vigorously pursued, moving beyond unilateral trade liberalization to the conclusion of bilateral free trade agreements with developed economies. The institutional framework was further strengthenedby the adoption of a floating exchange rate in 1999and a policy of structured fiscal s~~rplusesin 2001. The level of public sector debt was reduced still further. The indexation of debt to foreign currencies was gradually phased-out. Instruments were introduced to hedge against exchange rate risk and the solvency of the financial sectorwas further improved. 1.10 The success of export-oriented reforms and economic stabilizationcan be seen by comparing the growth performance of Chile with that of other countries (Table 1.2). In the 1960s and 1970s, Chile had underperfonned other countries in the LAC region. Starting from the 1980s, the situation was reversed. The 1980s-a lost decade for niost Latin American countries-was a time of strong growth in Chile. Growth accelerated even further in the 1990swhen, like other countries in the Latin American region, Chile's economy benefited from a favorable external environment, including large capital inflows and low interest rates. However, unlike other countries in the LAC region, Chile's gains were not eroded by the financial crisis that hit in the late 1990s.Real GDP grew by 6.5 percent in the 1990s in Chile, compared to 2.9 percent in LAC as a whole, and 2.4 percent in high-income OECD countries. Chile's growth performance has been similar to niany of the top-performers in East Asia and the Pacific since the 1980s. Growtli rates in Chile have been similar to those found in Finland and Ireland-the high performers of the EU--over the period 1980-1999. An important difference, however, is that Chile started from a much lower income level. 1.11 The economic performance of Chile during the 1980s and 1990s was radically different from the rest of the region, even though the external environment was similar. Chile was able to insulate its economy from the shocks of the late 1990sbetter tlian many other countries in the LAC region. It was the only country in the LAC region that, during this period, reduced the gap between its GDP per capita and that of advanced economies. It did not experience the deep crises--often fiscal in origin-that afflicted Argentina, Brazil, Ecuador, Mexico, Uruguay and Venezuela. Cross-countrystudieshave shown that much of the better performance of Chle is attributable to the strength of its structural policies and of its institutions (Loayza at al., 2004; Corbo at al., 2005). Economic stability also played an important role in explaining the differences. The more steady growth pattern of Chile owed much to increased trade and trade openness together with a fiscal policy ruled by structural surpluses. An autonomous central bank with enhanced credibility assured the consistency of monetary, exchange rate and fiscal policies. In addition, the introduction of instruments to hedge against exchange rate risk, a sharp reduction of public sector debt and gradual de-indexationof debt to foreign currencies,as well as a well-regulated and supervised financial sector, all helped to enhance economic performance. Table 1.2: Chile and Comparator Countries:Annual Real GDP Growth Rates (percent) 1960s 1970s 1980s 1990s 2000 - 2003 Chile 4.4 2.5 4.4 6.5 3.2 Latin America & Caribbean 5.2 5.8 1.9 2.9 1.2 Argentina 4.1 2.9 -0.7 4.5 -1.8 Brazil 5.9 8.5 3.0 1.8 1.8 Mexico 6.8 6.4 2.3 3.4 2.1 East Asia & Pacific China India Indonesia Korea, Rep. Malaysia Thailand High income: OECD 5.6 3.7 2.9 2.4 2.0 Finland 4.5 3.9 3.6 1.4 2.6 Ireland 4.4 4.7 3.1 7.1 6.7 Italy 5.8 3.8 2.4 1.5 1.4 New Zealand 4.3 1.6 2.0 2.6 3.5 Spain 7.8 3.9 2.8 2.6 2.9 Source:World Bank, DevelopmentData Platform. Slowdown 1998-2003 1.12 Chile experienced slower growth in 1998-2003 as a result of the AsialRussia/Brazil crises and the slowdown in world economic growth at the turn of the century. During that period, GDP per capita grew at an average of 1.3 percent per annum-well below the level of the previous 10 years. The Asian crisis of 1997 was more serious for Chile than for most LAC counties because a substantial share of its exports went to the countries in crisis. Russia's default in 1998, followed by the turmoil of Brazil in 1999,brought a sudden and lasting increase in interest rate spreads for Chile. At the same time, the country faced a decline in net capital inflows over 1998-99which, at 7.9 percent of GDP, was greater than that of all other major Latin American economies (Calvo and Talvi, 2004). From 1990-98, corporate external debt and domestic consumer credit had boomed. Then, following a substantialterms-of-tradeshockwhen the economy was most in need of access to credit, came a credit crunch. Domestic bank credit to the private sector slowed down sharply. The abrupt adjustment of the current account and of the financial sector led to a decline in investment and in the growth rate (Figure 1.3). 1.13 Cowan and De Gregorio (2005) point out that the collapse in net inflows was not typical of other sudden stops experiencedby emergingmarkets, which were characterized by an interruption or reversal in gross inflows. In the case of Chile, it was gross outflows that caused most of the collapse in net outflows -- non-FDI gross inflows remained almost unchanged. The rise in capital outflows is explained by a shift in the investment behavior of pension funds in favor of foreign assets. In addition, and prior to the peso being floated in 1999, banks increased the proportion of their portfolio held in foreign assets to take advantage of opportunitiesfor arbitragepresented by a gradual depreciation of the peso in real terms. The economic downturn was deepened by the contractionary monetary policy adopted by the central bank in 1998 as it sought to support the nominal value of the peso and to close the initially large current account deficit. It was acting in an uncertain environment, made all the more difficult by the previous expansionary domestic demand and fiscal policies of 1997-98. Figure 1.3: The Contribution of Investment,Domestic and External Demand to GDP, 1997-2004 Domestic demand 0 Investment I 0 exports Net -CDP 1 U -4.0 -6.0 Source: Banco Central de Chile and World Bank staff estimates. Figure 1.4: Chile Real GDP (Percentage deviation from trend) Real GDP (left axis) -wM- 50,000 25% -% Trend GDP (left axis) 45,000 1 deviation from trend (right axis) - 20% I I Note: Trend GDP is calculatedusing a Hodrick-Prescottfilter on annual data with h=100. Source: Banco Central de Chile and World Bank staff estimates. 1.14 Notwithstandingthe adverse impact of the external shocks of the late 1990s,Chile managed to emerge from the crisis better than most countries in the LAC region. Furthermore, the Chilean economy was far more successful in withstanding adverse external shocks than in the past. The downturn in Chile in response to the shockswas less severe than on previous occasions (Figure 1.4).There is no doubt that structural reforms, the continuity of economic policy and good institutions go a long way toward explaining this performance. Yet, the downturn illustrated the continued vulnerability of the economy to external shocks, prompting the introduction of a series of measures in 1999 to strengthen the macroeconomic framework. They included the adoption of a fiee- floating exchange rate; the adoption of an explicit fiscal surplus rule; a move to inflation- targeting; and a deepening of the foreign exchange derivativesmarket. 1.15 Following somewhat improved external conditions in 2000-2001, the international climate deteriorated and output in the major advanced economies slowed down during 2001-2002 after the collapse of the US asset price bubble and the September 11 attacks. Favorable prices for copper-Chile's main export commodity- helped the economic recovery in 2004, with real GDP per capita growing at an estimated 6.1 percent. This performance has led to a renewed wave of optimism among both Chilean entrepreneurs and public officialsregarding growth prospects. However, they are well aware that Chile, a small open economy dependent on primary commodities, is still vulnerable to external events. The downturn of 1998-2003 underlines the importance of continuing to strengthen the economy's resilience to external shocks. 1.16 Evidence suggests that microeconomic flexibility-defined as the gap between actual and desired employment-suffered during the shock of the 1990s (Caballero, et al., 2004; Micco, 2005). An economy is deemed flexible if the gap between actual and desired employment is closed fast. For Chile, the authors find that a reduction in microeconomic flexibilityafter the Asian crisis explains a large part of the slower growth in total factor productivity (TFP) over 1997-99. Job security, the minimum wage, and wage indexation became increasinglybinding constraints during this period. The impact on job turnover was magnified by the credit crunch associated with the shocks, which affected small and medium enterprises most of all. Throughout the period fi-om 1985 to 1997,small establishments adjusted far less quickly than large establishments. 1.17 Studies show a significantdecline in the contributionof TFP to growth during the recent slowdown. From 1985, an expansion in TFP became the dominant source of growth, followed by the increase of physical capital (Gallego and Loayza, 2002). Growth in TFP declined in the late 1990s,to a rate in the range of -0.6 to 1.32percent compared with 2.3 to 4.4 percent for the period 1986-1997 (Table 1.3). The recovery in growth in 2003-2004 has been based on a large rise in investment.Employment expansion has been muted compared with previous periods. Table 1.3: Measures of TFP Growth Contribution of (%) Period Output Growth Capital Labor TFP Beyer and Vergara (2002) 1986-1990 6.80 2.00 2.50 2.30 De Gregorio (2004) Fuentes et al. (2004) 1990-2003 5.18 1.76 0.81 2.61 1990-1997 7.14 1.61 1.14 4.40 1998-2003 2.08 1.48 0.16 0.44 Gallego and Loayza (2002) 1986-2003 6.64 2.46 2.22 1.95 Notes: *Calculatedusing GDP estimates for 2004. Controls for input utlization and human capital are included by Fuetes et a1(2004) and Gallego and Loayza (2002) studies. Beyer and Vergara (2002) and De Gregorio (2004) do not adjust for the utilization and quality of labor and capital. Source: Beyer and Vergara (2001) and De Gregorio (2004), Fuentes et a1 (2004), Gallego and Loayza (2002). 1.18 It is not yet clear whether the recent decline in TFP is cyclical or structural. Nevertheless, there is some consensus on the need for a second generation of reforms to foster high and sustained growth (Lagos, 2005; Eyzaguirre et al., 2003; De Gregorio, 2004; OECD, 2003; Beyer and Vergara, 2001; Gallego and Loayza, 2002). In order to spur TFP growth in the future, Chile will not only have to enhance microeconomic flexibility, but also iniprove education and increase the adoption of innovative technologies. For a country like Chile, rich in natural resources, it is critical to have the skills and technology available to fully exploit those natural resources. Maloney (2002) compares and contrasts the strong growth record of natural resource-rich countries such as Sweden, Finland and Australia with the experience of Latin America. He concludes that an inward-looking industrialization policy has held back Latin America-a policy now renounced completelyby Chile. He further concludes that a deficient human capital and innovation environment in Latin America has weakened the ability to innovate and adopt new technologies. It is that latter impediment to growth that Chile would do well to avoid. POVERTY AND SOCIALSECTOROUTCOMES CHILE IN 1.19 Chile has dramaticallyreduced poverty since 1990. It now has the second lowest poverty headcount in the LAC region, afterUruguay (Figure 1.5). 1.5: Poverty Headcount Ratio: LAC countries,Late 1990s, early I I Source: Szkkely(2001); Giovagnoli,Pizzolitto and Trias (2004). Table 1.4: Poverty Headcount: Chile, 1990-2003 OfficialPovertv Line National Urban Rural Poor and Poor and Poor and indigent Indigent indigent Indigent indigent Indigent 1990 38.6 12.9 38.4 12.4 39.5 15.2 2003 18.8 4.7 18.6 4.5 20.1 6.2 Source: MIDEPLAN, Encuesta CASEN (1990, 1992,1994,1996,1998,2000 and 2003). 1.20 Poverty reduction in Chile has been stiking, particularly when compared with other counties in the LAC region.3 The official headcount of poverty fell from 38.6 percent in 1990 to 18.8 percent in 2003. Over the same time period, the official headcount of indigence fell from 12.9 to 4.7 percent (Table 1.4). Looking across the range of poverty indicators assembled by Giovagnoli, Pizzolitto and Trias (2004), the story remains the same whether it relates to rural or urban poverty, to the World Bank definitions of moderate and extreme poverty or to the well-known poverty thresholds defined by a per capita income of US$1 and US$2 dollar a day. By any standard, Chile has been extremely successhl at reducing poverty since 1990. Social indicators such as enrollment in primary education,youth literacy, infant mortality and life expectancyhave also improved, reaching levels close to those of advanced economies (Data Appendix). Primary and secondaryschool enrollment is close to 100percent. 3Chile had the second largest reduction in headcountpoverty over the 1990s (with the DominicanRepublic showing a slightlylarger fall) (Szkkely,2003). Pro-poor Growth 1.21 Growth in Chile has been associated with a substantial reduction in poverty. In absolute terms, growth has benefited the poor. First, growth has been associated with a reduction in the proportion of poor people (as measured by the poverty headcount). Poverty reduction tracked the rate of growth. It was strong in 1990-94, slowed down in 1994-98,stopped in 1998-2000, and bounced back in 2000-2003. Second, the poor have enjoyed a substantial increase in real incomes. The Ravallion and Chen rate of pro-poor growthwas high at around 3.6-4.1percent for 1990-2003 (Table 1.5).4 1.22 Growth has also been associated with a slightly higher rate of increase in the incomes of the poor. Income growth for the poor was a little higher than for the rest of the population for the period 1990-2003 (around 0.1 percentagepoints higher). However, that was not the case consistentlythroughout the whole period. In the 1990sthe growth of incomes for the poor was 0.4percent lower than the overall growth rate (Table 1.5). 1.23 The reduction in poverty since 1990 is due to econonlic growth in combination with Government policies targeted at reaching the poor. Rapid growth of GDP per capita enabled the poor to enter the labor market. Social services were dramatically expanded and public expenditures were targeted more effectivelyto the needy. Table 1.5: Pro-poor Growth and ProgressiveGrowthMeasures, Chile, 1990-2003 - - 1990-1994 1994-1998 1998-2000 2000-2003 1990-2000 1990-2003 1. Growth rate Household s w e y 6.2 5.1 -0.4 0.9 4.4 3.5 National Accounts 6.4 5.5 0.4 1.7 4.8 4.1 2. Change in poverty headcountratio USD 1 -1.5 -0.9 0.1 -0.6 -2.3 -2.9 USD 2 -6.4 -3.9 -0.4 -1.6 -10.7 -12.3 Officialextreme -5.3 -2.0 0.1 -0.9 -7.2 -7.8 Officialmoderate -10.7 -5.8 -1.1 -1.8 -17.6 -19.8 3. Pro-poor growth Ravallion and Chenpro-poor growth rate USD 1 7.8 2.5 -1.8 5.3 3.9 4.1 USD 2 6.2 3.5 -0.6 4.4 4.0 3.9 Officialextreme 6.3 2.9 1.0 2.9 3.8 3.6 Official moderate 5.7 3.8 1.3 2.1 4.0 3.6 Poverty equivalent growth rate USD 1 5.0 4.4 -1.2 -1.2 4.1 4.1 USD 2 5.3 4.3 1.1 1.1 4.1 4.1 Officialextreme 6.7 7.1 4.4 -1.5 4.0 3.9 Officialmoderate 6.4 7.2 6.4 1.3 4.4 4.3 4. Progressivegrowth Difference in growth ratespoor-all USD 1 1.6 -2.6 -1.4 4.5 -0.5 0.6 USD 2 0.0 -1.6 -0.2 3.5 -0.4 0.4 Officialextreme 0.1 -2.2 1.4 2.0 -0.6 0.1 Officialmoderate -0.5 -1.3 1.7 1.2 -0.4 0.1 Source: Gasparini (2004) based on microdata ftom the CASEN. The Ravallion and Chen rate of pro-poor growth measures the average of the income growth rates of the poor. For example, if the definition of poverty is income of less than US$la day, then the average income growth for those individuals in the period 1990-2003was 4.1 percent. Figure 1.6:Persons Active, Employed and Unemployed, 1990-2004 Employed (lhs) 6,500 - 0 - Labor force (lhs) 6y000 - -+Unemployed (inverted rh 5 5,500 - cd rn s 5,000 - --400 4,500- #,*' 4,000 I I I I 1 I I I I I I I I I 600 I I Notes: Includes active share of total population over 15. Source: INE, Chile. Impact of Labor Market on Poverty 1.24 The incorporation of the poor into the labor market was a major factor in the reduction of poverty. Between 1990 and 2003, the total number of people active in the labor market increased by 28 percent, from 4,824 thousand to 6,199 thousand (Figure 1.6).The rise in labor-market participationwas due in a large part to the incorporationof unskilled and semi-skilled women into urban labor markets. The gap between the participation rate of the well educated and the poorly educated fell substantially over the 1990s (Table 1.6).Over the same time period, employment in the formal sector increased by 27 percent, from 4,450 thousand to 5,653 thousand. Employment (as a share of the population over 15) grew particularly fast in the high growth years of the early 1990s (Figure 1.7).The World Bank (1997) found that the main factorbehind the fall in poverty from 1987 to 1994was the incorporation of the poor into the labor market as a result of economic growth. The findings were based on a quantitative analysis of poverty over 1987-94using the CASEN surveys. The analysis also suggests that it was the increase in labor incomes, rather than in non-labor incomes, that played a significantrole in poverty reduction. Table 1.6: Share of Adults in the Labor Force, Chile, 1990-2003 Gender Age Education Area Total Female Male ... (16-25)(26-40)(41-64) (65 +) Low Medium High Rural Urban ... (i) (11) (111) (IV) v) (vi) (vii) (vm) (ix) (x) (xi) (xii) 1990 0.598 0.377 0.842 0.531 0.684 0.562 0.312 0.547 0.597 0.776 0.564 0.605 1994 0.624 0.409 0.857 0.541 0.702 0.608 0.295 0.568 0.642 0.718 0.566 0.635 1996 0.634 0.428 0.852 0.525 0.713 0.625 0.378 0.575 0.636 0.764 0.570 0.645 1998 0.646 0.456 0.853 0.532 0.729 0.639 0.358 0.588 0.647 0.766 0.575 0.658 2000 0.645 0.463 0.838 0.491 0.732 0.655 0.342 0.586 0.643 0.763 0.570 0.657 2003 0.665 0.499 0.843 0.498 0.749 0.675 0.348 0.610 0.666 0.771 0.586 0.677 Source: Gasparini (2004)based on microdatafrom the CASEN. Figure 1.7: Employment Growth and Unemployment in Chile, 1990-2004 (i) GDP and enlployrnentgrowth (as a share of population over 15) 15 Employment---GDPgrowth 10 h s 5 Y 2 0 0 0 2 0 0 1 2 0 0 2 2 0 0 3 2 0 0 4 ~ -5 (ii) GDP growth and unemployment (as a share of population over 15 active in the labor market) 1 1990 1991 1992 1993 1994 1995 1996 1997 I998 1999 2000 2001 2002 2003 2004 -5 Source: INE, Chile. 1.25 Output growth was not particularly concentrated in those sectors with the largest proportion of unskilled. The sectors that contributed most to output growth prior to the downturn in 1998 were the service sectors of commerce, financial services, industry and transport (Table 1.8). These sectors had a smaller proportion of low-skilled workers (Figure 1.8).That is not to say that the sectorsin which the unskilled are concentrated did not gain from growth. Output growth brought gains to all sectors and, indeed, employment grew fastest in those sectors with the highest participation of the unskilled. Communal services, financial services, commerce, construction and industry were the sectors that contributed most to employment growth (Table 1.7). In particular, the construction sector-with the second highest proportion of unskilled workers- experienced a boom in employment. However, employment declined in agriculture. Nevertheless, agricultural growth led to a substantial reduction of poverty and indigence rates in the rural sector from 1990 to 1998, and indeed over the entire period of 1990- 2003 (Table 1.9).It also contributedto the decline in overall poverty, because agriculture is the sector that accounts for the largest share of the poor in Chile (Figure 1.8).~ Ignoring the communal services sector that by definition includes a large share of the poor, as it includes those on work programs. Figure 1.8: Proportion of Low SkilledWorkers by Sector in Chile, 1990 and 2003 76.8 80 Shareof low sldlled workers 1990 Share of low sldlledworkers 2003 70 57.4 60 50 h 34.6 35.4 35.3 40 30 20 10 0 I AGR MIN IND EGA CONS COM TRAN SFIN SCOM I Source: World Bank staff calculations on rnicrodatafrom CASEN. Figure 1.9:Distribution of Indigent and Poor by Sector in Chile, 1990 and 2003 h o El Indigent 1990 HIndigent 2003 30.0 8 2 20.0 3 10.0 .B k 0.0 EO AGR MIN IND EGA CONS COM TRAN SFIN SCOM I B4 Poor, not indigent 1990 IHPoor, not indigent 2003 30.0 I AGR MIN IND EGA CONS COM TRAN SFIN SCOM - Source: World Bank staff calculations on rnicrodatafiom CASEN. Table 1.7: EmploymentDistributionby Economic Sector, 1990-2004 AGR MIN IND EGA CONS COM TRAN SFIN SCOM Total Decom position of change in employment 1990 0.2 0.1 0.3 1991 0.1 -0.1 0.3 1992 0.3 -0.2 0.9 1993 -0.3 0.0 1.O 1994 -0.5 0.1 -0.1 1995 -0.4 0.0 0.0 1996 -0.3 0.1 0.2 1997 -0.7 0.0 0.3 1998 0.1 -0.1 -0.3 1999 -0.3 -0.2 -1.5 2000 -0.1 -0.1 -0.1 2001 -0.6 0.0 0.0 2002 0.0 0.0 0.2 2003 0.5 -0.1 0.5 2004 0.0 0.1 0.0 0.0 0.2 0.2 -0.2 0.5 0.9 1.8 Source: Institute Nacional de Estadisticas (INE) and Bank staff estimates. Table 1.8: Growth by Sector, 1990-2004 AGR MIN IND EGA CONS COM TRAN SFIN SCOM Total 1990 0.8 0.1 0.2 -0.2 0.5 0.7 0.5 0.2 0.3 3.7 1991 0.2 1.1 0.9 0.5 -0.1 1.5 0.6 1.9 0.5 8.0 1992 1.1 -0.1 1.9 0.6 0.7 2.7 1.2 1.5 0.7 12.3 1993 0.3 0.0 1.2 0.1 1.1 1.2 0.4 0.9 0.4 7.0 1994 0.6 0.7 0.7 0.2 -0.1 0.8 0.4 1.O 0.4 5.7 1995 0.6 0.7 1.3 0.2 0.5 2.3 1.1 1.3 0.4 10.6 1996 0.2 1.2 0.5 -0.1 0.5 1.6 0.8 0.9 0.5 7.4 1997 0.2 0.8 0.8 0.2 0.6 0.8 0.7 0.9 1.0 6.6 1998 0.1 0.6 -0.4 0.1 0.2 0.4 0.4 0.7 0.6 3.2 1999 0.0 0.8 -0.1 -0.1 -0.9 -0.5 0.1 -0.1 0.5 -0.8 2000 0.4 0.3 0.8 0.3 -0.1 0.5 0.6 0.6 0.6 4.5 2001 0.4 0.5 0.1 0.0 0.3 0.3 0.5 0.5 0.6 3.4 2002 0.4 -0.3 0.3 0.1 0.2 0.1 0.4 0.4 0.5 2.2 2003 0.0 0.4 0.5 0.1 0.4 0.5 0.4 0.4 0.6 3.7 2004 0.6 0.5 1.1 0.1 0.4 0.7 0.4 0.7 0.7 6.1 Notes: For 1990 to 1996, Communal services (SCOM) includes the categories: Propiedad de vivienda, Servicios personales and Administracibn Pziblica. The sectors in this table are agriculture (AGR); mining (MIN); industry (IND); electricity, gas and water (EGA); construction (CONS); commerce (COM); transportation and communications (TRAN); financial services (SFIN); and conlmunal services (SCOM). The total growth rate is not equal to the sum of the sector shares because total GDP excludes indirect taxes. Source: Banco Central de Chile. Table 1.9:Reductionin Poverty and Indigence by Sector, 1990-2003 AGR MIN IND EGA CONS COM TRAN SFIN SCOM Indigence rate (% of sector) 1990 9.4 5.8 6.1 - 8.5 5.0 6.4 - 5.5 1992 6.7 2.6 3.7 3.7 5.5 3.2 4.9 0.9 3.8 1994 7.4 3.7 2.6 1.5 3.8 2.5 3.5 0.9 2.6 1996 6.3 2.5 1.7 2.3 3.2 1.7 2.1 0.4 2.0 1998 5.0 1.3 1.3 1.1 2.8 1.2 0.9 0.3 1.6 2000 5.8 1.4 1.7 1.3 5.0 1.5 1.9 0.6 1.9 2003 3.8 0.9 1.4 1.7 2.7 1.3 1.4 0.5 1.5 Poverty (not indigent) rate (% of sector) 1990 22.1 21.4 22.4 - 26.4 17.9 19.7 - 16.7 1992 22.0 15.9 18.6 12.0 23.7 14.7 16.9 7.5 17.3 1994 21.7 13.1 14.9 12.9 18.6 11.7 15.5 4.8 11.5 1996 20.2 8.6 11.9 8.2 17.6 9.0 12.2 3.9 9.1 1998 17.6 7.2 10.5 6.4 14.6 8.8 9.5 3.0 8.1 2000 16.9 5.5 9.8 5.8 15.0 8.4 10.0 3.8 8.1 2003 13.3 5.4 9.6 4.9 12.1 7.4 8.9 3.2 6.9 Change over 1990-2003 (%) Indigence rate -5.6 -4.9 -4.8 - -5.8 -3.7 -5.0 - -4.1 Poverty (not indigent) rate -8.8 -16.0 -12.8 - -14.3 -10.5 -10.8 - -9.8 Note: For 1990, the classification of employment by sector differs in the CASEN. For these years, the categories of servicios de gobierno and sewiciospersonales are included in total Communal services. Source: World Bank staff calculations on microdata from CASEN. 1.26 There is evidence of an increasing skills-premium in wages over 1990-2003. As far as labor income is concerned, the gap between the income of high-skilled versus low- or medium-skilled workers increased substantially from 1990 to 2003 (Table 1.10). The real monthly wages of the well-educated worker were 3.4 times those of the poorly- educated worker in 1990.By 2003, that ratio had climbed to 4.4. Table 1.10:Wages, Hours and Labor Income in Chile by Level of Education, 1990-2003 (pesos of November 2003) Wages Hours of work Labor income Low Mid High Low Mid High Low Mid High 1990 705 991 2,457 50.7 50.7 45.7 124,585 188,040 422,870 1994 801 1,185 3,150 49.1 48.7 45.2 145,045 217,961 558,236 1996 908 1,487 3,819 48.1 48.4 44.6 140,378 240,959 626,254 1998 878 1,377 3,804 47.2 48.4 45.4 141,488 240,164 666,674 2000 830 1,274 3,589 48.2 48.8 46.7 141,918 225,693 687,718 2003 965 1,330 3,847 44.9 46.3 44.2 147,992 226,665 657,500 Note:Labor incomerefers to real monthly wages. Source: Gasparini (2004)based on microdata from the CASEN. 1.27 Following the stagnation of employment over 1998-2002,employment recovered 5 percent over 2003-04. Unemployment increased after 1998, rising from 6.1 percent in 1998 to 9.7 percent in 1999. The recent economic recovery did not improve the employment situation commensurately, and unemployment still stood at 8.8 percent in 2004. The poor are particularly vulnerable. The rise in unemployment over the 1990s disproportionatelyaffected those in the lower tail of household income distribution,with unemployment being an important determinant of extreme poverty (Litchfield, 2002). The loss of employment income falls particularly hard on those near the poverty line. The World Bank (2004b) fo~mdthat the losses of employment income together with health hazards are the two most serious risks confronting low-income households in Chile. Unsurprisingly, the unemployed and the inactive had much higher-albeit falling-rates of poverty and indigence than the employed throughout the period 1990-2003 (Table 1.11). Employment has grown since the downturn, though growth has been at a slower rate than in the first half of the 1990s. 1.28 Evidence suggests that microeconomic rigidities-particularly the minimum wage-may have amplified the impact of the crisis on the poor. Micco (2005) argues that the minimum wage discouraged employment particularly in small firms after 1998. That is because the proportion of unskilled wage-earners and low wage-earners is higher in small firms than in big firms. He also found that these small firms, with a greater proportion of unskilled workers, were more likely to collapse after the crisis. Cowan, Micco and Pagks (2004) found the negative impact of minimum wages on employment to be greater in those sectors with the highest share of unskilled workers (agriculture and construction). They are also the sectors that account for the largest share of poor and indigentworkers. Table 1.11:Proportion of Employed, Unemployed and Inactive over 15who are Poor or Indigent, 1990-2003 Em~loved Unemploved Not active Poor, not Poor, not Poor, not Indigent indigent Indigent indigent Indigent indigent (%> ("h> ("h> (%> ("h> ("h> 1990 6.5 19.9 28.3 31.1 11.8 25.8 1992 4.2 17.8 20.6 31.2 8.5 23.5 1994 3.4 14.1 19.0 27.1 7.5 19.9 1996 2.6 11.8 17.1 26.4 5.8 17.2 1998 1.9 10.1 19.0 24.4 5.7 16.2 2000 2.5 10.0 16.1 23.5 5.4 14.3 2003 1.8 8.5 14.9 22.0 4.7 15.1 Source: World Bank staff estimatesbased on microdata from CASEN. 1.29 Employment in the construction industry shrunk dramatically in 1999 and has failed to recover filly. Large reductions in employment were also seen in industry and transport in 1999 and 2000. Subsequent output growth in these sectors was not accompanied by a complete recovery in employment. As a result of the slowdown, the pace of poverty reduction was slower, with indigence even increasing in most sectors over 1998-2000. While this trend has since reversed, it illustrates the vulnerability of the extremepoor to a downturn and particularly to a fall in employment. Much of the output growth since 1998 has been in the financial services and mining sectors. But the mining 6This excludes the communal services sector, which by design (due to employment programs) accounts for a large proportion of the poor or indigent who are working. sector has only a small workforce and employs few of the poor or indigent; as a consequence, growth in'this sector has little effect on poverty or indigence. Poverty Impact of Government Policies 1.30 Meller (1999) attributes approximately 40 percent of the reduction in poverty to social policies, and the remaining 60 percent to higher economic growth. Social expenditures almost doubled to about 16 percent of GDP in the period 1987 to 2002 (Arellano, 2004). The rise in social spending was financed through increased tax revenues and by reducing the proportion of expenditures devoted to defense, debt service and social security. The increase in social expenditures focused on extending the coverage of health and education services, and social protection for the most vulnerable. Those expenditures have also been targeted effectively towards the lowest income quintiles. 1.31 As a share of GDP, central Government health expendituresrose from 1.9percent to 3 percent over 1990-2003. That followed a period when health spending declined in the 1980s. Chile's health indicators are now similar to average OECD levels. Despite these gains, there remain challenges in the provision of healthcare. The high cost of healthcare can make it difficult for poor households to escape from poverty and can drive others into poverty. 1.32 Central Government public expenditure in education as a percentage of GDP almost doubled between 1990 and 2003, increasing from 2.3 to 4.3 percent. The ratio of education expenditures to GDP is similar to that of developed countries and, in fact, greater than the average of OECD countries.Average years of schoolinghave risen since the 1980s. The increase in education investment that took place since 1990 has led to faster progress in educationalachievement at all income levels. 1.33 A remaining challenge in Chile is to provide educational opportunitiesto under- schooled adults, especially to those adults from low-income backgrounds. Some 53 percent of the adult population in Chile has not completed secondary education. Low schooling levels are particularly prevalent among the poor. Years of schooling increased from 12.2in 1990to 13.6in 2003 for the population in the lughest income decile. During the same period, years of schooling also increased for the population in the lowest income decile, but only from 6.7 to 8.2 (Table 1.12). Reducing the gap in years of schooling between high and low income adults is important if Chile is to enhance its competitiveness and reduce income inequality. Enrollment rates also differ according to income. While the secondary school eivollment gap between high- and low-income students has narrowed between 1990 and 2003, the gap between high- and low-income students in pre-school and higher education enrollment remains substantial despite the significant increase enrollment rates in these levels. Table 1.12: Years of Education for Adults aged 25-65 in Chile, 1990-2003 By household income quintiles 1 2 3 4 5 Average 1990 6.7 7.3 8.0 9.3 12.2 9.0 Source: Gasparini(2004) based on microdatafrom the CASEN. 1.34 Effective targeting of social protection services helped greatly to reduce poverty. A range of institutions exist in Chile to protect households fi-om adverse income shocks such as job loss, the loss of earnings or costly medical treatment. A new unemployment insurance scheme was begun in 2002. In 1981, a pioneering retirement benefit system was put in place. The health system was enhanced to include a guaranteed basic package of services. 1.35 Poverty did not grow during the downturn of the late 1990s,despite a substantial increase in unemployment and slower employment growth. This is attributable to Government programs. Active employment programs played an important role in cushioning the effect of the downturn on the vulnerable. In the late 1990s, the Government gradually began to shift resources into indirect forms of employment assistance and away fi-om direct work programs. These indirect programs included subsidies for employment creation in the private sector (SENCE), skills-renewal, and financial assistanceto would-be small-scale entrepreneurs (FOSIS). 1.36 In response to stubbornly high rates of indigence in the mid-1990s, the Government launched a program entitled Chile Solidario in 2002. This is a conditional cash transfer scheme that targets the 225 thousand most vulnerable households in Chile. Participating households are given priority access to primary health care and the schools attended by their children receive special vouchers. Preliminary evidence based on the poorest households that have passed through the first phase suggests that there has been progress in meeting a range of socialtargets. 1.37 In conclusion, vulnerabilities remain for households with incomes close to the poverty line. Poverty can ensue when household members losejobs, particularly if those jobs were in the informal sector or the member was a self-employed worker whose business failed during the downturn. For the informal or self-employed worker, there remain institutionalbarriers to even the most basic forms of socialprotection. Inequality in Chile 1.38 In terms of income distribution, Chile remains among the most unequal countries in World. Its income distribution parallels the most unequal countries in Latin America and Africa, and is much more unequal than in OECD countries. The only country with a higher level of inequality and a higher level of income is South Africa (Figure 1.10). Chilean inequality is almost twice that of the advanced economies and 1.5 times that of the US (the most unequal of the advanced economies). Figure 1.10: Chile is One of the Most Unequal Countriesin the World r I I Gini index I I I Source: World Bank, Development Data Platform. 1.39 Inequality is not a new phenomenon in Chile. Statistics of earnings in Santiago, dating back to 1957, show that inequality has persisted at high levels throughout the second half of the 2othcentury. There has been some variation over time. In particular, during the period 1974-1987, inequalitybetween wage earners in Santiago increased, but then fell in the 1990s(Meller and Tokman, 1996). 1.40 Wherever there is extreme income inequality, GDP growth is less effective in reducing poverty. Higher inequality tends to entail a lower rate of poverty reduction for any given positive rate of growth. Inequality in Chile changed very little from 1990- 2003: The level of inequality as measured by the Gini coefficient has remained stable over 1990-2003 at about 0.56. Table 1.13: Distribution of Household Income per capita in Chile, 1990-2003 Gini Theil CV N.5) A(1) A(2) E(O) E(2) ... (i) (11) (111 (iv) (v) (vi) (vii) (viii) 1990 0.562 0.676 1.960 0.265 0.435 0.686 0.570 1.920 2003 0.561 0.705 2.365 0.267 0.432 0.668 0.565 2.795 Source: Gasparini(2004) based on nlicrodatafrom the CASEN. 1.41 Income distribution in Chile is highly skewed. The share of income held by the poorest decile of the population is 1.2 percent. In contrast, the top decile has a 46.4 percentage share of income. Table 1.14 shows income shares by decile in Chile from 1990-2003. The structureis similarto the rest of the LAC region, with an extremelyhigh concentration at the top end of the scale and large differences between the rich and the middle class (De Ferranti et al. 2003). The distribution has been relatively stable, with only small changes. Table 1.14:Income Distribution by Decile in Chile, 1990-2003 (percent) National average 88,561 112,520 127,683 137,445 136,441 139,359 Source: Gasparini (2004) based on microdata from the CASEN. 1.42 While the structureof the income distribution is similar to the LAC region, Chile has a particularly high concentration at the top. However, differences in income between the poor and middle income earners are much less than in many advanced economies. The ratio between the poorest and the second poorest deciles in Chile is half that of the US and the UK. If the wealthiest decile is excluded, inequalitybetween the remainder of the population is much less and Chile appears more equal than the US (Szekely and Hilgert, 1999). 1.43 A recent study of social mobility in Chile finds that the income concentration at the top creates a barrier to mobility between the elite and the rest of the population. However, there is significant mobility among the rest of the population due to the low income differentials between them (Torche, 2005). That points to a crucial aspect of inequality in Chile: the skewed income distribution means that a large proportion of the population is vulnerable to poverty. While poverty is limited to the bottom two income deciles, the income shares of the bottom five deciles are relatively close to that level and, therefore, a large proportion of the population may be vulnerable to poverty in the absence of mechanisms to help them deal with income shocks. 1.44 Evidence shows that greater educational attainment is associated with higher wage earnings. Therefore, improving access to quality education and increasing the participation of the poor in adult training programs would seem to be one way of targeting inequality. BLI~increasing equality of educational attainment is not the sole strategy for reducing inequality in Chile. The functioning of the labor market is also important. There is a positive correlation between a large informal sector and income inequality. For Chile, the majority of employees in the informal sector earn substantially less than their counterparts in the formal sector. By making it more difficult for workers to move from the informal sector to the formal sector, rigid labor market institutionsmay contributeto creating more rather than less inequality (Heckman and Pages-Serra,2000). CONCLUSION: DEVELOPMENT CHALLENGES 1.45 The twin challenges for Chile are to achieve high and equitable growth. 1.46 Achieving high growth. The aim of Chile is to achieve convergence with the lower tier of advanced economies by 2020. Meeting this objective will require sustained and high growth: If GDP per capita in Chile were to grow at 4 percent per annum, its incomes would rise to the present level of Spain by 2023 (Table 1.15). A dynamic enterprise sector, with the ability to rapidly adjust to shocks, is critical if Chile is to achieve the growth needed to reach its goal of advanced-country income convergence and social equity. Innovation and the enhancement of human capital will also be crucial elements to spur TFP growth, and to bring Chile closer to the knowledge-based economies of advanced OECD countries such as Australia, Finland and Sweden. Growth was central to reducing by half the level of poverty in Chile since 1990. Higher employment was also crucial in order for growth to raise the incomes of the poor. A continued focus on employnieiit generation will be necessary if poverty is to be fkther reduced. Table 1.15: Chile: Convergenceof GDP per capita with Spain (year of convergence under different growth assumptions) Chile (% GDP growth) 3 4 5 h 0 2030 2023 2019 2017 Spain (% GDP 1 2043 2030 2024 2020 growth) 2 2083 2044 2030 2024 Note. Calculatedusing 2003 GDP per capita expressed in US$ PPP. Source: World Bank, Development Data Platform and World Bank staff estimates. 1.47 Achieving equity. The high level of inequality in Chile is a cause for concern. Income is concentrated in the hands of the top 10percent of the population. The incomes of the poorer deciles are clustered close to one another. This means that while only one in five is poor, many more remain vulnerable to poverty. There is a strong case therefore for focusing efforts on decreasing the likelihood of those in the bottom deciles falling into poverty. That could be done through increasing incomes and by strengtheningthe social safetynet. 1.48 State intervention,particularlythe effectivetargeting of social protection services, played a large role in the reduction of poverty since 1990. While there has been much progress in protecting the vulnerable, the social protection system needs to be further strengthened. Many households remain vulnerable to poverty fiom adverse income shocks, such as job loss, costly health hazards, and loss of earnings-ability in old age. This not only threatens equity, it also damages prospects for growth. Where households do not have a sufficientlywide array of tools with which to insure against risks, resources that could otherwisebe put to productive use may be preempted by the need for adjusting to shocks. In some households, the recovery from shocks can even impair human capital or hinder investment. 1.49 Better education is a means to achieve both high growth and income equity. In Chile, there are great differences in adult educational attainment. A majority (53 percent) of the adult population in Chile today has not completed secondary education, and low schooling levels are particularly prevalent among the poor. That is a great barrier to innovation and equity. But raising adult education levels is not a matter of education policy alone. High levels of informality in the Chilean economy are a challenge to be addressed because they act as a barrier to on-the-job training. 1.50 In striving for equity among future generations, education is fundamental. The educational reforms beg~min 1990 brought about significant increases in educational investment and led to a substantial expansion in the quantity and quality of educational inputs. But two important challenges remain: Improving still fixther the quality of education and raising equity of educationalopportunityin Chile. 2. PROMOTING SUSTAINABLEPRODUCTIVITYGROWTH 2.1 Recent studies highlight the importance of a dynamic process of entry and exit of firms as a way to proniote growth andjob creation in industrial and developingcountries alike. As new firms enter and less efficient ones leave the market, there is a reallocation of factors which results in higher factor productivity. The key is having in place a business environment or investment climate that facilitates the entry and exit of firms. "The investment climate is the set of location-specific factors shaping the opportunities and incentives for firms to invest productively, create jobs, and expand. Government policies and behaviors exert a strong influence through their impact on costs, risks and barriers to conipetition." (World Bank, 2005). Bergoeing and Repetto (2003) shows that factor reallocation explains more than 50 percent of total productivity growth in the Chilean manufacturing sector. The percentage increased to 80 percent between 1990 and 1997,a period of macroeconomic stability. 2.2 For Chile to achieve its goal of convergence with the lower tier of advanced economies by 2020, a key challenge is to eliminate impediments to the reallocation of factors of production -particularly in response to shocks (Eyzaguirre, 2004). More generally, an improvement in the policies and institutions of a competitive market will contribute to sustainable growth. Chile has done well in establishing good competition policies, ranking on a level with developed countries. Nevertheless, recent studies show that high closing costs, prevailing labor policies and difficult access to credit have reduced the dynamism of firms after the shocks of the 1990s. Moreover, according to a recent World Bank Investment Climate survey7 covering firms of all sizes and in all sectors, about 25 percent of entrepreneurs believe that labor regulations, access to finance, and anti-competitive practices are severely limiting the expansionof their firm. 2.3 The objective of this chapter is to identify shortcomings in the market for factors of production and to propose policy options to help remedy them. The chapter primarily examines two issues. First, the dynamism with which Chilean firms contribute to higher factor productivitythrough the efficient reallocation of factors (Section A). Second, how business dynamism may be sustained during shocks, with special attention to closing costs, labor flexibility and access to credit (Section B). Section C. discusses institutions to foster competition and Section D concludes with a summary of key issues and recommendations. The chapter draws'heavily in new work commissionedfor this report8 It also draws upon other World Bank reports9 and presents new empirical results to support their conclusions. 'Annex 2.1 in this report. Micco (2005) and Galetovic (2005). 9 Chile-Report on Observance and Standards & Codes. Insolvency and Creditor Rights Systems. The World Bark, June 2004. BUSINESS ENVIRONMENT THE DYNAMISM FIRMSlo AND OF 2.4 Chile has afavorable business environment relative to many other countries and firms are generally dynamic." As a result of sound Government policies the macroeconomic environment is secure. Privatizatioii, market liberalization and good financial sector management has lowered the cost of doing business. As a result, and as shown by an analysis of business behavior, there is a rapid process of entry and exit and labor turnover is also rapid, on a level with OECD economies. 2.5 The size structure of Chile's private sector is top heavy. There are some 6470 large and mega-large firms of which about 200, for the most part export orientated, are generally efficient and well-managed (Table 2.1). At the other end of the scale there are about one million micro-enterprises that mostly serve the domestic market. They are labor intensive and only marginally profitable and about half of them operate in the informal sector. Large firms (includingmega-large firms) account for 17 percent of total employment while small and micro-firms account for a much greater 70 percent. There was a notable increase in the number of micro firms during the late 1990s (Figure 2.1). That appears to be the result of improved credit availabilityfor micro firms, as well as the collapse of many small firms into micro firms with the economic slowdown of the late 1990s (World Bank, 2004). Employment, on the other hand, has remained relatively stable with increases in the share of larger firms at the expense of employment in small and micro firms (Figure 2.2). Likewise, during 1998-2001, sales of small and medium firms increased only 3 and 6 percent while sales by large firms increased 35 percent (in contrast to the period 1994-1998 when sales increased around 40 and 45 percent, respectively). '* Table 2.1: Chilean Firms by Size, Sales, Exports and Employmenta Definitions: Mega large firms are defined as those with annual sales net of VAT above UF600, 000, large firms have sales between UF100,OOO and UF600,000, medium firms have sales between UF25,OOO and UF100,OOO small firms have sales between UF2,400 and UF25,OOO and micro firms have sales below UF2,400. UF 1=US$31.31. a. The Chilean classificationof firms is unique, making internationalperformancebenchmarking difficult. Based on the Chilean definitions, many firms labeled as "micro businesses" would be considered small enterprises in other countries (World Bank, 2004). Source: IMF-WB (2005) based on FSSA data. lo This sectiondraws heavily in Micco (2005) whlch was commissioned for this report. l 1 World Bank (2005a and b) and Annex 2.1 of this chapter. l2 Micco (2005). Figure 2.1: Number of Firms by Size, and GDP Source: Corfo (2003) and Fundes-Sercotec(2005). Figure 2.2: Employment: Employees by Size of Firm I I Note: In household surveys, around 5 percent of respondents cannot be classified by firmsize. This, plus the different nature of household and employment surveys, explains why employment by size does not sum to total employment reported by the INE. Source: Household SurveysCASEN (MIDEPLAN) and Employment Survey(INE). 2.6 The dynamism of Chile'sprivate sector is evidenced by thefollowingfindings: During 1995-200 1, 11 percent of medium-size firms grew into large firms. Some 54 percent of micro firms entered the market and 33 percent exited or did not report sales in 2001 (Table A1 in Annex 1). The dynamism of Chilean firms is comparable to that of other emerging economies, i.e., Mexico, Brazil and Hungary and Slovenia. The dynamism of micro, small and medium firms in Chile is coniparableto those in the USA, Japan and other OECD countries. Exit rates for medium and large firms are high by OECD standards (World Bank, 2004). On average, employment turnover affects one of every four jobs each year. Also, on average, one of every eight jobs is newly created each year. Annual gross job creation is 12.9 percent of total employment, while annual average gross job losses amount to 13.2 percent of total employment (Table A2 in Annex 1). In the manufacturing sector, job redeployment is on average 30 percent, and the entry-exit of firms accounts for almost 50 percent of this rate. That is to say, one out of every eight jobs is either created by a newly established firm or is lost by a plant that is closing (Table A3 in Annex 1).13 In manufacturing, Chile's job turnover (26 percent) is comparablewith that of United States and Canada (20 percent) as well as that of other emerging economies in the region such as Mexico and Brazil (30 percent).14 The survival rate of firms and the speed of job reallocation are related to the size of firms. In small and medium firms, the entry-exit process accounts for more than two thirds ofjob redeployment, while in large plants it accounts for only one quarter. 2.7 In the manufacturingsector, aperiod of slow growthfrom 1998-2002 affected the pace of entry and exit and ofjob turnover. During the previous period of high growth, almost 30 percent of manufacturingplants exited the market after four years of operation. The proportion went up to around 40 percent during the period of slow growth. Further, during the high growth period, the average annual growth rate of employment for medium and large firms was 1.5 higher than the rate for small and medium firms (Table A4 in Annex 1). Employment in medium and large firms actually declined during the period of slow growth. On average, small plants reduced employment by 10percent and medium plants by 6 percent. 2.8 In sum, Chilean firms are dynamic and heterogeneous, i.e., many firms increase their sales and the number of their employees in any particular year, while others reduce them. On average, one of every four jobs is either created or destroyed each year. The entry-exit process is mainly concentrated in firms that are less than large. The market behavior of firms and their productivity varies greatly according to size which augurs well for further productivity gains and sustaineddynamism if the business environment is further improved. Afier the shocks of the late 1990s, small firms did not recover as l3Similar results are reported in Ferrada and Reinecke (2004)based on multi-sector data. 14Job turnover is defined as the sum of creating new jobs and eliminating old jobs (Table A.3 -Annex I). See Micco and Pages (2004) for a discussion of data issues. readily as large firms in terms of sales and employment. The business slowdown provoked by the shock of the late 1990s suggests that smaller firms may face high costs of adjustment due to structural and institutional impediments that need to be removed, as discussed in the next section. SUSTAINING DYNAMISM FIRMSDURING SHOCKS THE OF l5 2.9 Competitionpolicies are generally sensible and appropriate in Chile. According to an index issued by the WB Institute in 2004, Chile ranked in the league of developed countries with respect to the quality of policies. Indeed, Chile's percentile rank of 94.1 is above that of the OECD (90.6) and well above the average of Eastern Europe (50), East Asia (45), and South Asia (30). But, recent research and entrepreneur surveys suggest there is room for the improvement of competitionpolicies and institutions.16 A reduction in microeconomic flexibilityafter the shocks of the late 1990swas due to such factors as the high cost of closing a firm, the poor enforcement of creditor rights and the high cost of reallocating labor and financial resources between firms. Removing the obstacles to microecoiiomic flexibility for firms other than large firms is particularly important as they account for the largest share of total employment. 2.10 Improving competitionpolicies to reduce costs,particularly during shocks, could help increase and sustain business dynamism. High entry and exit costs reduce productivity growth. They also encourage informality as a means of avoiding the costs of entry and exit. The legal and institutional frameworks governing the entry and exit of firms are generally appropriate. There are, however, a number of shortcomings to be addressed if Chile is to enjoy the full benefits of competition. The high cost of bankruptcy proceedings and the enforcement of creditorrights merit particular attention. 2.11 Entry costs could be further reduced. For an enterprise that seeks to enter the market, Chile ranks better than SouthKorea and Spain and Ireland in terms of entry costs but not as good as Australia, Finland, New Zealand and the USA (Table 2.2). Reducing start-up costs for Chilean firms would encourage informal enterprises to formalize their structure and, according to Djankov et a1 (2002), that could reduce the size of the informal sector by almost 10 percentage points of GDP and informal employment by 15 percentage points. Furthermore, if Chile simplifies its business start-up procedures and reaches the rank of New Zealand, it could increase annual growth in value added per worker by 0.25 percentage points in industries such as retail and food production. Those industries have naturally low entry barriers relative to industries such as chemicals or paper-pulp which have high natural entry costs (Klapper et al, 2004). At the present juncture, the notarization of articles of incorporation,recording them in a public deed and l5This section draws heavily from Micco (2005) and Galetovic (2005) which were commissioned for this report. 16World Bank (2005) on business environment, and Micco (2005). obtaining a municipal operating license all increase unduly the costs of entry for an enterprisethat seeks to incorporate.17 2.12 Closing a business remains costly. International rankings point to Chile's high cost of closing a business relative to other industrial and emergingeconomies(Table 2.2). Key sources of high exit costs are the bankruptcy law and the poor enforcement of creditorrights. One of the main objectivesof bankruptcy is to maximize the total value of proceeds received by all stakeholders. That is, businesses should either be rehabilitated, sold as a going concern, or liquidated, whichever generates the most total value. A successful rehabilitation not only satisfies the creditors, but protects employee incomes and provides all stakeholders with an opportunity to continue in the market. The successful rescue of a business saves jobs, and often provides creditors with a greater return. In short, viable businesses can be rehabilitated while non-viable ones should be liquidated. Furthermore, effectiveinsolvencyproceedings provide for the orderly transfer of business assets from non-viable firms to other firms that are more efficient and that perform better in the market. 2.13 Chile's bankruptcy law does not provide expeditious procedures for shutting down an enterprise. To address some of the problems, a new law was proposed to Congress in 2004. Improvements in thejudicial area are not easy to implement.However, the experience with the "Reforma Procesal Penal" shows that it is possible.18Similarly, enforcement of creditor rights is important for market entry and good business performance. Chile does not rank well internationallyin that respect. Even though Chile fares better than other Latin American economies, it still lags behind such countries as Australia, Korea and New Zealand. Countries with strong and effective creditor rights have a more secure and predictable market for credit, which is important for investment planning (Table 2.3). The Second Capital Market Reform Bill (SCMR) is a good first step towards addressing some of the problems involved. A detailed account of the issues and recommendationsto address them can be found in Annex 2.2. l7The World Bank (2005). l8Valdivieso (2002). Table 2.2: Cost of Starting and Closing a Firm, Job Security and Cost of Hiring and Firing Workers Cost of (Botero et (Botero et al, 2004) Closing a Business Starting a al, 2003) Cost of Time to Cost of Recovery Business Dismissal Firing complete bankruptcy rate (cents Job Procedures Workers closure proceedings on the (% GNI Security Index Index (years) (% of estate) dollar) per capita) Index Chile 10.3 0.31 1 0.81 1 0.29 5.6 14 23.1 Australia 1.9 0.14 1 0.53 1 0.14 1.O 8 80.0 Canada 0.9 n.a. 0.8 4 90.1 Finland 1.2 n.a. 0.9 4 89.1 Ireland 5.3 n.a. 0.4 9 88.0 New Zealand 0.2 0.04 1 0.00 1 0.14 2.0 4 71.0 Spain 16.5 n.a. 1.O 14 77.9 South Korea 15.2 0.26 0.62 0.29 1.5 4 81.7 USA 0.5 0.08 0.07 0.14 2.0 7 76.3 OECD 6.5** 0.33 0.41 0.46 1.5** 7.6** 73.5** Latin America 56.2 0.50 0.5 0.36 3.5 17 28.2 & Caribbean East Asia & 41.7 0.26 0.52 0.24 3.3 27.6 26.4 Pacific South Asia 39.7 n.a. 4.2 7.3 19.7 ** High income country. Definitions: Job Security: the average of (i) protection of grounds for dismissal; (ii) protection of procedures for dismissal; (iii) notice and severance payments; and (iv) the constitutional right to job security. The job security index ranges from zero to one. Cost of firing workers: the cost of firing 20 percent of a firm's workers (assuming that 10 percent are fired for redundancy and 10 percent without cause). Cost is defined as the sum of the period of notice, severance pay, any mandatory penalties established by law or mandatory collective agreements for a worker with three years of tenure with the firm. If dismissal is illegal, this cost is one year's wage. The new wage bill incorporates the normal wage of the remaining workers and the cost of firing workers. The cost is the ratio of the new wage bill to the old one. Dismissal procedures measures worker protection granted by law or by mandatory collective agreements against dismissal.The index ranges from 0 to 1and is the average of the followingseven dummy variableswhich sum to one: the employer (1) notifies a third party before dismissingmore than one worker; (2) needs the approval of a third party prior to dismissing more than one worker; (3) must notify a third party before dismissing one redundant worker; (4) needs the approval of a third party to dismiss one redundant worker; (5) must provide relocation or retraining alternatives for redundant enlployees prior to disnlissal; and (6) priority rules applying to dismissal or layoffs, and (7) priority rules applyingto reemployment. Source: Doing Business (World Bank, 2006), unless otherwiseindicated. Reducing the Cost of Reallocating Factors: Labor Flexibility and Access to Credit 2.14 Less labor flexibility and restricted access to credit has been associated with a reduction in microeconomic flexibilityand, hence, in the late 1990s,a slower adjustment of manufacturing industry to shocks and to the policy response (Caballero, Engel and Micco, 2004). Microeconomic flexibility is assessed in terms of the labor adjustment or the speed at which a firm closes the gap between its current and desired level of employment. The difference between wages and the marginal productivity of labor was adopted as a proxy for that gap for the period 1979-1992.19 An economy is said to be flexible, if faced with an unexpected shock, the employment gap is quickly closed a result of prompt adjustment. Table 2.3: Creditor Rights, Rule of Law, Days neededto Enforce a Contract, Davs to Collect a ~ouncedkheck.Effective creditor Rights. and Volatilitv of credit Definitions: Creditor Rights: degree to which secured creditors are protected duringbankruptcy procedures. A score of one is assigned when each of the following rights are defmed in laws and regulations: (i) there are restrictions, such as creditor consent or minimum dividends, for a debtor to file for reorganization; (ii) secured creditors are able to access their collateralafter the reorganizationpetition is approved(no "automatic stay" or "assetfreeze."); (iii) secured creditorsare paid first out of the proceeds of liquidatinga bankrupt f i ; and (iv) management does not retain administration of its property pending the resolution of the reorganization. Rule of Law: Includes several indicators that measure the extent to which agents have confidence in and abide by the rules of society. Days to Enforce a Contract: number of days to resolve a payment dispute through courts according to Djankov et a1 (2004). Total Duration of the Procedure: the number of days needed to collect on a bounced check. Effective Creditor Rights is the product of Creditor Rights and Rule of Law (both normalized between 0 and 1). Source: Djankov et a1(2004),Kaufmann et a1(2003), World Bank Governance Indicator (2004), and Galindo and Micco (2004). 2.15 Using a larger time series (1979-2001), Micco (2005) shows that, within a year, plants close 72 percent of the employment gap, confirming that microeconomic flexibility in Chile is similar to Colombia and Brazil, higher than Mexico and Venezuela, but lower than in the USA. It also confirms that larger firms have greater flexibility than smaller ones. This result is in line with the hypothesis that credit constraints determinethe speed of adjustment according to the size of firms. Changes in employment for small firms, which are more credit constrained, respond not only to the gap between wages and the marginal productivity of labor, but also to the availability of finance. For small employment gaps, internal resources may be sufficient. As discussed below, in the period from 1997 to 2001 l9An analysis of data from the Chilean Manufacturing Census (Encuesta Nacional Industrial Annual, ENIA in Spanish)was published by the Institute Nacional de Estadistica covering all manufacturing plants in Chile with more than ten employees. when growth was slow, there was a reduction in microeconomic flexibility in Lack of access to credit may be the explanation. Figure 2.3: MicroeconomicFlexibility in Chile (Average adjustment coefficient) I I Source: Micco (2005) based on ENIA. Facilitating the Redeployment of Labor: Job Security, Minimum Wageand Wage Indexation 2.16 Labor markets in Chile are flexible, ranking better than industrial and emerging economies. But there is growing consensus in Chile that the minimum wage and wage indexation also became increasingly binding following the shocks of the late 1990s. Country experience show that job security, minimum wages and the price indexation of wages could be potential barriers to fast labor reallocation, particularly during shocks.'l Labor regulations aimed at job security increase the cost of laying-off workers and therefore lead to fewer dismissals when confronting negative shocks. Conversely, when confronted with an economic up-swing, employers respond cautiously, taking into account that workers may have to be fired in the fbture. Hence, the employment response is smaller. Similarly, a high minimum wage and the indexation of wages to prices do not allow downward wage adjustment and that, in turn may lead to the dismissal of employees if the firms need to retrench. Recent studies identified job security, minimum wage and wage indexation as sources of loss of labor flexibility following the shocks of the late 1990s. Also, recent surveys indicate that labor regulations are perceived by 30 percent of large firms and 23 percent of small firms as one of the factors that negatively affects their business (Annex 2.1). Box 2.1 summarizes Chile's situation with respect to job security, the minimum wage and wage indexation. 2.17 Job security could be improved. The costs of firing workers (i.e., the cost of advance notice required, severance payments and penalties due when dismissing a redundant worker expressed in terms of weekly wages) is at the level of Sub-Saharan Afiica, although somewhat higher than in East Asia & Pacific economies. That confirms 20Micco (2005)using the same observations as Caballero, Engel and Micco (2004) for the 1986-1999 period. 21IADB (2003). 37 the poor position of Chile in global rankings. Although kles governing the termination of employment in Chile are less restrictive than in other Latin America economies and similar to those prevailing in OECD countries, they are nevertheless more restrictive than in other shock-prone economies such as New Zealand and Australia (Box 2.I), where labor-market flexibilityis key to rapid adjustment. Box 2.1: Chile. Job Security,Minimum Wage, and Wage Indexation Job security. Chile's regulatory provisions for job security consist of severance payments and a type of unemployment insurance. Severance payments have been the traditional legal mechanism to protect employees because they are easy to implement. However, it is inefficient as an insurance mechanism and it distorts hiring and firing decisions. Severance payments account for more than 80 percent of the benefits. Furthermore, unemployment insurance is inefficient as an insurance mechanism and generates moral hazard problems. See Chapter 5 for details. Minimum wage. The minimum wage has increased since the late 1990s. In May 1998, Congress set the minimum wage to increase 12.7 percent by June, 1998, 12.4 percent by June, 1999, and 10.5 percent by 2000, to achieve a minimum monthly wage of 100,000 pesos by the year 2000. This policy entered into effect right before growth slackened in the third quarter of 1998 and when growth was still projected at over 6 percent per year." Around 6 percent of Chilean workers were affected by the minimum wage hike that took place between 1997 and 2000 (Cowan et al, 2004) based on microeconomic data from the Chilean employmentsurvey.A 12percent increasetook place on July 1,2005. Wage indexation. Information on indexed wage contracts is limited, but reportedly wage indexation has decreased in recent years. Collectivebargaining has led to wage contracts extending on average two years. They include clauses for full backward indexation to inflation every six months, and they have not changed significantlyover the last 15 years, despite falling inflation.In the manufacturing sector coverage was 35 percent of total sector employees. a. See Cowan and Micco (2005a) and Cowan et a1(2004)for a discussion on the rationale of this measure. 2.18 The problems engendered by the framework of job security in Chile impair labor mobility and increase administrative costs (Box 2.1). For example, the judicial process which settles disputes between employees and employers is extremely slow in Chile and, hence, it increases the administrativecosts of the system. Gazmuri (2004) shows that labor disputes take around 250 days to be settled in Chile. If one party appeals, the process can last an additional 230 days. A recent study based on a sample of 60 countries during the period 1980-1998 found a lower speed of adjustment of industrial employment in countries with high legal protection against dismissal, especiallywhen such protection is likely to be enforced (Caballero, Cowan, Engel and Micco, 2004). Based on ths study, if Chile were to adopt a framework of job security comparable to that of New Zealand, it would increase the speed of adjustment of employment by 14 percentage points and accelerate growth by 0.3 percent. Other recent studies confirm this result.22 2.19 Chile's statutory minimum wage and long-termprice-indexed wage contracts may have become increasingly constraining in the years following the Asian crisis. Chile's statutoryminimum wage increased in the late 1990s.The increases affected an estimated 2 22Micco and Pages (2004). percent of workers with high levels of education and experience and 13percent of workers with low levels of education and low levels of e~perience.~~Minimum wages expressed as a proportion of wages of unskilled workers doubled to around 60 percent between December 2002 and December 1997(Figure 2.4). In the constructionsector, the rise in the minimunl wage affected 20 percent of unskilled worker positions and 16 percent of total jobs. The increase in wages may come from either an increase in wage rates or a reduction in the number of employees with low wages (a change in the composition of the labor force). The results suggest that a binding minimum wage puts more pressure on the payroll of small companies, and therefore on their employment level. This could partly explain the slow growth of employment in small firms after 1998. Those firms have a higher proportion of unskilled labor, and therefore of low wages than larger establishments, consequently,a binding minimum wage affectsthem more.24 2.20 Recent studies covering theperiod to 2000point to these twopolicies as the source of Chile's slow adjustment to the shocks of the late 1990s (Cowan, Micco, Mizala, Pages, and Romaguera, 2003, and Cowan, Micco and Pagks, 2004). High costs-including factor reallocation costs-- hinder productivity gains and prevent firms from making a fast adjustment to shocks. The impact of shocks in the late 1990s was more pronounced for small firnls in Chile (Micco, 2005).~~During this period, small manufacturing firms with 10to 25 employees reduced their payroll more than larger firms. Employment dropped 17 percent in small firms compared with 1.4 percent in larger firms. This contrasts with the performance of the 1991-96 period when average eniployment in small firms and large firms grew at 8 and 0.3 percent respectively. Figure 2.4: Minimum Wages relative to the Wages of Unskilled Workers (~ercent) 70 M) 50 40 30 g z t g k z z 9 z z z g $ $ $ $ > 7 7 7 7 7 7 7 Source: INE. 2.21 Reducing the severance-payment component and increasing the employment insurance component ofjob security and eliminating wage rigidities would contribute to a faster adjustment to shocks. To preserve job security while preventing distortions in the 23LOWlevels of education and experience are defined as 0-11years of schooling and 0-9 years of experience 24Cowan, Micco, Mizala, Pages and Romaguera (2004)provides details in unions-firms wage negotiations. 25A very flexible economy is one in which gaps disappear quickly as a result of prompt adjustment (Caballero, Engel and Micco, 2004, based on manufacturing employment data). 39 hiring and firing decisions of firms, consideration could be given to reducing the severance-payment component and increasing the employment insurance component. To sustain labor flexibility, an alternativepolicy to annually adjusted minimum wages and two year-indexedwages would be to tie wage changes to observabletriggers such as changes in unemployment. In order to select appropriate triggers, transparency of data concerning public and private contracts would be a first step. Further analysis is needed regarding private sectorpractices of collectivebargaining to make more specificrecommendations. Improving Access to Credit 2.22 Chile has a well-functioning financial sector that ranks with other countries at a similar level of development. But bank credit to the private sector stagnated in 1998-2000 with large firms crowding out smaller firms. Confidence in Chile's financial markets abroad is evident from the continued availability of foreign credit during the international capital market crisis of the late 1990s(De Gregorio and Cowan (2005). In comparison with OECD countries, Chile's OECD index of restrictivenesstowards foreign direct investment is lower than that of all b ~ one, ~ t the United Kingdom (OECD, 2004). Given the ready availability of external financing, it would appear that the causes of credit scarcity during the late 1990s and early 2000s were either domestic policies andlor shortcomings in the domestic financial markets. The scorecard on this is but all studies highlight the presence of microeconomicimperfectionsin Chile's credit markets. 2.23 Overall, Chile'sfinancial system ranksfavorably by regional standards, but it still somewhat weaker than industrialized countries (Table 2.4). For example, Chile's market capitalization, at 119 percent of GDP, is higher than the average industrial country.27But, despite improvements introduced by the First Capital Market Reform-the market is still illiquid with only a 10per cent stock tuniover annually. This contrast with a stock turnover in industrial countries of 86 percent. Likewise, private bonds as a percent of GDP, although the highest in the region, lag behind Korea and kustralia. Also, bank intermediation,while the highest in the region at aro~md63 percent, is still below the above-mentionedcountries. Interest rate spreads, i.e., the difference between lending and deposit rates, while lower than regional spreads,remain stubbornlyhigher than more developed countries.28 26Gallego, Smith-Hebbel and Serven(2004); Caballero (2003); Calvo and Talvi (2004). 27This percentage is even larger than the level of industrial countries(i.e., 70 percent). Brock and Rojas Suarez (2000). Table 2.4: Size of FinancialMarkets in Selected Countries,2003 % GDP % GDP % % GDP Chile 63.26 119.16 10.25 22.8 Australia 98.97 112.08 76.54 29.6 New Zealand 118.63 41.54 38.25 0 SouthKorea 103.851 54.45 236.77 44.5 Brazil 34.64, 47.64 32.43 9.6 Mexico 18.47 19.57 20.75 2.5 Hungary 43.04 20.22 57.64 2.3 Slovenia 41.19 22.09 0.34 n.a. Middle East and North Africa 58.78 32.84 13.73 n.a. Europe and Central Asia 21.87 9.81 15.97 n.a. Industrialized 105.15 73.91 85.71 38.4 Latin America and the Caribbean 40.65 22.97 3.55 7.4 - East Asia & Pacific 35.48 51.26 45.37 31.6 SouthAsia 27.25 17.5 86.63 n.a. Note: Data for regionscorrespondto the median. Source: World DevelopmentIndicators 2005, BIS and IFS. 2.24 Competition in the banking sector has increased in recent years with new entrants, including foreign banks, and a further development of the capital market. This has contributed to a sustained availability of credit, although principally for larger finns that also have access to foreign credit as well as the local capital market. The recent LMF-WB Financial System Stability Assessment (FSSA, 2004) characterized Chile's financial system as deep, well-diversified, sound and resilient to Private pension funds, the life insurance sector, the mortgage industry and commercial paper and corporate bonds have developed significantly in the last 20 years, although at different stages and speeds. Sources of vulnerability are difficult to identify. Dollar indebtedness is low, bonds are characterized by long maturities, and the corporate sector enjoys credibility abroad as evidenced by large corporate external liabilities and large holdings of private assets abroad. The equity market is large also. On the other hand, the scorecard is mixed when it comes to measuring the impact on competition of financial sector taxes such the tax on loan transactions (the stamp tax) and the tax on check transactions (Honahan, 2002).~' Chile is not alone in this. Brazil and Ireland also have taxes on financial transactions which are an effective way to collect fiscal revenue. Further analysis is need on the impact of these taxes because they seem to penalize short-term borrowers, which are mostly small finns, more than long-term borrowers, which are mostly large firms (Arrau, 2005). The evidence is also inconclusive when it comes to perceptions of reduced competition attributed to bank concentration. Cross-country comparisons suggest that Chile is not different from the international norm with regards to bank concentration. But high profitability reinforces 29OECD (2004); IMF(2005). 30 The stamp tax amounts to up to 1.608 percent of the amount transacted, while the check tax amounts CLP$138 per check or transaction. It is equivalent to around 15 percent of the cost of making a check transaction and is between 60 and 650 percent of the cost of other types of transactions. perceptions of low competition. IMF (2005) tentatively associates low competition with high effective entry costs (in the form of higher capitalization ratios notwithstanding low regulatory barriers to entry), limited competition from other non-bank financial intermediaries and regulatory restrictions on the investment of private pension finds (which are the largest institutionaldepositors). Further analysisis needed on ths issue. 2.25 Competition has not contributed signzficantly to improve access to credit by small and mediumfirms. As highlighted in FSSA, this is a worldwide phenomenon,not confined to Chile, that reflects the special characteristicsof these firms. It is more difficult to verify information on small firms and, even with a well-developed financial sector, small firms would have less access to credit than large firmsbecause they are riskier and there are fixed costs in the lending process. (World Bank, 2005 and Benavente et al, 2005). But, while large firms enjoy easier access to bank credit than smaller firms, the cost difference is smaller and collateral requirements are lower in Chile than in some other countries regardless of firm size (thus the interest rate on loans in Hungary and Slovenia (2002) are almost twice the rate in Chile for any given size of firm). Also, the value of collateral as a percentage of loan value in Chile is lower than that in Hungary and ~lovenia.~' 2.26 Two explanations have been advancedfor the sudden reduction of credit to small and mediumfirms in the late 1990s: competitionfrom less risky largerfirms and the higher vulnerability of small firms to rollover risk due to poor enforcement of creditor rights. Bank credit to the private sector stagnatedin 1998-2000 (Figure 2.5); large firms crowding out credit for smaller firms (Figure 2.7). Costlier credit for smaller and medium firms relative to large firms is partly the result of their higher credit risk and the higher cost to creditors of monitoring them. After 1998, large firms that needed working capital as a consequence of the tight international credit market crowded out riskier smaller firms. In Chile, bank loans account for around 20 percent of short-term financing for all firms, which is typically used for working capital (Figure 2.6). This makes small fimis vulnerableto cuts in working capital during periods of restricted credit, because banks prefer to lend to larger firms than to riskier smaller firms that may already be over indebted (World Bank, 2004). This appears to have been the case in Chile's recent credit crunch.32 The leveling-off of credit to firms other than large h s reflects this (Figure 2.7).33Also, banks were reluctant to rollover short-term credit to small and medium firms, due partly to high credit risk. Chile fares better than the Latin American economies on the enforcement of creditor rights, but it is still behind countries like Australia, Korea and New Zealand. These countries have been able to weather crises more successfilly than Chile, and have experienced less credit volatility. The trend of total credit by firm size suggests that after the AsianRussian crises small and medium sized firms faced a tight credit market. During this period, preliminary results suggest that firms became increasingly credit constrained forcing them to exit the market more for reasons of liquiditythan for lack of profitability. 31Hungary--micro firm: 190 percent; large firm: 139 percent. Sloveni-micro fm:146percent;large firm: 90 percent. 32A similarphenomenon took place in Argentina during the Tequila crisis of 1995(Calvo, 1997). 33Micro credit grew during the whole period, at the beginning at 4 percent and after 1998 at 6 percent. The number of micro loans grew at around 9 percent a year during the whole period compared to 1.7percent for other credit after 1998.Governmentpolicies oriented to this sectormay explainths expansion. Figure 2.5: Bank Credit to Private Sector: Stocks and Flows (millionpesos of December 1998) Stocks:blue line, right axis Flows: red bars, left axis Source: Banco Centralde Chile % GDP % GDP Yo % GDP Chile 63.26 119.16 10.25 22.8 New Zealand 118.631 41.541 38.251 0 South Korea 103.85 54.45 236.77 44.5 Brazil 34.64 47.64 32.43 9.6 Mexico 18.47 19.57 20.75 2.5 Hungary 43.04 20.22 57.64 2.3 Slovenia 41.19 22.09 0.34 1. n.a. Middle East and North Africa 58.78 32.84 13.73 n.a. Europe and Central Asia 21.87 9.81 15.97 n.a. Latin America and the Caribbean 40.65 22.97 3.55 7.4 East Asia & Pacific 35.48 51.26 45.37 31.6 South Asia 27.25 17.5 86.63 n.a. Note: Data for regions correspond to the median. Source: World Development Indicators2005, BIS and IFS. Figure 2.6: Bank Commercial Figure 2.7: Bank Commercial Loans Loans by Size of Firm by Size of Firm ST financing structure KW I Mega Large Medium Smll Mcro 1 H % bankloans % public securities % supplier credit % intraconpany loans % other STfin. sources + ~ + W rn-k@I Source: IMF-WE3 (2005) and FECUs data. Source: SBIF,FUNDES and Sercotec2005. 2.27 Further development of the financial sector with the better enforcement of creditor rights, and better credit information systems will enhance access to credit, especially for smaller firms. Specific recommendations from the FSSA to further develop the financial sector include the following (i) broaden and diversify investment opportunities for private pension funds; (ii) fill gaps in the market structure to improve market liquidity and development; and (iii) adapt oversight to meet the needs of an increasingly integrated and complex financial system. Banks have recently increased their lending to small and medium firms through leasing. Measures such as unifying the legislation on movable collateral and creating a single registry for pledges, as discussed, will further promote bank lending. The World Bank report A Strategy to Promote Innovative Small and Medium Enterprises makes the following recommendations: (i) evaluate the feasibility of market driven schemes such as mutual guarantee associations; (ii) credit scoring (measuring the probability of a loan applicant repaying his or her loan on the basis of the past record of all debtors) and (iii) credit-self evaluation (involving a series of standardized formats developed by the financial institutions). Building on the success of Bunco del Estado with microcredit, other commercial banks could be encouraged to develop specialized lending programs for small and medium enterprises. 2.28 Barriers to greater competition exist in particular at the institutional level. The Fiscalia Nacional Econdmica and the Competition Tribunal are the two main institutions that oversee competitionin Chile. The Fiscalia is charged with monitoring market behavior and prosecuting anticompetitive acts. Its aim is to defend the public interest. The Tribunal cannot initiate an investigation.The Tribunal hears and decides on cases involving alleged anticompetitivepractices. The Tribunal can suggest changes to laws if they are deemed to harm competition. Unfortunately, many times the rulings of these institutions have protected firms that were cutting prices more aggressively, as opposed to those involved in anticompetitive practices. Government also affects competition through discretionary administrative decisions by sector ministries, as well as through regulations applied by specific agenciesthat affect. 2.29 Strengthening institutions will help guarantee competition. Two potential safeguards that might be used to prevent firms misusing the legal system and forestalling competition are (a) higher standards of economic analysis, for the claims of plaintiffs are seldom supported by rigorous analysis, let alone by empirical evidence; and (b) allowing the Trib~malto dismiss cases that are have no economic justification, hence are not well justified by the plaintiff. CONCLUSIONS RECOMMENDATIONS AND 2.30 Even though Chile has a favorable investment climate relative to many other countries and a relatively high degree of microeconomic flexibility, a record of sustained high growth rates since the late 1980s was interrupted by the slowdown of 1999-2002. Recent business surveys suggest that further action may be needed if the Government is to succeed in increasing GDP so that it convergeswith the lower tier of advanced economies by 2020. 2.31 The analysis of business behavior presented here shows that the dynamism of Chileanfirms and the rates ofjob turnover are similar to OECD economies. During normal times (as opposed to times of shock), 30 percent of firms exit in a period of four years, and on average, one out of every four jobs are either gained or lost each year. This analysis also shows that after the shocks of the late 1990s,the dynamism of firms slowed down and microeconomic flexibility diminished, i.e., it took longer to close the gap between actual and desired employment. There are wide differences in market behavior and productivity for firms of all sizes, which points to the likelihood of further and sustained productivity gains if the business environment is further improved. 2.32 How can Chile's business environment befurther improved,given already adequate competition policies, flexible labor markets and a well-developedfinancial sector? Chile has done well in establishing credible macroeconomic policies and institutions and generally good competitionpolicies, ranking on a level with developed countries.But there remain barriers to competition (or microeconomic inflexibilities) which, if reduced or removed, would enable the economy to become more resilient to shocks because adjustment would be facilitated. It is very important for Chle to move in that direction because the region is vulnerable to shocks and there is a potential for contagion, despite a strong macro-financial framework and credibility in international markets. Removing barriers to competition is particularly important for small firms. They account for the larger share of employment and, typically, they are disproportionatelyaffected by shocks, as shown in this chapter. 2.33 Recent surveys and studies and the reduction of microeconomicflexibility after the shocks of the 1990s point to a number of conclusions. Small and medium sized firms suffered a loss of dynamism due to high closing costs (including bankruptcy, enforcement of contract andjob security costs), wage rigidity, and dgjcult access to credit. Financial resources are needed for labor flexibility, i.e., to sustain the desired level of employment. It facilitates entry and exit andjob reallocationthrough the assured provision of credit. That is particularly important during credit crunches provoked by shocks when tight financial conditions do not allow the creation of new production units or the expansion of existing ones to absorb laid-off workers.34 Furthermore, credit crunches can lead to business closures due to liquidity problems. During shocks when gaps emerge between actual and desired employment, it is more difficult for sniall firms to access credit than larger firms, hence the adjustment is slower for them. In this regard, a well-developed financial market, improved enforcement of creditor rights, and better credit information systems will help sustain access of credit, in particular for smaller firms. 2.34 Firms of all sizes benejtfrom improving the investment climate. But, as highlighted in WorldBank (2005), the same investment climatepolicies may have different impacts on SMEs because of their size. Additional policies beyond improvements in the general investment climate may be needed to strengthen SMEs. Such policies would not provide special preferences or subsidies because they have proven both costly and ineffective in strengthening SMEs. Successful policies would help level the playing field for firms of all sizes. The World Bank report on SMEs accompanying this report provides specific recommendations based on country experiences to help SMEs better cope with circumstances that affect them more than larger firms. These include credit scores (measuring the probability of a loan applicant repaying his or her loan on the basis of the past record of all debtors); credit-selfevaluation (involving a series of standardizedformats developed by the financial institutions); aid possibly the development of mutual guarantee associations (MGAs) -- business networks that provide guarantees and offer technical assistance to members on business matters. Some 70 such guarantee schemes have been operating in Western Europe and the United States. However, they may create market distortions and are prone to moral hazard. Hence, careful analysis of their effectiveness is needed before they are introduced in Chile. Also, building on the example of Banco del Estado, more commercial banks could be encouraged to develop specialized lending programs. 2.35 Finally, institutions are as important as policies to guarantee competition. The practices of the Fiscalia Nacional Econdmica and the Competition Tribunal could be further improved to promote greater competition. 2.36 A summary of the most important recommendations discussed in this chapter to remove remainingmicroeconomicinflexibilitiesis presented below: (a) Reducing costs where technicallyfeasible Enforce creditor rights and the bankruptcy law. The Second Capital Market Reform Bill could be enacted promptly (i) to create a Unified Registry for all movable assets pledged as collateral, as well as other information pertinent to lenders; and (ii) to introduce modem rules dealing with the treatment of off-sets and netting-out 34Caballeroand Hammour (1996); IADB (2004); World Bank (2005). of financial contracts in the context of insolvency proceedings as well as provisions which clarify the treatment of subordinatedebt agreements in bankruptcy. Annex 2 provides detailed recommendations. (b) Reducingthe cost of reallocatingresources across firms Labor issues. (i) Redesign the framework for job sec~~ritywith a view to reducing the role of severance payments and increasing the role of the unemployment insurance, thereby keeping the overall package neutral with respect to the labor market; (ii) in order to improve labor-market flexibility during shocks, consider tying the minimum wage to measurable variables such as unemployment. Evaluate alternative wage indexation options with the aim of having at least one wage component flexible. A first order of priority is transparency of data in the private and public sector. Credit issues. Follow FSSA recommendations to increase efficiency and financial sector resilience to shocks. Evaluate the feasibility of establishing in Chle some of the widely-used market-oriented schemes such as mutual guarantee associations that could ensure access to credit for SMEs during shocks. Follow the recommendations of the World Bank study on SMEs. Banking sector. Undertake further sector analysis to investigate the relationship between high profitabilityand a perceived lack of competition between banks. (c) Improvinginstitutionsruling on competition policies Fiscalia Nacional Econdmica and the Competition Tribunal. When ruling on cases involving uncompetitive practices, it is important that rigorous economic analysis be employed and that the respective decisions of each institution be enforced. Allowing the Tribunal to dismiss cases that are not well justified by the plaintiff would improve efficiency. Table A2.1.1. Transition Matrix By Size for Firms Between 1995-2001 Whole Economy (Size defined by Sales) (in percent) 1995 \ 2001 Micro Small Medium Large W/O Sales Exit Micro 40.8 4.0 0.2 0.0 14.7 40.3 Small 21.7 39.2 4.6 0.6 10.6 23.3 Medium 8.2 21.6 37.3 10.7 6.0 16.2 Large 5.0 6.7 12.1 59.6 3.3 13.3 WIoSales 14.7 3.4 0.6 0.3 35.5 45.6 Entry 53.7 9.7 1.2 0.4 34.9 0.0 Total 36.9 8.5 1.4 0.7 27.1 25.4 Note: w/o sales: firms with records in the SII but without sales during this year. Source: Gustavo Crespi (2003). Table A2.1.2. Job Creation and Destruction and Net Changes in the Economy Note: Size is defined by the averagenumber of employeesbetween t and t-1. Source: Reinecke and Ferrada (2005) based on ACHS. Table A2.1.3. TransitionMatrix by Size of Firm, 1993-1997 and 1997-2001 ManufacturingIndustry 1993-1997 Small Medium Large Exit Total 1993 Small 62.5% 5.5% 0.2% 31.8% 3212 Medium 14.1% 57.4% 6.0% 22.5% 1358 Large 0.8% 17.6% 67.4% 14.2% 472 Entry 77.6% 18.9% 3.6% 0.0% 1595 Total 1997 3439 1340 462 1396 6637 1997-2001 Small Medium Large Exit Total 1997 Small 49.6% 3.6% 0.0% 46.8% 3439 Medium 12.6% 45.1% 5.0% 37.3% 1340 Large 0.4% 12.3% 53.5% 33.8% 462 Entry 71.2% 20.9% 7.9% 0.0% 1100 Total 1997 2659 1016 401 2265 6341 Note: Bold numbers refer to the number of firms in each period. For matrix 1993-1997,Entry refers to plants that are in operation in 1997 but not in 1993, and Exit refers to plants that are in operation in 1993but not in 1997. Size categoriesdefined by sales and employeesare not comparable.For example, only 25 percent of firms with 200 or more employeeshave sales above 1 o6UF. Source: Author's calculationsbased on ENIA. Table A2.1.4. Net and GrossFlows in Manufacturing (percent) Net Flows Year Small Medium Large All 2001 -3.6 4.1 12.2 6.8 Avg.1993-97 1.5 0.5 -0.3 0.3 Avg.1997-01 -10.1 -6.3 0.3 -3.8 Gross Flows Avg. 1993-2001 Year Small Medium Large All POS 19.8 15.4 11.2 14.2 POS Entry 13.7 7.0 2.4 6.1 NEG 23.5 17.9 11.2 15.8 NEG Exit 15.9 9.5 3.6 7.9 NET -3.6 -2.5 -0.1 -1.6 SUM 43.3 33.3 22.4 30.0 EX-SUM 36.1 27.7 18.1 25.8 Note: Size is defined by the averagenumber of employeesbetween t and t-1. Source: Authors' calculationsbased on ENIA. In this annex new evidence is presented on the investment climate in Chile based on data from a recently completed survey carried out by the World Bank between August and December 2004. The sample consist of approximately 1000establishmentsin nine industries, including IT services, biotechnology and fish-farming, as well as six manufacturing industries. It covers firms of all sizes, including micro, small, medium and large firms. A World Bank team and the Instituto Nacional de Estadistica (INE)selected the sample. The survey collected data on the perceptions of managers as well as hard data on the costs associated with investment issues. An assessment of survey data will provide the tools and analytical framework to identify priorities for reform of the country's investment climate (ICA). It will analyze constraints to costs and productivity at the level of the firm. The Chile ICA is expected to be completed in 2006. The Investment Climate Survey inquires into the opinions of entrepreneurs concerning the obstacles to growth facing their firms. It shows the percentage of firms that consider the obstacle to be "severe" and "extremely severe." Table A5 below presents the results for Chile's overall sample as well for large firms and for a group including medium, small, and micro enterprises. Labor regulations, macroeconomic instability, labor skills, anti-competitive or informal practices and access to finance and tax rates emerge as major concerns for the managers of about 26 percent of Chilean firms surveyed (Table A5). Labor regulations and access to credit are the most serious concerns for large and smaller firms, respectively. Not surprisingly, the weights attached to the various obstacles in Chile clearly differ from the usual pattern observedin developing countriesand in Latin America in particular. Table A2.1: Obstaclesto Growth-Entrepreneurs' Perceptions Obstacles to growth All firms Large SMEs Licensing and permits Legal system1conflictresolution Tax administration Economic & regulatoryuncertainty Laborregulations Anti-competitive or informal practices Transport Tax rates Corruption Telecom Access to finance(e.g., collateral) Crime, theft, disorder Labor skills Macroeconomic instability Electricity Cost of financing (e.g.,interestrates) ANNEX2.3 INSOLVENCY CREDITOR AND RIGHTS Predictable mechanisms for creditor protection and expeditious insolvency proceedings contributeto market confidence and attract investment.Furthermore,well-defined legislation governing matters of collateral and efficient enforcement procedures help lower the cost of financing, and encourage entrepreneurship and commerce. Chile's legal and institutional framework governing insolvency and creditor rights is typical of a modem, credit-based economy. There are, however, a number of shortcomings to be addressed if the economy is to manage credit risk effectively. Specifically, three areas of regulation could be improved, namely those affecting (i) market entry; (ii) business operations within the market; and (iii) market exit. These issues are important if the private sector is to drive economic growth and encourage both foreign and domestic investment. This annex identifies some of the major weaknesses with respect to creditor rights and insolvency procedures in Chile and offers options for reform. Creditor Rights and Enforcement in Chile: Removing Barriers to Market Entry and Optimal Market Performance Creditor Rights. The enforcement of creditorrights is important to encourage new businesses to enter the market and perform well. In Chile, there are shortcomings that create barriers for both lenders and borrowers alike. For instance, although Chilean law permits many types of asset to be used as collateral in obtaining loans, there is an over-reliance on the use of real- estate as collateral, such as mortgages on housing or land.35This type of "immovable" security is generally preferred because it can easily be authenticated at mortgage registries36 and because verification and enforcement (in the event of need to pursue debt collection through judicial proceedings) is relatively inexpensive. Creditors that hold real-estate as collateral are granted "secured" status because other types of'collateral, such as "prendas" are deemed to be less reliable.37 There are many different laws governing assets other than real estate38and there is no single registry for them. Moreover, the law does not always requireprendas to be registered and, even when registration is required, there may be many different registries (resulting in higher administrativecosts). The preference for real-estate as collateralputs SMEs at a relative disadvantage and is a significant barrier to obtainingcredit. To overcome such problems, a legal framework could be established to provide for the creation, recognition, and enforcement of collateral agreements for all types of assets - movable and fixed, tangible and intangible. The Second Capital Market Reform Bill (SCMR) takes the first steps in addressing these problems. It proposes the creation of a Unified Registry for all movable assetspledged as collateral, as well as other informationpertinent to 35In rare cases, mortgages on ships,mines, or aircraft are also accepted as collateral. 36Registration takes place at the Real Estate Registry where the property is located and publicly discloses the legal status of the property. 37"Prendas" refer to the use of movable assets such as cars, and other tangible or intangible property as collateral, but not real-estate. 38At least eight different laws govern security rights on movable assets, creating a number of inconsistencies and shortcomings that limits the widespread use of movable assets as collateral. lenders.39If enacted, the Bill would replace some of the present legislation governing collateral and introduce provisions to facilitatetlie use of other types of ~ollateral.~~Even the framework for "secure" real-estate collateral is not entirely satisfactory. Mortgages are filed in registries that are not interconnected and which still use traditional written methods of record-keeping. They need to be replaced with more transparent and effective databases which share informationthrough IT connectivity. Enforcement Procedures. In Chile, creditors also face difficulties in securing repayment. When a debtor defaults on a loan, creditors must initiate lengthy and costly judicial proceedings in order to collect. For example, "executor proceedings", in cases where the loan contract identifies an executor, can last between one to three years. Ordinary civil proceedings, in which proof of the credit must be provided, can last between three to five years.4 Even the auction of collateral or other attached assets can be lengthy and 1 complicated. Procedures for the recovery of debt in default should be efficient, transparent, and reliable, including the seizure and sale of immovable and movable assets and sale or collection of intangible assets. Drawn out enforcement procedures and uncertainty about the enforceability of contractual rights reduce creditor confidence and increase the interest rate on lending, thereby impairing private sector development. Improving the efficiency of civil proceedings (especially for unsecured lenders), streamlining formalities, and shortening procedures are key reform options to better enforcethe rights of creditors. Insolvency Procedures: Maximizing Valuewhen exiting the Market In Chile, the insolvency law allows a court to order the reorganization of an enterprise. However, creditors are not permitted voting rights. According to the law, reorganization plans are negotiated only with unsecured creditors. Secured creditors only vote if they relinquish their security. This legacy of the previous regulatory framework governing reorganizations is particularly important because most financial corporate credit is secured.42 Moreover, the current law does not take into account the ongoing business needs of the debtor d~~ringreorganization,making it more difficult for the firm to remain afloat during the transition. Formal reorganization procedures would be significantly improved by: (i) introducing provisions to meet the ongoing business needs of the debtor, such as authorizing the use of cash-in-hand that may be pledged or constitute security, or to obtain new financing subject to assurances and safeguardsfor its eventual repayment; (ii) allowing creditors a vote in reorganizations and; (iii) enhancing the participation of creditors in insolvency 39Information also recorded in the Registry includes: other loans for which the same asset has been pledged, any previousjudicial decisions limitingthe ownership of asset, any previouspolice reports regarding said asset, etc. Access to such information would increase verifiability (reducing costs for borrowers) as well as decrease costs of enforcement,if necessary. 40Guaranties other than mortgage-backed securities commonly used in Chile are: personal guarantees, such as bonds Cfianzasimple),joint-and-several guaranties Cfianzasy codeuda solidaria) and guaranteeby endorsement (aval), whch usually only complement other guarantees and securities over movables (such as accounts receivables, intellectual property rights, debtor's credits against third parties, proceeds, future and to-be- purchased assets). 41When compared to one week in the United States, Germany, Ireland, and Tunisia. "Doing Business 2004", pg. 61. World Bank. 42As discussed earlier,there is an over reliance on real estate, "or secured", collateralin Chile. proceedings, mainly through the creation of creditor committees in reorganization and liquidation. Informal workouts and restructuring (extra-judicialreorganization) are minimally regulated by the Chilean Insolvency Law. Although workouts are often used in practice, a significant shortcoming of the current system is that it does not provide for an expeditious way to convert a workout approved by a majority of creditors into a prepackaged restructuring plan, binding on dissenting minorities. If the commercial viability of a firm cannot be restored and liquidation is the final resort, then this process should be carried out as soon as possible. However, in Chile, liquidationproceedings are quite inefficient. The duration of an average liquidation is 24-36 months. Once privileged and secured claims are paid, unsecured creditors typically receive a very low proportion of their due. Legislation could impose time limits to ensure that the process is conducted without delay so that creditors can maximize recoveries. Furthermore, specialized courts with jurisdiction over commercial law could be created in order to reduce congestion in the courts and large caseloads. Since the law does not address subordinated claims in insolvencyproceedings, the effect of a debtor's bankruptcy on subordinated debt agreements--quite frequently used in syndicated loans or project finance credits-is uncertain. The current legal treatment of off-set and netting-o~~tof financialcontractsin bankruptcy is poorly defined, especiallywhen a financial contract establishes the conditions for all derivative transactions between the parties ("open contract" or convenio /contrato marco). If enacted, the Second Capital Market Reform Bill would amend the Insolvency Law by introducingprovisions which clarify the treatment and effect of subordinated debt agreements on bankruptcy. The Bill also introduces up-to-date rules dealing with the treatment of off-set and netting-out of financial contracts in insolvency proceedings.43 Finally, Chilean laws dealing with the international aspects of insolvencies are incomplete and outdated. Adopting up-to-date rules, such as the UNCITRAL Model Law on Cross- Border Insolvency, to deal with international aspects of cross-border insolvencies would appropriatelyaddress the weaknesses of the current system. 43AS originally drafted, the Second Capital Market Reform Bill is an evolution of the legal regime currently in force. However, regulation to be issued after the enactment of said Bill should incorporate several provisions to make the system fully workable. 3. INNOVATION 3.1 The slowdown in growth during the last half decade has raised concerns that Chile may have exhausted previous sources of growth. Cross-country analysis shows that differences in total factor productivity (TFP) account for fully half of all differences in income and growth. Many governments in the OECD, and increasingly in Latin America, have come to identify barriers to innovation as a constraint to further growth. The European Council, when it met in Barcelona in March 2002, aimed at "turning the EU into the most competitive knowledge-based economy in the world" by increasing expenditures on Research & Development (R&D) from 1.9 percent of GDP to 3 percent by 2010, thereby closing the gap with the United States (2.7 percent) and Japan (3 percent).44Chile, with a share of 0.7 percent, and the rest of Latin America (0.4 percent), is in a different league altogether45and President Ricardo Lagos has targeted an increase in R&D expendituresto 1.5percent of GDP by the year 2010.~~ 3.2 This chapter examines three issues. First, how can we know if Chile really has an "innovation problem" that explains its weak TFP? Second, given that the evidence suggests that Chile does indeed have an innovation problem, what do we mean by innovation and in what respects does Chile lag? Third, what does recent experience and literature suggest in the way of principles and broad policy measures to foster innovation? Even in the OECD, there is little consensus on which specific policies governments should adopt, and even less on the suitability of those policies for developing countries. This chapter attempts to organize thinking around some basic principles and to present alternatives based on international experience, rather than providing specificpolicy advice. DOESCHILEHAVEAN INNOVATION PROBLEM? 3.3 Why has TFP been relatively low in Chile and why has it not grown in recent years? Other chapters have examined the role of cyclical factors, such as microeconomic inefficiency, labor market rigidities and barriers to the entry and exit of firms. However, there is also evidence of what can be called "innovation failure" or a deficiency of ii~novation.~~ 3.4 Chile lags substantially behind comparator countries with respect to many indicators of innovation. By OECD standards, expenditure on R&D is low in Chile, the 44See OECD Science, Technologyand Industry Outlook 2004. 45See OECD, World Bank Institute,De Ferranti et al. 46 Ricardo Lagos, Discourse on the State of the Nation, May 21, 2005. (see http://www.gobiemodechle.cl/2lrnayo2004/indice~discursos.asp). 47See Box 1. private sector participatesvery little in R&D, few patents are issued4' and too much effort goes to basic research as opposed to applied resear~h.~'Chile also performs relatively poorly with respect to education and human resources devoted to research and development.50 3.5 However, innovation is best understood in the overall context of growth. Innovation analysis often assumes that the accumulation of knowledge is independent of other factors affecting a country's development.That, however, is not the case. Globally, there is a very strong positive correlationbetween TFP and the capital-laborratio which suggests that the accumulation of knowledge is complementary to, and driven by, many of the same forces that determine investment in physical assets.51 Hence, the low innovation indicators for Chile may simply reflect the country's relatively low level of income. In turn, that raises the question of whether the perceived shortfall in innovation results from problems common to overall growth, or whether innovation itself is sonlehow especiallyimpaired. 3.6 Policies aimed at the modernization of industry or the promotion of innovation should be pursued within the context of broader strategies for the growth of knowledge. There is a precedent for that strategy in the dual approach adopted by New Zealand -- upgrading its education and science and its technology infrastructure, but also focusing sharply on business mentorship, competitionpolicy, and incentivesto "get off the island" and export; in other words, creating a demand for innovation. Chile faces a similar challenge and, because of its small domestic market and its distance from larger markets, it is more like New Zealand than like Spain or Italy. When the latter two countriesjoined the European Community they were exposed to dynamic forces of competition, new ways of doingbusiness and the potential for expanding into broader markets. 3.7 Aside from issues relating to human and physical capital formation, Chile also appears to have an innovation problem. A model developed by Kleiiow and Rodriguez Clare (2004)~~shows that, given its level of human and physical capital formation, Chile has relatively low levels of TFP and, implicitly, high barriers to innovation. In other words, Chile has an innovation problem apart from a general problem of capital formation. 3.8 An important caveat is that the model cannot distinguish between static inefficiencies and innovation problems. Other literature in Chile has stressed the importance of microeconomic impediments to adjustment as an explanation for poor TFP performance over the last five years (Box 3.1). However, if the gap in TFP were due solely to adjustment costs, the inefficiencies would appear to be surprisingly large relative to the United States given the extensive micro reforms that have taken place in 48 This holds even when correctingfor income per capita. See Ledermanand Maloney (2003). 49 Chile can benefit enormously from the adoption and adaptation of new technologies through applied research, as Repetto and Bergoeing (2004) argue, and recent theoretical studies (Scavia (2004)) show that applied research has a higher impact on growththan basic research. 50 See Tokman and Zahler (2004) for a diagnosis on the differentaspects of innovation in Chile. 5 1See Rodriguez-Clare (2004). 52 See Maloney and Rodriguez-Clare (2005) for this exercise. Chile. Circumstantial evidence suggests that insufficient innovation was a barrier to growth in post-war Chile. Likewise, analysis in comparator countries such as Spain and New Zealand indicates that insufficient innovation is likely to be a barrier to future growth. This suggeststhat innovationdeservesto be a central issue in the policy debate in Chile. Resolving the impediments to microeconomic adjustment is a central part of the innovation agenda along with raising the capacity of firms to introduce productivity- improving measures. Thus, the agenda can go much further than a policy for supply-side science and technology. It is important to focus, in particular, on R&D. 3.9 Economic structure is onlypartly responsiblefor the low indicators of innovation. A recurring theme expressed in Latin America is that the region's dependenceon natural resource based industries constrains innovation because those industries have a limited potential for technological progress. Proponents of this view point to the experience of the Asian NICs (and to a lesser extent Finland) with an emerging comparative advantage in knowledge-intensive sectors such as advanced electronics. R&D investments in sectors that figure pronlinently in the Chilean economy tend to be lower than, for instance, R&D investment in the electronics or pharmaceutical industry.53There are, however, good reasons of comparative advantage that explain why Chile does not have a presence in those sectors, and it is far from obvious that Chile would benefit by fostering them. More importantly, within the OECD, only half the variance in aggregate R&D investment can be explainedby the sector structure of the economy. In Finland there has been a striking increase in R&D as a share of GDP and it is due to an increase of R&D investment across all sectors.54 Table 3.1: Aggregate R&D investmentwhen OECD sector investnlent rates are applied to the Chilean econon~icstructure (1995-99 average) Estimated R&D Inv. Estimated1 Country Rate using Chilean Observed rate (%) Observed Structure (%) Australia 0.7 0.8 89% Canada 0.7 1.I 65% Germany 0.4 1.7 26% Spain 0.2 0.5 51% Finland 0.8 2.1 37% Italy 0.5 0.6 84% Korea 0.6 1.9 33% Netherlands 0.6 1.2 51% Sweden 1.4 3 48% UnitedStates 1.1 1.9 57% Average OECD 0.8 1.4 57% Source: Rodriguez-Clare and Maloney (2005). 3.10 Taking Chile's industrial structure and applying to it the sector R&D investment rates of several OECD countriesreveals two important conclusions (Table 3.1). 53 Maloney andRodriguez-Clare (2005). 54 See Maloney (2005). 3.11 Structure matters. If Finland and Korea had the same economic structureas Chile, their aggregate investment in R&D would be two-thirds less. Even if Chile invested in R&D at a rate comparable to OECD countries, aggregate R&D would be almost 50 percent lower than the OECD average because the electronicsand transport industries are relatively small in 3.12 Even within the existing sector structure, Chile could investfar more in R&D than is presently the case. Some countries, like Australia and Norway, would still invest at roughly 90 percent of their present aggregate investment rate even if they had Chile's industrial structure. This suggests that a large part of Chile's disappointing TFP performance arises from low R&D investment in the sectors where it presently has a comparative advantage. Whattype of innovation does Chile lack? 3.13 Although the literatureoften uses R&D as a proxy for innovation, in practice it is not clear which measures of innovation best explain TFP and best serve to focus policy. Despite a large literature on the importance of R&D, it may be less important for LDCs than for more advanced countries because progress can occur by adopting improved business practices as well as by inventing new technologies. Even within the OECD, the role of R&D may be small compared to the introduction of new organizational and managerial practices.56The question arises, therefore, whether there is a single successful strategyfor TFP growth. Have strategies differed from countryto country? 3.14 Four commonly cited indicators of technological transfer are shown in Figure 3.1 and they point to a number of conclusions.57First, Latin America and the high-tech "miracles" have followed very different paths for dealing with R&D, FDI, licensing and education. Finland, Israel and Korea relied very little on FDI and very heavily on R&D and licensing of foreign technologies, a pattern that China is now following.58In part this reflects their specialization in electronic products which are especially R&D intensive. By contrast, Latin American countries, and Chile in particular, have followed a path that has relied little on R&D or licensing and heavily on FDI. This niay be problematic given the low rates of technological transfer associated with FDI, as documented by numerous authors. 55Ths is not to say that Chle should restructure its economy towards innovation intensive sectors. It is simply to say that, even if Chile were to lead the world in sectors where it has a comparative advantage, it may still have less R&D than other countriesjust because it specializes in sectors that are less innovation- intensive. Forcing specialization in sectors without an underlying comparative advantage can result in a large loss of social welfare, as the IS1experience showed for LAC. 56In the US, for example, Comin (2004) argues that less that 3- 5 tenths of 1 percentage point of the 2.2 percent annual growth in productivity is attributableto R&D. 57The analysis is based on data from a panel 60 countries with data adjusted for level of income. See Maloney (2005). 58Ireland's reliance on extremelyhigh levels of licensing reflects its unique position as an English speaking bridge between the US and the EU and hence it is a dubious model for the region. Box 3.1: An alternativeview of the TFP gap: microeconomicflexibility and the costs of institutional adjustment Although there is a large and growing literature stressing the importance of innovationfor long-run growth in productivity, recent research suggeststhat microeconomic rigiditiesplay a central role in explaininglow TFP in Chile and in some other countries. Parente and Prescott have argued that the capture of policies by special interest groups at the level of the f m and of the State can explain most of the difference in TFP between countries. Bergoeing and Reppeto (2003) contend that those finding fits well with the experienceof Chile. Chapter 2 of this report analysesthe most recent findingsconcerning the negative effects of microeconomic rigidities for flexibilityand productivity growth in Chile. It cites the importance of constraints in access to credit, the difficulty of closing firms and labor market rigidities - chiefly labor firing costs. It also cites evidence attributing more than 60 percent of manufacturing productivity growth to factor reallo~ation~~, which drivesthe creative-destruction process. Recent empirical literature has also shown how institutional adjustment costs (as opposed to technological ones) can hinder TFP growth. Caballero, Cowan, Engel and Micco (2004) argue that microeconomic flexibility is at the core of creative destruction, increased productivity and economic growth. One of the main costs is labor regulation. They estimate the effects of job security over the period 1980-98 using a sample of 60 countries. They find that, for countries like Chile with strong rule of law, increased job security lowers the speed of adjustment to shocks by a factor of thud and reduces productivity growth by almost 1 percent. Caballero et al. (2004) measure microeconomic flexibility as the speed at which firms close the gap between current and desired levels of employment. They suggest that the reduction in productivity growth after 1997was largely due to greater microeconomic inflexibilitywhich can potentially reduce long-run annual growth by 0.5 percent. However, the dichotomy between rnicroeconomic flexibility and innovation as alternative explanations of TFP may not be as sharp as is sometimes drawn. Barriers to competition in different markets may also inhibit innovation at the level of the firm. Parente and Prescott studied barriers to the adoption of new technologies -- an issue that is central in the innovation agenda. How do countries grow after all the existing high productivity firms have taken over? Innovation policy should ensure that new high- productivity firms emerge to replace the old ones. 3.15 Spain and Italy appear to be success stories closer to the Latin model. They showed little effort in either R&D or licensing and yet they attained relatively high levels of growth.60Spain, in particular, has converged rapidly to the income level of other West European countries over the last 30 years and it provides an important alternative to the Asian NICs as a benchmark for Latin America. 59See Bergoeing,Hernando and Repetto (2003). 60Italy's TFP growth in manufacturing is the third highest after Japan and Belgium in the 1970-87period, despite below-average R&D investment in most sectors (Bernard and Jones). Figure 3.1: "High tech" success stories: Deviations from predicted values for 4 indicatorsof technological advancement Finland Israel '1 Figure3.2: The Latin model . Argentina Chile 2 -2 1972 1977 1982 1987 1992 1998 1972 1977 1982 1987 1992 1998 Italy Spain 2 - -2 1972 1977 1982 1987 1992 1998 1972 1977 1982 1987 1992 1998 -- FDI R&D Royalties Education I Source: Maloney (2005). 60 Figure 3.3: Advanced resource-abundantcountries New Zealand Australia 1 -2-1 1972 1977 1982 1987 1992 1998 1972 1977 1982 1987 1992 1998 Canada Denmark 1 -2-1 1972 1977 1982 1987 1992 1998 1972 1977 1982 1987 1992 1998 FDI R&D - Royalties ' - Education L-. - 3.16 The experience of Italy and Spain suggest that there is a large potential for non- R&D sources of TFP growth such as improved organization and management. Repetto and Bergoeing (2003) argue that the best way for Chile to increase TFP is by adopting technologies from abroad. In both Italy and Spain, productivity gains were largely a result of moving the work force from low productivityto higher productivity sectors and, at the same time, increasing yet further the productivity of high productivity sectors through advances in organization and management.61Marimon (2002) suggests that innovations in processing and in organization (not separately identified in official statistics) have facilitated the growth of outward-looking industries such as banking, tourism and textiles. He argues that the system of Technological Centers in Spain has been central to enhancing the capacity of SMEs through clustering, networking and cooperatives. (Box 3.7). 3.17 More attention deserves to be paid to these lesser known aspects of innovation. Benavente (2004b) shows that innovations in processing, design, organization etc. are as common in Chile as are innovations that emerge from R&D (Figure 3.4). The data in figure 3.5 suggests that Chile is not so unusual in its pattern of innovation. Other countries devote a similar share of investment to innovations in processing. However, the similaritybetween Chile and the EU in that respect is no cause for complacencybecause expenditure as a share of turnover is very low (Figures 3.6 to 3.8). 3.18 Data limitations preclude a detailed disaggregation of innovation-related expenditures but it is likely that design is an important focus of innovation in Europe. In Italy, for example, sales of the fashion industry exceed those of the automobile industry. For countries such as Mexico or Colombia with a strong tradition of design, investments in those areas may make as much sense as iiivestment in S&T. In every sector of the economy in Chile, there are major differences from one firm to another which suggests that many of the less efficient firms have a potential for increasing productivity and T F P . ~ ~ Caselli and Tenreyro(2005). 62See Chapter 2 of this report and Bergoeing, Hernando and Repetto for evidence for Chile. This argument has been stressedby Dahlman (2004). Figure 3.4: ManufacturingEstablishmentsDeclaring that they have Made Innovationsin aspects of production and marketing (percent) Product Process Package and Orsign Organizational Wrapping b n a grmrm Source: Benavente (2004) Figure 3.5: Enterpriseswith successfulinnovation I I Enterprises with succesful innovation (as a % o f total) ....-..- Chile I I Source: Maloney (2005). Figure 3.6: Innovation effort 1 Innovationeffort (% of turnover 2000) Chile I I Source: Authors' elaborationbased on enterprise innovation surveys in Europe and Chile. Figure 3.7: Total innovation expenditures and total turnover I I Total innovation expenditure as a share of total turnover ManufacturingSector Total 10-49 50-249 0250+ - E z 2 V) I! u 'U. '< m rn P. Data for every countrycorrespondsto 2000 except Argentina (2001) ua Source: Authors' elaborationbased on enterprise innovation surveys in Europe and Chile. 64 Figure 3.8: In-house Expenditureon R&D, as a share of total turnover Expenditure in intramural R&D, as a share of total turnover Manufacturing Sector I * Data for every countrycorresponds to 2000 except Argentina I I Source: Authors' elaboration based on enterprise innovation surveys in Europe and Chile. 3.19 Investment in S&T and R&D is essential over the longer term. First, as argued by Cohen and Levinthal (1989), R&D is crucial not only to generate new knowledge, but to enhance "absorptive capacity" - the ability of a firm to use and benefit from existing information.This means that policies to strengthentechnology transfer have little value if industry does not possess the critical mass of human capital and highly technical skills to turn new knowledge into business opportunities. 3.20 Second, the recent slowdown of growth in Italy and Spain suggests that even though Chile would benefit from a broad approach to innovation in the medi~unterm, there is no substitute for investments in science and technology and R&D over the longer term. The Economist recently characterized Italy as the "Sick Man of Europe" because of declining TFP due, in part, to weaknesses in science and technology. Spain has been aggressivelyupgrading its R&D and its collaborationwith research centers in Europe to avoid a similar fate. The convergence club model of Howitt and Mayer indicates that, if Spain were to limit itself to adopting technologies from other countries, it would remain at a lower level of income than countries which rely on invention. That prospect also appears to haunt other countries even though they may have average levels of R&D. Denmark, New Zealand, Australia and Canada are increasinglyconcerned that they have become good "adopters" with little chance of staying at the front of the innovation game where the big gains occur. Blomstrom, Kokko and Sjoholm (2002) argue that a similar logic pushed Singapore towards a policy of knowledge generation in contrast to its previous reliance on technology adoption through FDI and licensing (as in Latin America). 3.21 In Chile, a long-term focus on S&T makes sense especially in sectors where the country has a clear competitive advantage,where it is already has a prominent position in the market and where M h e r advances in technology are key to sustain Chile's position - salmon production, metal mining, forestry, wine production and fruits. These sectors employ a relatively low embedded technology with potential for advancement and the key is to dynamically sustain their competitive advantage through R&D, especiallywhen confrontedwith low-cost emerging competitors such as China, India and Eastern Europe. Authors in Chile and elsewherehave raised this concernrecently.63 Summing up: Theneedfor a comprehensive view 3.22 Chile has an innovation problem that is manifest both in the low level of innovation, and in the way that innovation spreads throughout the economy. Thus, innovationmerits an importantplace on the policy agenda. 3.23 Innovation policy goes beyond science and technology. To the extent that a low level of TFP reflects a low pace of technological progress, the possible impediments to progress may include barriers to the creation of more innovative firms, deficient absorptive capacity at the level of the firm, internal or external credit market barriers to the adoption of existing technologies, and weak science and technology. Hence, promoting innovation requires a broad vision and a broad set of diagnostics that goes beyond science and technology per se. To foster innovation in this broader sense it is important to focus on the firm: developing entrepreneurial skill and a taste for innovation, and removing both innovation and non-innovation related barriers to productivity growth. 3.24 The capability of individualJirms is central to upgrading technology. Although the discussion of technological progress is ofien expressed in aggregate terms, productivity growth takes place at the level of the firm. If there is no demand for innovation on the part of the private sector, efforts to improve the S&T capacity on the supply side will be ineffective. Public spending on public goods will be "pushing on a string" and will not lead to productivitygrowth unless it is taken-up by the private sector. 3.25 Chile's private sector faces the challenge to prepare for this transition. Since larger firms are less likely to suffer from a lack of information, credit or other resources, it is reasonable to expect that their investment in innovation would be greater than small and medium enterprises. However, Benavente's work suggests that this is not the case (Figure 3.7 and Figure 3.8). These comparisons do not take into account of country characteristics or economic structures which are an important part of the story. They nevertheless suggest that large firms do not invest sufficientlyin innovation. This general observation also extends to training which, both in the aggregate and for Chilean large firms, is especiallylow relative to comparator OECD countries. 63See Eyzaguirre,Marcel, Rodriguez and Tokman (2005) and Tokman et. al. (2004). 3.26 It is important to go beyond increasing the number of researchfacilities and the stock of well-trained human capital and to improve the diffusion of knowledge through the economy. Evidence from the OECD suggests that, at the micro level, there are decreasing returns to knowledge generation while, at the aggregate level, there are constant returns to scale. That is because there are substantial spillovers of knowledge between individuals and firms. In Chile, however, there are diminishing returns to knowledge creation at the national level, suggesting that spillovers may be less prevalent.64The degree of spillovers appears to be correlated with such factors as the level of education, the perceived quality of research institutes, the degree of collaboration between research institutes and the private sector, and intellectual property rights. These are central aspects of policy design in the context of a National Innovation System. 3.27 Chile can introduce reforms that will lead to a more effective National Innovation System and to participate more fully in the international science and technology community. Even though, at present, less exotic sources of productivity growth can be harnessed, it makes sense to start now on the reform of the National Innovation System, particularly as it relates to Science and Technology System -- 'including institution building and program identification- because it may take several decades. The followingsectionexaminesthe agenda of options for reform. INNOVATION POLICYAND REFORM IN CHILE 3.28 As indicated above, fostering innovation requires a broad reform agenda extending beyond science and technology to the support of markets and business entrepreneurship (Figure 3.9). Figure 3.9: Basic componentsof a Core NIS and "extended" policy frameworkfor innovation 64See Bosch, Lederrnanand Maloney (2005) 3.29 The following agenda begins with a narrow agenda focused on long-run reforms and then proceeds to the elements of a National Innovation System (NIS), focusing on R&D and issues relating to the disseminationof technology. Addressing the innovation problem 3.30 Market failures with respect to invention and the dissemination of knowledge are frequently cited as a justification for public sector involvement in innovation issues. The most commonly cited failure is the inability of the market to protect intellectualproperty rights. Considerable effort may be needed to discover a new product or establish the appropriatenessof a new technology for Chile and low-cost imitations can quicklyreduce benefits for the individual innovator. While spillovers are good for growth, insufficient protection of intellectualproperty rights discouragesinvestmentin both the invention and adoption of new technologies. 3.31 In the OECD, the gap between social returns and private returns to R&D are thought to be so large that investment is only a quarter of its optimal Even the adoption of already known technologies (for instance, the salmon industry) may entail large scale investments and long gestation periods that are beyond the capacity of any individual firm. Inadequate resolution of these market failures constitutes an innovation- specificbarrier to technologicalprogress. 3.32 The OECD itself has not yet reached a consensus on the optimal policy frameworkto address these market failures. Even if there were a consensus, it may not be applicable to middle income countries with more serious institutional constraints.Hence, rather than propose specific policy measures, the next sections seek to explore general principles and considerations, common to OECD and middle-income countries, that may serve as a basis for Chile to deepen and improve its policy framework. Basic Tasks of the NIS 3.33 The NIS could usefully provide a policy framework for innovation with a particular longer-term focus on S&T. Redressing market failures requires establishing institutions and policies whose efficacy and efficiency depends on a coherent overall strategy. There are two reasons for this. First, many innovation institutions - universities, public research institutes, grant giving agencies - are non-market institutions. They are not subject to market discipline and coordinationbetween them is weak. Hence, Finland, Japan, and New Zealand have put a strong emphasis on Government coordination of the national innovation system. Second, private and social returns may not always be the same. Subsidies may need to be provided at the margin and targeted at areas where the country may develop a comparative advantage. The identification of areas with a potential comparative advantage is best achieved through an integrated approach at the national level, and is only possible through a national innovationpolicy. 65See Jones and Williams for an overview. 3.34 The NIS can foster an environment consistent with capital formation more generally. It could redress bottlenecks in such markets as credit and labor and it could encourage competition. 3.35 The NIS could encourage innovation as well as the entrepreneurial and scientzJic capacity required to effectively innovate through the acquisition, adoption and adaptation of new knowledge. The latter is particularlydifficult since there is only a very limited understanding of how to increase a firms' appetite for innovation or its absorptive capacity. Achieving a "critical mass7' is important both to foster innovation and to develop the human resources that can adopt and create useful knowledge. The private sector stands at the center of effective innovation policy in order to ensure the link to productivity. 3.36 The NIS could encourage both research into new products and the broader adoption of improved technologies into existing industries. A different policy framework is appropriate to each of those objectives. 3.37 The NIS can ensure thatpolicies and programs are subject to constant evaluation in order to eliminate ineficiencies and redundancies. Finland, with a highly regarded record of innovation, subjectsall its principalprograms to regular outside evaluation. 3.38 The overall structure of the NIS may be less important than well-designed incentives within individual institutions. Successful countries have experienced a wide variety of initiatives and programs many of which have grown out of earlier industrial structures as opposed to being designed from a "tcEbula rasa". There are widely differing styles and approaches and there has been little evaluationof which mix works best. 3.39 In Chile, as in other parts of Latin America, the balance between market mechanisms and Government intervention would need to be managed carefully. The size and nature of market failures with respect to innovation are very difficult to establish. Yet, most advanced countries behave as if such market failuresneed to be corrected, and severalhave given great attention to incentivesand measures to addressmarket failures at their source. As potential models, New Zealand has a comnlitment to the market similar to that of Chile - as well as a similar economic structure. However, its approach to the innovation agenda is more characteristic of the OECD. Thirty years ago, Finland had a level of education and income similar to Latin America but is has since become a highly competitive economy by creatively exploiting its resource base and making far-ranging reforms in its NIS. Diagnosis of Chile's NIS and the Needfor Reform 3.40 Chile's institutionalframework for innovation lackspolicy direction and strategic goals. Diverse studies and assessments have warned about the lack of an overall policy framework, coherent organizational structure and policy coordination in Chile's N I S . ~ ~ As Figure 3.10 shows, several ministries - including the Ministries of Education, 66See for example, Eyzaguirre et.al. (2005), Tokman et. a1(2004),Benavente (2005). 69 Planning, Agriculture, and Foreign Affairs - are involved in the formulation of S&T policies, the design of programs and the management of research in Chile. Innovation policy is defined each year through the budget process, not by a coordinating body that defines a medium- to long-run policy. As a consequence, Chile has a wide array of uncoordinated policies and programs for R&D and for technology diffusion with significant overlap between content and objective^.^^ This reduces the effectiveness of public expenditures, results in duplication and leads to excessive diffusion of programs including to some areas with relatively low social rates of return. The National Commission of Science and Technology (CONICYT) - the entity mandated by law to define an S&T policy -is, in practice, unable to coordinate effectively. Figure3.10: Present organizationof supportfor innovation Source: Adapted from Benavente (2004). 3.41 The lack of an innovation strategy and a coherent policy framework has also hindered the effective evaluation of existingprograms. To date, evaluationshave focused narrowly on the disbursement of funds and on the internal consistency of individual initiatives. Only recently have there been attempts to assess broader program effectiveness. In an evaluation of matching grants, Benavente (2002) found positive effects in terms of encouragingprivate R&D. He found that firms which used matching funds were more likely to experience improvements in productivity and en~ployment compared with those that did not. He also found that the overall NPV was positive for 67 A well-known case is the overlap between FDI (CORFO) and FONDEF (CONICYT) or the similar structure and objectivesof FONDAP Institute (CONICYT)and The MillenniumInitiative (MIDEPLAN). matching grants. On a more systemic basis, however, there are no evaluations of policies and performance for the NIS as a whole. 3.42 Chile's NIS would benefitfrom an overallstrategyfor science and technology. An overall strategywould help to prioritize reso~rces~~,clarify the goals of funding agencies, strengthen coordination between them, provide stronger incentives for public-private collaboration; and strengthen data collection on R&D for purposes of evaluation. Such a strategy is more likely to be effective if all stakeholders - including representatives from industry - are involved in its design and implementation. A coherent policy is the most effective way to redress and improve the perfornlance of the different tasks of the NIS highlighted above. 3.43 Benavente (2004) has analyzed options for an institutional reform, based on international experience. One possible alternativeis shown in Figure 3.11. A Ministry of Science and Technology may be able to reorganize institutions and define a clear and feasible policy, but a better alternative would be to establish a directive Council with authorityover policy definition,implementationand evaluation.Such a structurecan lead to better overall coordinationand less redundancy, at a lower political cost. 3.44 There are different institutional arrangements that have been adopted by successful innovative countries. The key lies not in a particular framework, but in the incentives defined to effectively overcome the market failures that hinder the innovation process. Pigure 3.11: Possible alternative structureof support for innovation Source: Benavente (2005). 68 Between for example, basic and applied research, horizontal policies and strategic sector resource allocation, incentives for R&D and capacitybuilding, etc. Box 3.2: Discovery and Deepening In countriesthat have a record of successfulinnovation, the NIS typically encompassesinstitutionsdevoted to the discovery of new products, either independently invented or adapted from elsewhere. It also has other institutions that seek to deepen innovation in existing industries with a mix of vertical and horizontal policies. The appropriate balance between the two is a matter of trial and error. A good example is Finland's TEKES which divides its activities into "projects" that seek to explore and evaluate new areas, and "programs" which seek to advance productivity growth in existing sectors or clusters. Policy reform would clarifythe distinctionbetween fostering mscovery and deepening the adoptionof new technologies. Chile lacks Finland's comprehensive policy framework. Both horizontal and vertical policies are defined and executed within particular funding agencies. The Chile Innova Program under the Ministry of Economy is the only real effort to prioritize strategic areas and to coordinatefunding for inno~ation.~~That program has also attempted to identify sectors with a current or prospective future competitive advantage. However, funding for the program amounts to only $3 million. Fundacidn Chile ranks among the most successful institutions in the developing world for fostering discovery. It redresses several key market failures: identifying possible products appropriate to Chile; undertakingresearch in needed complementary technologies;and providing sufficient long-run financing to successfullycommercializea product for purposes of denionstration. Where there may be room for experimentation is in clusterpolicy. That could involve helping the f m s in a given cluster to address bottlenecks that hinder the further expansion of a cluster. This is a central activity of TEKES with a focus on increasing networking among members of the NIS. Chile has made relatively few efforts in this area, but it is beginning to address clusters through its new consortia program. Rationalizing NIS Institutions: a Micro View 3.45 Recent literaturehas examined the constituent parts of a typical NIS and how they relate. Link and Scott (2004) and Hall (2005) both reviewed market failures and the advantages and disadvantages of particular types of Government intervention. Hall, in particular, reviewed numerous issues concerning the design of institutions to foster innovation. She argues that, once innovation is recognized as a public good in part, then policy design becomes a matter of weighing the cost and effectivenessof taxes, subsidies, and other instruments for "internalizing the externality," including intellectual property. She divides the policy space in two: a. Government sponsored innovation related tasks: Governments often directly fund or undertake basic research which may have a high social value but low private returns. Universities and public research institutes (PRIs) may be engaged to identify and evaluate areas with prospectivelyhigh socialrates of return. b. Private sector-led tasks: Private initiatives are preferred where governments are unable to identify projects and where the sociallprivate gap is smaller. The Government may support private initiatives with tax credits, intellectual property generatedrents7' and public-privatehybrid arrangementssuch as matching-grants. 69It currentlyprovides support for biotechnology, IT, clean production, and quality improvement. 'OThe Bank has recently published a book entitled IntellectualProperty and Development by Maskus and Fink (2005) which addressesthat issue in some depth. See also Maskus (2000). Universities 3.46 Universities develop human capital, undertake long-term, high-risk, basic research, and collaborate with partners in the business sector. In most of these respects, Chile is lagging. 3.47 InsufJicient development of advanced human capital. The number of research- trained scientists and engineers remains low in Chile. In the prestigious 25 Council of Rectors Universities (CRUCH), students at the graduate level comprised only 4 percent of the student body in 2003 (DESUP 2005). In the same year, only 144 PhDs graduated from Chilean universities. However, this number is likely to increase in the near future due to efforts by CONICYT and the Ministry of Education to strengthen doctoral programs through the MECESUP program. Between 1999 and 2003, enrollment in PhD programs almost doubled from 1,144to 1,930. The effect upon the economy will depend on the availability of appropriate employment opportunities for PhD graduates, particularly in industry. Doctoral students have currently few incentives to orient their research towards a more diversified career, which includes training in such disciplines as business and technology management. 3.48 Toofew researchers. Weak graduate education translates into a dearth of human capital for research and development. Measured in full-time equivalent, Chile employs about 5,700 researchers, 700 research assistants with fellowships, and 4,700 supporting and technical staff. This places Chile at a clear disadvantage compared to OECD benchmarks. Whereas Chile had 0.47 full-time equivalent RR& personnel per 1,000 population, the corresponding number for Finland and Japan was 10.16 and 7.07, respectively (IMD 2003). 3.49 Academic excellence is paramount. Despite their prominent role, universities are severely constrained by insufficient faculty with advanced education, and by a lack of research equipment and laboratories. Less than 17percent of universityprofessors hold a PhD, and only about 24 percent of lecturers conduct research (DESUP 2005). The salaries of full-time academics are low and it has been traditional practice for teachers to hold more than onejob. Hence, there is little incentive for academics to use their time to do research. Unappealing retirement benefits discourage the retirement of faculty members and the hiring of new talent. Efforts to stimulatequalityresearch are slowedby deep-rooted practices and compensation structures that emphasize seniority rather than performance. 3.50 Publications are fewer than expected. Scientific publications are an imperfect indicator of R&D because they do not always lead to productive innovation. Figure 3.12 shows publications relative to income for a sample of 123 countries. It can be seen that the number of publications in Chile is slightly below the level expected on the basis of per capitaincome. 3.51 Chilean researchers are productive. The lower-than-expected number of publications in Chile reflects the short supply of scientists and engineers. Figure 3.13 shows the annual SCI publications relative to the n~~mberof full-time equivalent (FTE) researchers for a sample of seven countries. In 2000 Chile featured about 41 articles per 100 FTE researchers. This makes Chilean researchers more productive than their counterparts in Canada, Spain, and Latin America. In fact, Chilean researchers publish about twice as often as their colleaguesin Argentina, Colombia and Mexico. Figure 3.12: Publications per million people relative to GDP per capita, 1999 - 3.00 2.50- Hungary 2.00- Bulgana Argentina Gambia orocco Pananla 1.00- Kenya Thailand Guinea-Biss . . Colombia Bohv'a 0.50, Malawi /-,in 16Cl Honduras Ecuador Sudan Guatemala 0.00 2.00 2.50 3.00 3.50 4.00 4.50 Log ofper Capita GDP, 1999 Note: Some countrynames have been omitted to make the graph more legible. Source:Author's calculationbased on data from NSF 2002 and WDI 2003. Figure 3.13: Scientific articles per 100 FTE researchers, 2000 I Chile Canada Spain Brazil Argentma Colombia Mexico Source: RICYT 2005. Note: Based on publicationsregistered in SCI search. 3.52 Potential for improving the quality of research. The average country citation impact factor (ACCIF) for Chile improved slightly between 1990 and 1999 from an average 1.1 to 1.4 citations per publication (Figure 3.14).~'Chilean scientists are cited at the same level as their colleagues in Latin America. However, Chile performs considerably below countries such as Spain, Australia, Finland and particularly the United States, suggesting a gap in the quality of scientific output between Latin America and high-income OECD countries. Figure 3.14: Research citations, 1990 and 1999 W e d Fmknd Austral$ Span C h k Argentina Meldca Cobmbia Brad States Source: Author's calculationbased on date fiomNSF 2002. Strengthening University-Private Sector Collaboration 3.53 With particularreference to Latin America, recent work by Banda et al., Canton et al. and Thorn et al. focuses on the incentives to createknowledge and link it to the private sector. Several conclusions and policy implications for Chile emerge from interviews with universities as well as standard data analysis. 3.54 The universities' mandate can benefit from greater clarity. Universities are the principal agents of research (45 percent of R & D ~ ~and ) the principal employers of researchers (60 percent of all researchers), but they contribute little to the policy agenda and, more particularly, to turning science into business. Given that more than 55 percent of finance for R&D is dedicated to basic research (compared with 18 percent in the US), there is a clear need to clarify the mandate of the universities. They would benefit from a practical vision of university research as the handmaiden of industry. Well focused universityresearch would support the country's comparativeadvantage. 7'ACCIF is defmed as the ratio between citations of Chilean research articles received in one year and papers published in the two previous years. By allowing for a time lag and using two years as a reference, a rough match is made between articles and citations. 72The relative participation of universities has steadily risen from 27 percent in 1990. Data of this paragraph comes from RICYT. Box 3.3: The Importance of Collaborationbetween Universities and the Private Sector Canton et. al. (2004) identify three principal shortcomings in higher education in the OECD that may hinder the transfer of knowledge.All three have their parallel in Latin America: Mismatch between the knowledge supplied by universities and that demanded by businesses. Inward- oriented universities may be the cause of the mismatch. In the USA the system is more competitive and de~entralized.~~In that regard it is similar to the Finnish approach where academic departments receive limited basic financing supplementedby competitive financing and incentives to undertake work useful to business. In Chile, nine cooperative research consortia have been established and they could provide an alternative and equally successfulmeans of providing a link with the business community. Ineffective reward structure for academics. Many of the incentives that promote creativity may, to some degree, deter researchersfrom addressing the needs of business. An emphasison publishing,while essential for guaranteeing quality of output, does not necessarily facilitate disseminationsince local industry would benefit more from direct contact than from publications. The opposing cultures of disclosure in the academic world vs. protection of trade secrets in the business world can work against each other. However, the importanceof this effect is mixed.74 Lack of an entrepreneurial culture. Even if the inventor of an academic idea has the ability to take it to market, there is often little incentive for him to pursue the idea beyond "proof of concept". In the United States, it does not appear that encouraging the greater commercialization of results will lead to less fundamental research. US policies have included the development of science parks and Technology Transfer Offices (TTOs) designed to develop networks of industrial partners set up guidelines for the commercialization of research findings and the manage university intellectual property. The latter, in conjunction with their royalty sharing systems, appear to have increased commercializationparticularly in private universities,which are less constrainedby regulations. 3.55 Chilean universities could be made more accountable to an external audience. Two reforms would help in adapting the university toward a more "innovation friendly" posture. Expand stakeholders in the governance structure: Few universities have a concrete plan for contributing to national development. University boards and academic councils are mostly dominated by internally elected academics answerable to staff and students. Expanding the role of other stakeholders, especially the private sector, can shift incentives toward making research responsive to the needs of the productive sector. In this regard, lessons can be learned from OECD countries, many of which have made it a requirement to include private sectorrepresentativeson the managingboards of universities. 73The US has a successful system of Industry/University Cooperative Research Centers, small academic centers designed to foster research with strategic interest to industry. They require proposals to be jointly agreed by industryand the university. 74 Experience with the Bayh-Dole act of 1980 that allowed universities to patent and license inventions from federally funded researchhas been inconclusive although generally positive. When a university shares the licensing fees from their inventions it encourages research. However, there is some sentiment in the OECD that IP incentives are distorting the role of the university. Increase the role of competitivefunding: A stable source of basic, untied f~mding is essential for the most theoretical, long term, risky research projects which are unlikely to have a commercial application, at least in the medium term. However, countries can influence the thematic composition and the modality of research by shifting funding away from unconditioned basic salaries to competitive funding, whether public or private. In Colombia, a new university fimding system places a strong emphasis on the relevance of research for society. New Zealand moved from allocating the educational budget on the basis of "bums on seats" or the number of students, to objective measures of academic productivity and quality. Chile is in the process of introducing institutional performance agreements for publicly funded universities that tie "Aporte Fiscal Directo" to the results achieved in such endeavors as strengthening collaborative research between universities and industry. Shifting funding in part to a competitive basis makes university research more accountable to the needs of the private sector and may encourage more applied research.75To be politically feasible, this shift could be implemented on an incremental basis. Such policies would need to be routinely evaluated externallyto make sure that they achieve their objective. Building a critical mass. Further, even where countries have numerous world- class academics, there is often a tradition of insularity between institutions that prevents them from establishingthe critical mass necessary to progress in a given field. Sometimes this may be due to the idea that every field needs to be represented in every university.In Chile recent data shows that bigger universities produce more publications not only in aggregate terms, but also in relative terms, when adjusted for the number of full-time equivalent research staff.76This is only to be expected because bigger universities are often well-established and more research-oriented compared with smaller teaching-oriented universities. The well- established universities are also better able to attract the most productive scholars and build a critical mass. Excellence in research is necessary if the private sector is to use university-produced research. The Millennium Science Initiatives in Chile, Brazil and Venezuela are examples of how large competitive research grants can create a critical mass that extends beyond individual universities and that contributesto an environment for excellence in Latin America. Recent UniversityDevelopments in Chile 3.56 Universities have begun to develop institutionalpolices for IP. Using data from Chile, Thorn and Holm-Nielsen find a correlation between institutional polices and 75The down side of excessive reliance on competitive financing is that it may add hrther to the volatility of financing and make it more difficult for universities to become a credible research partner for the private sector. An appropriate balance between both types of research is important and should be guided by international experience. 76Indeed, the quality of the research staff is directly related to research productivity. The share of academic staff with PhD degrees and the number of PhD graduates has a statistically sigmficant effect on the publicationrate. Interestingly, R&D expenditure per full-time research staff and publication activity is not significantly correlated. See CONICYT (2005). patenting. University of Concepcion is a case in point: The University actively protects intellectualproperty generated within its research program and is also a leader in terms of the number of patents issued per researcher. Nevertheless, regulation of intellectual property rights between universities and firms remain inchoate, and this is an area in need of attention. Universities in Chile are not very active in claiming patents. The total number of patents has increased slowly but steadily over the last decade, but the number of international patents for individual universities still ranges from zero to four. Among the 25 traditional institutions that form the backbone of the university system, only nine have any patents at all. It is arguable that patenting should not, in fact, be the primary objective of universities in developing countries if their role is conceived more as disseminators,rather than creators,of knowledge. 3.57 There is growing interest in incubators. The most successful example is found at the Engineering department of the Catholic University in Chile, which has supported the establishment of 5 new spin-off companies.However, in general, universities do not have a tradition of stimulating entrepreneurship. PhD students in science and engineering generally have few incentives to orient their research towards broadly-based careers, which could include training in such disciplines as business and technology management. This is part of the reason why most young researchers stay in the public sector, working either for Government research institutes or universities. 3.58 There is a lack of strategic thinking with respect toprogram evaluation: In none of the universities interviewed had there ever been a formal evaluation of policies to foster a linkage with industry. Evaluation is particularly important since initiatives such as incubators can be very costly. Lack of a long-term perspective is also evident in the absence of institutionalstrategiesfor the disseminationof knowledge. 3.59 Incentives to establish linkages are still weak. When universities in the region undertake an appraisalof their performance,they are slow to recognize that linkages with industry are an important aspect of performance. liicentives such as laboratory space, access to young researchers etc. remain limited. Generally, graduate students are able to participate in private sector activities. Again, the Catholic University School of Engineering in Chile is held up as the best in terms of promoting interchanges. Thorne and Holm-Nielsen also note that Chile and Mexico run programs that offer partial scholarships to post-doctoral or other early career researchers who undertake research in industry (Box 3.4). Box 3.4: Policiesto link researchand industry-The Mobility of Human Capital "Human mobility" refers to the abilityof technicians and scientiststo move between institutions in the NIS. Mobility strengthens innovation within firms and enhances the transmission of knowledge from science-to- business knowledge. It also has the potential to create entrepreneurs who are technology-literate,who know where the technological frontier is and who can increase a firm's research capability in a constructive way. One policy option is to place researchers in industry, i.e. sponsor the employment of young researchers in industry or provide scholarships to doctoral students who undertake their thesis work in industry. In 2004 Chile, with the support of the World Bank, launched a program that awards competitive scholarships to doctoral students who wish to undertake a substantialpart of their thesis work in industry.A staff member of the company acts as an associate supervisor of the student and the company contributes a small supplement to the scholarship. In addition, the program offers partial scholarshipsto post-doctoralor other early career researchers who undertake research in industry. The scholarships are temporary and they are progressively phased-out over time, with the company taking an increasing share of the researchers' salary. So far, 17 researchers have been selected for placement in industry. The program is being publicized to increase interest among companies. An impact evaluation is being developed in order to test the effectiveness of the placement program in Chile. The key question is to what extent the program boosts company revenues from innovation and how many researchersfind permanent employment in industry. Experience from the OECD shows clear benefits from increasing exchangesbetween university and industry. Public Research Institutes (Pa) 3.60 The role of public research institutes in the NIS varies greatly from country to country, ranging from the relatively small in Swedento the protean ITRI in Taiwan. This section will consider the rationale for PRIs in terms of the market failures they are designed to address as well as important issues concerning their financialstructure. Market Rationale for PRIs 3.61 As Link and Scott (2004) argue, PRIs are most relevant where there is a need for an agency that can work with the private sector, but is independent of it. This applies to setting standards, to disseminating technologies and to coordinating institutions within the NIS (Box 3.5). In practice, some functions, especially dissemination, have been discharged by universities (for instance, agricultural extension by the Universidad de Chile). However, a separate entity may be called for. In practice, the problem in Latin America has been a lack of clarity of exactly what market failure the existing PRIs are redressing, how the outputs would be measured, and what incentive structures would be appropriate. An evaluation of PRIs from this perspective would be useful in Chile, since to date there are no comprehensive evaluations of how well they bridge the gap between private industry and the academic sector. In addition, the social return of their research deserves to be evaluated as well as those aspects of their output that may be considered a public good. Box 3.5: Market Failure and the Rationale for PRIs Link and Scott (2004) argue that there are three circumstances where PRIs with a research capability in technology and innovation can be the most efficient way to address market failures.Given the nature of the research they perform, the PRI may be considered an institution for "standards and infrastructure technology." The promulgation and adoption of standards would reduce risks for businesses that need to acquire new technologiessuch as software and high-tech equipment. The key issue here is that the standards need to be set by an "honest broker" with no market interest in the outcome. Hence the market itself cannot provide the service.Nor would it be an appropriatetask for academic institutions. Extension services to transfer technologies and management techniques to sectors such as light industry or agriculture are another appropriate task for PRIs. In those sectors small producers predominate. There may be imperfections in the credit markets, predatory behavior by larger firms and difficulties in effectively protecting intellectual property rights. The authors conceded that the case for an independent PRI to engage in those issues is not as strong as in the case for a PRI to manage the promulgation and adoption of standards. Nevertheless,there may be synergies if, for instance, appropriatetechnologies were developed jointly with the PRI. In the USA, public research institutions often stand mid-way between the basic science of universities and the applied needs of industry. They are thus uniquely suited to develop and transfer generic technologiesto those who need them. In Spain, Marimon argues that the universities have been weak in dissemination and that, instead, technological centers have largely played this role. The Government's role has been to coordinate and to provide modest levels of financing where the private sector is responsive. The sponsorship of cooperative R&D efSorts joining industry, universities, and Government in research projects subsidized by the Government is another task suitable for PRIs together with the oversight of private research that is partially hnded from public resources. A PRI is often needed to coordinate the development of infrastructuretechnologies as well as generic technologies for complex systems involving high cost and high risk. Again, the Technological Centersin Spainhave increasingly played this role. All these tasks require a neutral player that works with, but is not part of, the private sector. Public research institutions are most likely to be successll in correcting underinvestment when they focus on specific projects that addressparticular market failures. 3.62 A second question is how Chile has performed in fostering entrepreneurialtalent on the pure business side to advance these industries, once they are established. Chile seems to face a problem similar to the one Trachtenberg identifies in Israel - that successful ideas are purchased by foreigners and not developed locally. A developing country may not be able to take a new product all the way to market because distribution and other supporting networks are lacking. Hence, the price it receives for a successful idea is far less than would otherwisebe the case.77In that context, the role of technology instit~~tesas intermediateplayers between PRIs and firms, such as those in Spain, may be worth exploringmore in Chile. Competitive vs. Base Financing 3.63 For PRIs, the structure of funding is a central issue affecting incentives. That, in turn, is closelyrelated to the function of the PRIs within the NIS. The appropriatemix of competitive and base funding depends upon the extent to which the program includes lumpy, long term projects with a high rate of social return. 3.64 Competitive finance, particularly when it is linked to the private sector, has the advantage of getting the incentives right in terms of maintaining quality, encouraging linkages and aligning research with established national goals. However, competitive procedures are time consuming for the researcher, they may lead to a fragmentation of the research agenda, and they make it more difficult to take a long run view of high-risk projects. Hence there needs to be an appropriate balance between competitive finance, larger multipurposegrants and some degree of base financing. 3.65 New Zealand offers an example of erring too far on the side of competition. It reduced the base financing of its Crown Research Institutes (CRIs), leaving them to compete for private sector funding. Predictably, the CRIs behaved increasingly like consulting firms, focusing on short-term projects and giving short shrift to the public goods aspect of their charters. To redress the imbalance, new funds have now been developed to provide long-termblock grants for more basic research by the CRIs. 3.66 The Finnish VTT combines about 25 percent base financing with 25 percent private grants and 50 percent TEKES-financed matching grants. It requires some experimentation to get the financing right and to strike the right balance between quality control, responsiveness to social and business needs and the need to accommodate loiig- term risky projectswith higher socialrates of return. 3.67 In both Spain and Chile concern has been voiced of an excessive bias towards competitive financing. In Chile, most of the PRIs were transformed into legally private institutions, although with strong links to specific ministries (with the exception of Fundacidn Chile). Base hnding was severely reduced in the 1990s to provide an incentive for greater links to the private sector. An excessive reliance on competitive funding may have been at the cost of riskier, long-tern1research. In the past, Fundacidn Chile was able to identify new products suitable to Chile, undertake research to develop them, and introduce them to the market. It thereby played a central role in the discovery and introduction of new products, most famously the salmon. Now, Fundacidn Chile has private hnding and the fundamental question is whether there are other long-term investment projects in their portfolio that could be undertaken if greater financial resources were available. A bias towards competitive funding sometimes seems to have unintended effects because PRIs have had to behave like private entities seeking to maximize revenues instead of addressing broader issues of dissemination and private innovation. ThePrivate Sector as a Leader of Innovation 3.68 Private sector initiativesmay be preferable in areas where the gap between private and social returns is small and where governments may not easily be able to identify projects. As Hall (2005) indicates, two main instruments are commonly used to help companiesincrease their investmentin R & D . ~ ~ (i) Tax CreditsISubsidies 3.69 Tax credits lower the cost of innovation without specifying the sector, thereby leaving the evaluation of profitability to the individual firm. The econometric evidence suggests that one dollar in tax credits stimulates one dollar in additional private investment. 3.70 Nevertheless, tax credits have several drawbacks. First, tax credits typically apply to increnieiital R&D expenditures over and above a specified base level. This avoids subsidizing investments that would have been made anyway. Experience in the USA indicates that defining an appropriate base level of R&D expenditures can be problematic. Second, tax auditors frequently lack the expertise to distinguish between true R&D and other expenditures that firms may claim. Third, private and social returns to R&D expenditures may diverge. Private firms are likely to choose projects that have the highest private rate of return, whereas the Government may prefer those with the highest socialreturns. (ii) HybridsIMatchingGrant Schemes 3.71 In the case of matching grants, the private sector proposes projects and contributesto the cost. The Government co-finances a share of project cost and nionitors project performance, sometimes limiting its participation to strategic predefined sectors. Matching grant schemeshave several advantages. 3.72 They encourage networking. When the private sector presents a project proposal jointly with an academic or other research institute, the matching grant creates incentives for collaboration. Such contacts can lead to other linkages and can stimulate entrepreneurshipamong academics.Finland abandoned tax credits a decade ago precisely for this reason. 3.73 They bring together complementary expertise. The public sector may have neither sufficient technical capacity to evaluate the scientific merit of many R&D projects, nor the business experience to evaluate their commercial viability. Matching grant schemes mobilize the expertise of the private sector in that respect subject to the constraint that private returns may not always correspond to the social returns of a project. Finland's TEKES attempts to identify which areas are likely to generate the greatest social "bang for the buck". It maintains regular contact with MIT, the NSF and other global centers of cutting-edge research to identify promising new areas. That requires a very high level of skills on the part of the funding agency which may not, as yet, be available in many developing countries. Nevertheless, CORFO has developed important know-how in the last 15years and may be able to fulfillthat role. ''PlusIPR policy which, as alreadynoted, will not be discussedin this document. 82 3.74 They establish incentives for non-market institutions. Matching grant schemes have proved useful in guiding PRIs and universities towards socially useful research. However, client institutions must be capable of understanding the potential benefit of their services. 3.75 They may open up better ways of allocatingfinancial resources. As Link and Scott (2004) note, companies may compete for matching grants, but the amount of the grant is determined administratively.In theory, the amount of the grant could instead be determined by auction. Do These Programs Work in Chile? 3.76 Public support for innovation in Chile has relied on co-financing through grants, and currently there are no tax incentives for innovation. Until the 1980s, the main instruments of support were supply-side subsidies, i.e. through budget funding of technological institutes. Today these institutes, together with firms, universities and researchers, compete for a diverse array of demand-driven subsidies offered mainly by CORFO and CONICYT. They provide co-financingbetween 30 percent and 100percent, depending on a number of factors including the size of the project, the involvement of universities and technological institutes, and the ease with which project benefits may spread throughout the private sector. These fimds are allocated mainly through "open window" requests relating to specific areas of strategic interest. The eligibility of a project proposal for co-financingis determined on the basis of an independent evaluation which takes into account technical, economic and strategic considerations. 3.77 Matching grants stimulate R&D and increase productivity: Benavente (2004) examined the impact of severalmatching fund programs. He examined in particular detail FDI and FONTEC, both from CORFO. His findings are consistent with Hall's survey. First, public co-financing of innovation encourages R&D. Second, an evaluation of the finance provided by CORFO reveals a positive NPV and shows that firms which receive the finance increase their productivityfaster than firmswhich do not.79This indicatesthat the funds are accomplishing their objectives. 3.78 Getting the right mix of co-Jirtancingis tricky. In Chile, there is a strong case for the private sector to increase its share of co-financing and to contribute financiallyrather than in kind to better ensure the commitment of the private sector and to better fulfill the "commercialization test". Some 78 percent of the firms participating in CORFO's FDI- financed projects contribute only with non-monetary resourcesg0and in many cases the public share of project cost is greater than 60-70 percent. That clearly casts doubt on the depth of private sector commitment. Yet Thorn et al. (2005) state that World Bank and 79 In particular, an evaluation of grants from FONTEC found that the beneficiary firms significantly increased sales, employment, and investment in machinery. Nevertheless they did not show a significant difference in innovation compared with a control group. The most recent evaluation of FDI shows that firms and institutes that benefit from its hnding undertake more R&D than those who don't. However, it did not result in more innovative processes or products. See Universidad de Chile (2004a) and Universidad de Chile (2004b). Universidad de Chile (2004), "Evaluacidn Impacto Fondo de Desarrollo e Innovacidn ". IDB supported schemes in Latin America have generally provided less than the 50 percent co-financing that is standard in the OECD. They cite a lack of private sector capacityto utilize the funds. Experimentationwith alternativeco-financing arrangements, or even with some type of auction scheme as mentioned above, could provide useful evidence on this issue. 3.79 The benefit of a strategic vision. The advantage of hybrid funding over tax credits/subsidies is that they allow public financing to be targeted at areas where the social return is especially high. However the funding programs need to be placed within an institutional context that permits a coherent evaluation of the most socially valuable priorities. The decision to restrict Chile's matching grants to mid-size projects, limits the opport~lnitiesfor larger-scale innovative projects. The decision would have ideally emerged from a coherent overall strategy,but in all likelihoodit did not.81 Beyond S&TProjects 3.80 As previously noted, TFP can be increased through the dissemination of known technologies, improvements in management, and the elimination of obstacles to the adoption of new technologies. This section considers selectiveissues in that regard. (i) Human capital formation 3.81 Chile has relatively high standard of education by regional standards. The Chilean population aged 25 years and older has 7.9 years of schooling, which is well above the average for Latin America (5.8 years). A relatively high 36 percent of the Chilean population has had some secondary education and 16 percent has had some tertiary education. Nonetheless, a significant proportion of the population lacks basic skills. Forty three percent of all Chileans have no more than "some primary education" (De Ferranti et al. 2003). 3.82 The quality of basic science education is uneven. In recent decades, Chile has taken many steps to improve the coverage and quality of basic education. The efforts include in-service teacher training, new curricula and the introduction of a full school day. Enrollment in pre-school educationhas also increased. Even so, the 2000 PISA test revealed that the quality of basic education remains uneven at best. In mathematics and science, Chile ranked 35 out of 41 participating countries, behind Mexico and Argentina (OECD 2003). Shortcomingsin basic science education will restrict the future supply of young people with the skills needed for a career in research and innovation. 3.83 Training policies could be strengthened. Despite a relatively high average educational attainment, the Chilean workforce does not possess the skills needed for a complex advanced economy. The Third Innovation Survey in Chile showed that close to 70 percent of companiesperceive a lack of qualifiedhuman resourcesto be an obstacleto innovation. Moreover, the 1998 International Adult Literacy Survey test showed that only 20 percent of the Chilean population has "reading comprehension sufficient to deal 81This restrictionhas recentlybeen somehowrelaxed throughthe policies for consortia formation. 84 with the demands of everyday life and work in a complex advanced society" (OECD 2000). Innovation-related training in Chile is far below the standards of more developed and innovative societies. A tax rebate system operated by SENCE under the Ministry of Labor has been effective in increasing training activities in recent years. However, the quality of training coudl be further improved and better aligned with the needs of the business community(Araneda and Marin, 2002). (ii) Cluster Policies 3.84 Cluster building vs. leveraging. Feser argues that the benefits from what he calls "cluster building" are closely tied to the notion of spatial externalities, knowledge spillovers and economies of scale enjoyed by clustered industries. Unfortunately, cluster building initiatives in developing countries have produced disappointing results because clusters have been poorly selected and they have suffered from a lack of S&T support. The successful targeting of clusters requires detailed and reliable data on innovative or entrepreneurial strengths which may not be available. Hence, Finland and Israel have focused on ensuringthat the NIS hnctions well in general terms so that industries with a potential for innovation may grow even though they may not be on the Government's radar screen. Clusterpolicies cannot be a substitutefor a well-functioningNIS, including policies for human capital formation over the long run. 3.85 Feser (2002) offers an alternative approach that seeks to "leverage" innovative synergies among existing or emergingbusinessesrather than seekingto establishclusters. Leveraging implies mobilizing the most important characteristics of the cluster - cooperative competition and collective efficiency - to strengthen innovative synergies between interdependent firms and institutions, regardless of whether spatial clusters emerge. Hence, the goal is to enhance the synergies that drive clustering rather than the physical cluster itself. 3.86 More particularly, he argues that, the "cluster" part of "cluster analysis" comes down to systematically analyzing strengths and weaknesses, and opportunities and threats, facing a set of interrelated existing or potential future industries. The focus should be on a broad policy framework with a minimum of dirigisme and little emphasis on physical location. It should aim to identify key linkages to non-market institutions such as universities and laboratories, as well as possible areas of joint research. The business community should be consulted on the need for policy intervention. Feser also identifies a set of tools to prioritize matching grants or subsidiesfor basic research, and to design supportingpolicy measures. Clustering in LAC 3.87 The initiative to establish clusters rarely comes from SME's as is common in Europe. In Spain and Italy, it was SMEsthat largely contributed to a growth in TPF at an early stage of the development process when there was little R&D. Identifying and eliminating the barriers to SME growth and development offers more "bang for the buck" than conventional extension programs through PRIs. Thus, a major focus of cluster strategycould be to address the weaknesses of SMEs. 3.88 In LAC there is a decided lack of cooperation among clustered firms. This suggests that policies to promote networking through PRIs and matching grants would be more effective than tax credits. Several OECD countries have kick-started clusters by supporting cooperativeresearch among universities, Government laboratories and private companies. 3.89 Inspired by collaborative research in Australia, CONICYT launched in 2004 the Consortia Program that encourages public and private entities to undertake joint research projects. Through a partnership between CONICYT, CORFO and FIA - supported by the World Bank - grants are awarded on a competitive basis both to large groups of researchers (Cooperative Research Consortia), and to research groups of more modest size (CooperativeResearch Teams). These groups bring together researchers and research users from ~miversities,public research institutes and industry to work together on issues of common interest. To obtain funding, the private and public sector participants in the consortia must themselves make substantial financial commitments in support of the cooperativeeffort. Consortia are structured as formaljoint venture partnerships with clear agreementson the handling of intellectualproperty rights. 3.90 Although the program is still in the launch phase, the response of industry has exceeded expectations. 61 pre-proposals were received of which 19 were selected as full proposals for international peer review. The private sector has pledged to finance more than one-third of consortia expenditure, and the establishment of 9 research consortia is expected to increase total private R&D spending by as much as 5 percent. C~~rrently contracts are being negotiated for consortia in the areas of dairy products, fishing, forestry, fresh fruit, viticulture and health care. 3.91 The consortia initiative differs from a cluster policy in its bottom-up approach. Partners need to come together and formulate a proposal in order to obtain funding. Collaboration is based on a clearly formulated project whose success or failure will depend on the ability of the partners to conduct research together. Hence, strong incentives exist to make cross-sector communication and collaboration work. It is premature to assess the Chilean experience with consortia. However, evaluations from Australia - where support for collaborative research was initiated in 1990 - find that consortia have encouraged team work and helped develop new linkages nationally and internationally(Myers 1995,Mercer 1998and Howard partners 2003).~~ Quoted in Holm-Nielsenand Thorn (2004). Box 3.6: Box: Clusteringin Finland Faced with limited domestic resources for R&D, Finland saw it as imperative that public hnds be used to support programs that have widely-applicable results. This, in turn, depended on a close relationshipbetween innovators and the firms that would use their product. It also depended on the ability of firms to absorb new ideas. By focusing on industries and companies that belong to major clusters - where input and output linkages are strong - it was possible to concentrate public finance for R&D in sectors where the results may be widely adopted. Furthermore, by explicitly including small and medium-size f m s in publicly funded R&D projects, which are often dominated by multinationals such as Nokia, it was also possible to strengthen the absorptive capacity of these firms and to increase the pace at which new technologies are adopted. The basic idea was to support the development of the cluster as a whole, rather than supporting individual companies. Another goal was "improving integration, collaboration, and division of labor within the whole research system". Ths latter goal is particularly important because the Finnish Government sought to improve the hnctioning of the entire NIS. It was important to ensure that measures to eliminate bottlenecks caused by a shortageof skilled workers and venture capital were not sidelinedby the cluster approach. In this sense, Finnish policy with respect to clusters, like Israeli policy, aimed at "informed neutrality". The long-nu1 goal is to ensure a well-functioning NIS that will allow the emergence of new sectors while at the same time boosting those sectors in which the country has a clear comparative advantage. Source: Blomstrom, Kokko and Sjoholm (2002). (iii) Policies for Small and Medium Enterprises 3.92 Today in Chile, both in academia and in the public sector, there is a feeling that SMEs can be left to the forces of creative destruction so that the competitiveness of the economy may be strengthened.However, if SMEs are considered as a key component of the NIS and if policies are geared towards upgrading their capabilities and addressing market failures, productivity can be significantly increased without keeping alive unproductive firms. SMEs have been responsible for the bulk of TFP growth in Italy and Spain. Hence, a more engaged approach may be worth consideringin Chile. 3.93 Comprehensive approaches appear to be successful. Tan (2005) shows that the productivityof small firms varies greatly across countries, both in absolute temis and also relative to the relative to the best performers of each country. The differences can be reduced by appropriate policies. He argues that comprehensive and coherent policies need to be adopted in the areas of training, technological progress and the management of supplies. 3.94 Training, in particular, yields high returns. As discussed earlier, LAC firms train relatively little. In Chile, it appears that the SENCE tax incentive was not well-designed and it has not generated the expected results. A more hdarnental question is why firms do not invest more in training. Firms cite the use of mature technologies and the availability of informal training as the principal reasons. But other reasons may include high labor turnover, limited resources, a poor understanding of training methods and skepticism about the benefits. Tan concludes that there is a role for the public sector in terms of disseminating information, introducing labor market reforms to reduce high turnover and in developing capitalmarkets to increase the availabilityof finance. 3.95 An integrated approach. Both Malaysia and Mexico have integrated policies of upgrading skills, technology, production processes, the adoption of ICT, and quality certification. Both also had programs to prepare SMEs as suppliers to larger firms by through increasing their capacity, raising quality standards and disseminating information. Unfortunately, only 10 percent of SMEs in Mexico were aware of the program. Thus, improving disseminationis clearlyhigh on the list of priorities. Spainhas also had great success with its Technical Institutes (Box 3.7) that have served as an intermediary between PRIs and the business community. A development plan for the period 2004-2010 envisages that the share of resources dedicated to technical advisory services will decrease while the share dedicated to joint basic research with universities and PRIs will rise from 0 to 10 percent. This presumes that the clientele will become progressively more sophisticated so that there will be less need for simple dissemination as opposed to the introduction of new ideas. It also presumes that the universities will shake off their previous insularity and will be more open to partnership arrangements with the private sector. 3.96 Chile today spends hundreds of millions of dollars to help SMEs address market failures in many sectors, through several institutions and programs. Goals and funding overlap significantly. There is rent seeking by firms who know the system, and a lack of information for the majority of others. Hence, the system does not work well. The Government is in the process of simplifying the policy framework. That task could be undertaken in a comprehensivemanner to increasethe effectivenessof funding and, even more importantly, to ensure that policies towards SMEs form part of an overall strategy to increase competitiveness. Figure 3.15: Financial Structureof Tech Centers in Europe Fedit Sintef VTT Fraunhofer rPQblicosNo Competitivos rnPiiblicosCompetitivos Privados Source: Fedit (2005). Box 3.7: MobilizingSMEs: Technology Centers in Spain Spain's technical centers provide a link between public and non-market institutionsin their efforts to foster innovation at the level of the firm. Their underlying rationale is that, given that Spain's industrial structure consists largely of SMEs, technology transfer from the scientific institutions to the productive economy poses difficulties.The technical centers facilitate the assimilation, adaptation and transfer of technology to enterprises. Their official mandate is to enable companies to obtain the best return possible from available scientific and technological resources and thereby improve their competitiveness. The technologies often originate from within the technology centers themselves which specialize in different sectors or in cross- cutting technologies dealing with information, communications, or the environment. Technological Centers are private entities with the majority of their boards drawn from the private sector. 61 of the TCs are voluntarily grouped under the Spanish Federation of Technological Innovation Entities (FEDIT), which is also a private, non-profit institution. Today, Tech Centers are guided by Spain's 2000- 2003 National Plan for Scientific Research, Development and Technological Innovation. Their total budget was around Euro 302 million in 2004 and they serve roughly 18,000SMEs. Financing is 59 percent private and 41 percent public, and the relative share of public vs. private research is similar (Figure 3.15). The public, non-competitive contribution is very small compared to Norway's SINTEF, Finland's VTT or Germany's Fraunhofer. The 5-year plan envisages shifting the activities of TCs away from technological services and more into basic R&b, in cooperationwith universities.The need for stable financing will thus emergeparamount. (iv) Support for Credit and Labor Markets 3.97 Labor markets can be barriers to innovation. A prominent body of academic literature supports the view that recalcitrant unions, or legislation which prohibits adjustments to the work force, can act as a barrier to the adoption of new technologies.It is useful to distinguish between two constraints: those that discourage the reallocation of workers across industries, a view developed by Hopenhayn and Rogerson (1993); and those that block the adoption of new technologies, a view formalized by Parente and Prescott (1994). The "empirical" evidence is mostly derived from simulations and it points to the potentiallyhigh cost of both constraintsin terns of productivityforegone. 3.98 In the case of Chile, recent studies have focused on the reallocation of workers between industries. Chapter 2 in particular cites evidence attributing more than 60 percent of growth in manufacturing productivity to factor reallocation which, conceptually, is behind the creative destruction process. It also shows how constraintsin access to credit, the high cost of closing firms and labor rigidities - chiefly labor firing costs-can have a negative impact upon the ability of small firms to adjust to shocks. 3.99 Caballero, Cowan, Engel and Micco (2004) argue that microeconomic flexibility is the core of creative destruction which enhances productivity and economic growth. One of the main costs is labor regulation. They find that job protection significantly lowers the speed of adjustment to shocks. Caballero Engel and Micco (2004) go even further, suggesting that the reduction in productivity growth after 1997 is due largely to an increase in microeconomic inflexibilityand, in particular, an increase in labor market rigidities. 3.100 Credit markets can be a constraint to innovation. This study has not focused much on the issue of credit for two reasons. First, much of the credit problem seems part and parcel of the general shallownessof financial markets in Latin America. Second, the financing of innovation cannot be delinked from the other issues such as business mentorship and capacity building, which merit a study in themselves. A good first attempt to review the issues at stake has been made by Mani and Bartzokas (2002). 3.101 Little is known of the overall impact of credit markets. Interviews with analysts at the IFC suggest that in the USA, most start-ups are financed with credit cards. It is not known if the paucity of venture capital (VC) financing in Latin America reflects a lack of supply or a lack of demand. In Chile, as in Italy and Spain, there is very little venture capital. However, in Chile, nunierous VC firms have been started and have subsequently failed. The cause of their failure may lie in barriers to innovation. Some VC managers argue that the lack of a "deal flow" reflects a shortage of upstream financing by "angel" or other start up investors. Other managers suggest that the cause of failure is that VC firms have neglected the management dimension of financing that plays a central role in the USA. In addition, earlier legislation permitted pension funds to invest in VC companies and thereby restricted their ability to take risk (Arrau 2002). A more careful examination of innovation financing could be undertaken within the broader context of business start-ups. 3.102 Venture Capital as a substitute for PlUs? Martin and Scott (2000) have argued that, if venture capital companies were to bid for matching grants, the need for publicly financed subsidies would be minimized through the bidding process. In addition, the development of new technologies and their successful implantation into a commercial environment could be overseen by those with expertise in the development and nurturing of innovation. 3.103 A lack of VC can be symptomatic of other constraints. There are concerns in Chile that the weak demand for VC might be explained in part by shortcomings in public sector support for the earlier stages of product development, such as pre-seed or seed stages. Recently, CORFO has attempted to fill this gap with new financial instruments. The different financial requirements for product development at successive development stages, and the policies and instruments to address them and fill the gaps, could be an important focus of the NIS policy framework. 4. EDUCATION 83 4.1 The education reforms begun in 1990 greatly increased investment in education and also led to a substantial increase in the quantity and quality of education inputs. At the same time, compensatory programs were introduced aimed at increasing school coverage while expanding school infrastructure and facilities, especially for the most vulnerable students. 4.2 The rise in education investment has led to significant gains in terms of coverage. However, there are still problems of access at some levels, and there is a challenge to improve the quality and equity of the education system. The recent expansion of pre- school and higher education has yet to reach the population in the lowest income quintiles. The quality of Chilean education is an issue of concern because national standardized achievement tests shown little improvement and the performance of Chilean studentsin internationaltests is unsatisfactory. Investment in Education 4.3 Total expenditure on education almost doubled between 1990 and 2002, increasing from 4 to 7.6 percent of GDP (Table 4.1). Table 4.1: Total Expenditureon Education relative to GDP. 1990-2002 (percent) 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Public expenditure in educatiodGDP 2.4 2.5 2.6 2.7 2.8 2.7 3.0 3.2 3.5 3.8 4.0 4.1 4.3 Private expenditure in educatiodGDP 1.6 1.7 1.8 1.9 2.1 2.1 2.4 2.3 2.7 3.0 3.1 3.2 3.3 Total educational ex~enditures1GDP 4.0 4.1 4.4 4.7 4.9 4.9 5.4 5.5 6.2 6.9 7.1 7.3 7.6 Source: Arellano (2004)based on MINEDUC data.Private spending estimated by the Department of Studies, MINEDUC, based on information from the Central Bank of Chile. 4.4 The near doubling of education expenditures reflected large increases in both private and public expenditure on education.84Private expenditures increased primarily as a result of an increase in higher education enro~lrnents.~~At the same time, public investment in education infrastructure increased from 14 billion pesos in 1990 (in 2004 83 Many individuals generously participated in discussions that greatly benefited this review. We are especially grateful to JosC Pablo Arellano, Beatrice Avalos, Cristian Bellei, Sergio Bittar, JosC Joaquin Brunner, Jaime Crispi, Pablo Gonzilez, Juan Eduardo Garcia Huidobro, Maria Ariadna Hornkohl, Mario Marcel, Pedro Montt, Dagmar Raczynski, Jorge Rodriguez C., Julio Valladares, and participants of the seminar held in Santiago in December 2004 on social sector development in Chile. 84 Between 1990 and 2002, public expenditures as a share of GDP increased 1.8-fold, whle private expenditures as a share of GDP rose 2.1-fold. 85Unlike most Latin American countries, all universities in Chile charge fees. pesos) to 166 billion pesos in 2 0 0 4 . ~Indeed, as a result of Chile's efforts to increase ~ education investment, the level of education expenditure as a percentage of GDP is similar to that of developed countries and greater than the average of OECD countries. Private expenditure on education accounts for a greater proportion of the total than in most other countries (Figure 4.1). Figure 4.1: Total EducationExpenditures relative to GDP, 2000 (percent) Ireland Mexico OECD Argentina Korea USA Chile (average) Note: OECD data relates to the academic year from Septemberto May. MINEDUC data relates to the calendar year. Source: OECD (2004b). 4.5 Table 4.2 presents the average annual expenditure per student between 1987 and 2002 for primary and secondary education (grades K-12) and for the undergraduate programs of the "traditional" universities (Couiicil of Rectors). From 1990 to 2002, averageper student expenditurein primary and secondaryeducationincreasedby a factor of 2.6 in real terms. That largely reflects an increase of 156 percent in the average monthly subsidy. In the traditional universities, enrollment almost doubled between 1990 and 2002'~but average expenditureper student increased by a factor of only 1.2. 86Includes capital contributions for full school day (JEC), education expenditures from the National RegionalDevelopmentFund, (FNDR), and the maintenance subsidy, see Marcel (2004). Indeed,public and private universityenrollment increasedby a multiple of 2.7 between 1990and 2002. Table 4.2: Central GovernmentExpenditureper Student, K-12 and Higher Education, 1987-2002 Expenditures and K-12 enrollment Expenditures and higher education enrollment Annual Annual expenditure expenditure Enrollment per average higher per average average in Total university Enrollment subsidy education student in K-12 student (thousand)b traditional enrolment traditional (thousand (2002 universities universities (thousand) 2002 (thousandld Ch pesos) (2002 ~h pesos)c Ch pesos)" Year (1) (2) (3) (4) (5) (6) 2002 430, 3,295 29,911 1,174 226 349 a/ Total expenditurefor K-12 education per enrolled student. bl Includes pre school, elementary and secondary education in municipal schools,private subsidizedschools and schoolsrun by and private corporations. C/(3)= Total MINEDUChigher education expenditure1(4) dlOnly undergraduate enrollment from universitiesbelonging to Rectors' Council (calledtraditionaluniversities) Source: Arellano (2004)based on Educational Statistics,2002. MINEDUC. 4.6 Much of the increase in the student-based subsidy went to finance an increase in teacher salaries. Between 1991 and 2002, average full-time teacher salaries increased by 156 percent. The starting salary in municipal and private subsidized schools grew by 174 and 433 percent, respectively (Figure 4.2).88At present, teachers' salaries have recovered the ground they lost during the 1980s, and they are now well aligned with the market and with international comparators.The salary of a primary school teacher with 15years experience is equivalent to 133percent of per capita GDP in Chile. That compareswith 127percent of per capita GDP in advanced developed countries and 164 percent in countries such as Finland, South Korea and Sweden. The corresponding figures for a secondary school teacher is 133 percent in Chile, 132percent in advanced countries and 170percent in Finland, South Korea and Sweden (OECD, 2004b). 88This large increase is explained by the low level in 1990. 93 Figure 4.2: Monthly Salaries of Full-Time Teachers (44 hourlweek) ( ~ e s oof 2002) s P 1T -~5 rn 4- = - '0 P P wn - 7E a:: Ic 0 00 '0 N m m- P - % 5 2 1980 1881 1882 1883 1841 1885 1896 1997 1998 1998 2000 2001 2002 I+-Municipalaverage salary +Municipal startingsalary - Private-subsidizedstartingsalaryI a) Average salary includes:RBMN (nationalminimum wage) plus benefits (10 two- year periods, responsibility,upgrading, performance,and difficult conditions),UMP (professionalimprovement unit), proportionalbonuses, total handicap, excellence bonuses and additionalsalaries. b) Starting salary includesRBMN, UMP base; proportionalbonus and complementaryallowances. Source: MWEDUC Years of Schooling and Coverageof Education 4.7 Average years of schooling of the population have increased since the 1980s. It is likely that the increase in education investment that began in 1990 led to an acceleration of educational attainment of the population at all income levels. However, the greatest improvement was among the highest income deciles. In spite of this progress, the majority (53 per~ent)'~of today's adult population in Chile has not completed secondary education, and low schooling levels are particularly prevalent anlong the poor. Census data show that while adults in the highest income decile increasedtheir years of schooling from 12.9in 1990 to 14 years in 2002, adults in the lowest income decile increased their years of schooling from 4.1 to only 4.9 years over the same period (Figure 4.3). Indeed, a remaining challenge in Chile is to educateunder-schooled adults, especiallythose from low incomebackgrounds. 4.8 In order for Chile to enhance its competitiveness and reduce income inequality it faces the challenge to close the gap in years of schooling between high- and low-income adults. In 2002, the Government launched ChilecaliJica, a flexible program designed to provide opportunities for adults to complete basic education and attain technical skills for the labor market.90Preliminary evaluationsindicate that low income adults are taking advantage of the new opportunitiesto upgrade their skills. 89OECD,Education at a Glance 2004. 90Thls program is partially financed by The World Bank. Figure 4.3: Years of Schoolingby Income Decile, Adults Aged 25 years or More, 1992-2002 1 2 3 4 5 6 7 8 9 10 (poorest) (richest) SOCIOECONOMIC DECILE I Sampleapercentwi* lowerincomeand sample5 percent total ciuntry Base individual aged25 years or more I ' I I Source: 1992and 2002 Census. 4.9 Access to secondary schooling improved substantially in the 1990s, with secondary school enrollment rates reaching 87 percent in 2002 (Figure 4.4).91At the same time, there was a significant decline in secondary school dropout rates, from 10.3 percent in 1991 to 8.5 percent in 2002 (Figure4.5). Figure4.4: Enrollmentin Primary and SecondaryEducation, 1982-2003 (percent) +elementary --r---secondary Source: MMEDUC. Primary school educationwas expandedto all childrenin the 1980s. 95 Figure 4.5: Dropout Rates in Primary and SecondaryEducation, 1991-2002 (percent) Desertion rate in elementaryand secondary education &secondary A elementary Source: MMEDUC. 4.10 The increase in coverage was due to a notable expansion of private subsidized schools, especially secondary schools, as a result of reforms in the early 1980s.Indeed, since 1981, private subsidized schools have consistently increased their share school enrollment, while the share of municipal schoolshas steadily declined (Figure 4.6). Figure 4.6: Enrollment by Type of School, 1981-2002 (percent) I 1 HMunicipal Private Subsidized Private Fee-paying I Source: MINEDUC. 4.11 The coverage of pre-school and higher education has also increased greatly since 1990. Enrollment in pre-school education increased from 20.9 percent in 1990 to 35.1 percent in 2003, while higher education enrollment increased from 16 to 37.5 percent during the same period. However, pre-school enrollment rates are low by international standards (Figure 4.7). That is troubling given the effects of pre-school attendance on school and labor market performance later in life. Figure 4.7: Enrollmentby Age, Chile and SelectedOther Countries (percent) I ~ a l a y s kB d I 'Canada Chile Mexico Argentina Peru USA OECD Spain Italy 4 year-olds or less H 5-14 yearalds H15-19 yearalds 1 Source: Tokrnan (2005). Education Inequality 4.12 Not only are school enrollment rates (other than basic education) lagging with respect to comparator countries, there are also important inequities in schooling with respect to income and ethnic background. While the gap in secondary school enrollment between high- and low-income students has narrowed between 1990 and 2003, the corresponding gap in pre-school and higher education enrollment remains substantial despite a significant increase in enrollment rates. (Figure 4.8 - Figure 4.10). Understanding the reasons for persistent inequalities in access to pre-school education is necessary in order to develop more effective policies for incorporating children into the educationprocess. Figure 4.8: Pre-school Enrollmentby Income Quintile (percent) Coverage in preschool education by income quintile (%) I I Source: MIDEPLAN, CASEN Surveys 1990 and 2003. Figure 4.9: SecondaryEducation Enrollment by Income Quintile (percent) Coverage in secondary education by income quintile (%) I I Source: MIDEPLAN, CASEN Surveys 1990 and 2003. Figure 4.10: Higher EducationEnrollment by Income Quintile (percent) Coverage in higher education by income quintile (%) 73.7 80 70 60 n 50 1990 40 2003 30 20 10 0 Source: MIDEPLAN, CASEN Surveys 1990 and 2003. 4.13 In all regions of the country, there is a gap in the enrollment rate of different income quintiles. (Table 4.3). In addition, CASEN data show that only 26.2 percent of -theindigenous population enroll their children in pre-school education - a substantially lower proportion than average. In contrast, mean enrollment rates in primary education do not differ substantiallyacross income quintiles or between regions. 4.14 Enrollment in pre-school has a positive impact on subsequent schoolperformance and fut~~resuccess in the labor market, particularly for children of poor families. Pre-school education increases the likelihood of finding lucrative employment and higher earnings. Early-life experiences of socialization and education also enhance a child's future development (Heckrnan 1999).It is costly and difficult to make-up later for failure to attend pre-school. Thus, increasing pre-school enrollment rates in Chile, especially for the poor, would help improve education and labor market outcomes. It would also likely help reduce income inequalityin the medium term. 4.15 Income and ethnic background also determine the kind of higher education received. While most students attending universities (private and traditional) come from the highest income quintile, youth from the lowest quintile are more likely to enroll in Technical Training Centers (Table 4.4).92 Only 23.1 percent of youth from indigenous backgrounds enroll in tertiary education (CASEN). 92Technical careers are often chosen by students from lower socioeconomicbackgrounds because they require a shorter training period which makes them more attractive for those wishing to enter the labor force quickly; they are less costly relative to university careers; and it is easier to be accepted by technical schools. Table 4.3: Enrollment by Level of Education, Region, and Income Quintile (percent) stimated Table 4.4: Enrollment in Higher Educationby Type of Institution and per capitaIncome, 2003 Income Quintile I I1 I11 IV V Total Universities 6,o 10,2 16,l 25,5 42,2 100 - Traditional universities 6 8 13,O 18,2 26,O 36,l 100 - Private universities 4,7 5,6 12,7 24,6 52,4 100 Professional Institutes 6 3 12,s 22,4 30,4 28,2 100 Technical Training Centers 16,O 22,5 26,9 17,l 17,5 100 Notes: Traditional universities are those universities belonging to the Rectors' Council. Domestic live-in service is excluded. Source: MIDEPLAN, CASEN Survey 2003. Higher Education: Expansion, Quality and Equity 4.16 Since 1990, traditional universities have doubled their enrollnlent, while at the same time there has been a substantial expansion of enrollment in private higher education institutions. As a result, between 1990 and 2003, the gross enrollment rate in higher education rose from 16 percent to 37.5 percent. This sharp increase in enrollment has put pressure on student financing mechanisms. That, in t ~ nhas important budget implications , for public subsidies because university student loans have low recovery rates. In the coming years, it is anticipated that the demand for higher education will increase as a result of demographic pressures and because more students are completing secondary school. (Eyzaguirre at al., 2005).'~ 4.17 Efficiency, particularly in universities, has not improved commensurately with the recent expansion of the sector. While the graduation rate in non-university tertiary institutionsis comparable to that of OECD countries, it is substantially lower in universities. Only 29 percent of university entrants graduate in the stipulated time, compared with the 76 percent OECD average (Brunner et al., 2005). The low graduationrate implies that resources are not being used efficiently,and that Chile is developingadvancedhuman capital resources at a slower pace than its potential. 4.18 Equal access to higher education remains a challenge, but President Lagos' administration has taken important steps to expand options for student financing. Increased financial support to students in traditional universities (crkdito solidario, an income- contingent loan) during the past decade has permitted an increase in student enrollment, particularly among lower income students (Table 4.4). However, the crkdito solidario is confined to students attending traditional universities, which constitute only half of all students in higher education. In 2005, Parliament approved a parallel student financing system for students attending private and non-traditionaluniversities, as well as technical and 93 The expansion of university education has three facets: (1) an increase in the total number of institutions, with corresponding challenges related to comparabilitybetween degrees and programs; (2) an increased participation of the private sector in tertiary education, with corresponding challenges related to quality, equity and regulation; and (3) a separation within tertiary education, specifically between university institutions and non- university tertiary education institutes, which also raises issues of quality and equity (Brunner and others, 2005). professional institutes. For the future, it would be important to integrate those initiatives into a more comprehensivesystem of student finance for tertiary education. 4.19 The recent expansion in higher education has also created challenges for prospective students and employers due to insufficient reliable and uniform inforn~ationon the quality of new programs and new institutions. In an effort to reduce the information gap, a national accreditation system for higher education is currently being developed (Brunner 2004). In 1999, two commissions were established to develop quality assurance and accreditation at the undergraduate and graduate levels. Currently, about one-third of undergraduates are enrolled in an accredited program. Progress has also been made with respect to voluntary institutional accreditation, and a first roster of public and private institutions has been accredited. It would be important to sustain these efforts by developing an appropriate legal framework for the accreditation of tertiary education institutions and by ensuring adequate funding for accreditation. 4.20 Finally, since 1998, the Government has sought to improve the quality and relevance of tertiary education by introducing competitive funding. Resources from this source have been furnished to more than 300 university-based projects to improve the learning infrastructureand management of universities. Recent competitiverotinds have been used to redirect resources to the redesign of degree structures and curricula reform inspired by the European Bolognaprocess. The Quality of Primary and Secondary Education 4.21 Educational quality is as important as its coverage in terms of helping to raise both personal incomes and economic growth. Given the scarcity of information about the quality of higher education,this section focuseson the quality of primary and secondaryeducation. 4.22 National and international testing suggests that the quality of education in Chile is low and slow to improve. Average scores in the national SIMCE assessmentsfor the last few years-and in particular since 1998when the tests began to be comparable over time-show little or no improvement. Table 4.5 shows results from of the tests between 1994 and 2004. For the period 1994-1997, SIMCEhas adjusted the scoresto make then1comparablewith the current scale. The test scores tend to fluctuate around 250 points (the base score adopted for each test).94 - - 94Comparisons over time are difficult because SIMCE has changed its methodology and the timetable of tests. SIMCE modified the measurement in 1998,based on Item Response Theory, IRT. The SIMCE commission has since proposed a reorganizationof the timetable to allow the periodic evaluation of 4th grade primary students and that will pennit the measurement of school's value added in the future. Table 4.5: Average SIMCE Test Scores. 1992-2004 Testslyears 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 4&Mathematics 248 250 247 4' Language 251 250 251 8" Mathematics 250 250 253 8" Language 253 250 251 10&Mathematics 242 250 248 246 1O& Language 249 250 252 253 Source: MINEDUC. SIMCE. 4.23 In addition to a lack of progress in student performance, there is a substantial gap between the average test scores of Chilean students and those of students from other countries according to international assessments of student learning such as the Third International Mathematics and Science Study (TIMSS) and the Program for International Student Assessn~ent(PISA) (Figure 4.11 and Figure 4.12). Figure 4.11: PISA 2000 AverageMathematics Scores by Country Source: OECD (2h4b). PERU ALBANIA INDONESIA MACEDONIA BRAZIL CHILE ARGENTINA MEXICO ROMANIA BULGARIA THAILAND LUXEMBOURG ISRAEL LATVIA RUSSIAN FEDERAl'ION PORTUGAL GREECE POLAND HUNGARY LIECHTENSTEIN GERMANY ITALY CZECH REPUBLIC SPAIN SWITZERLAND DENMARK UNITED STATES FRANCE NORWAY ICELAND BELGIUM AUSTRIA SWEDEN JAPAN UNITED KINGDOM KOREA. REPUBLIC OF HONG KONG IRELAND AUSTRALIA NEW ZEALAND NETHERLANDS CANADA FINLAND Figure 4.14: PISA 2000 Math,ScoresandExpenditures per Pupil, by Country 3h 'BRA 300 - 0 PERU *P 200, 50 2050 4050 6050 8050 10050 12050 Annual expenditureon secondary education, per full-time-equivalentstudent, in equivalent US$ converted using PPPs Source: OECD (2004b). 4.24 When PISA scores are related to GDP per capita and to expenditureper pupil, Chile's results are well below expectations (Figure 4.13 and Figure 4.14).~~ 4.25 Furthermore, while 32 percent of 15-year old students in OECD countries have reading skills that are in the top two levels assessed by PISA, only 6 percent of students in Chile do so. Perhaps even more troubling is that 20 percent of Chilean 15-year olds perform below the first level (indicating that they cannot master basic reading skills) while in OECD countriesthe figure is only 6 percent (Figure 4.15). 95Due to space limitations, we only present the figures for mathematics. The language figures are very similar. 105 Figure 4.15: Distributionof Studentsby Level of Performance accordingto PISA 2000 reading scores,by Country (percent) I .Level 4 Level 3 Level 2 .Level 1 BlBelow I Source: OECD (2004b). 4.26 Chilean students who score in the highest percentile (and who, in most cases, attend private fee-paying schools) do not perform well when compared with similarly high-scoring students from other countries, or even with the average student from other countries. Thebest Chilean students achieve results that are little better than the OECD average and well below the average score achievedby SouthKorean and Japanese students (Table 4.6). Table 4.6: Distributionof PISA 2000 MathematicsScores,by Country Country Average Percentiles 5 25 75 95 Argentina 388 180 307 474 574 Chile 384 222 321 449 532 Mexico 387 254 329 445 527 Japan 557 402 504 617 688 Korea 547 400 493 606 676 USA 493 327 427 562 652 Latin America 357 177 286 432 530 OECD (average) 500 326 435 571 655 Source: MINEDUC. 4.27 Although mean test scores are low, there is great variation in student achievement and in school quality. Indeed, schools with stagnant scores co-exist with schools that show continuous progress.96 These differences are not necessarily associated with the socio- economic background of the students. Figure 4.16 shows the relationship between SIMCE 96Mizala and Romaguera (2005b) show that if the 10percent lowest performing schools in 2001 and 2003 are excluded from the computationof SIMCE average scores, then mean scores increase from 2001 to 2003 in both language and mathematics. scores and the schools' vulnerability index for the eighth grade tests taken in 2000.97The great diversity of school performance in Chile has important research and policy implications.If it were that all the schools suffer from poor performance, one could conclude that responsibility does not so much lie with scliools but, instead, on the Ministry, the parents, or even the children.In contrast,if some schools are progressing while others are left behind, different questions arise, such as: (i) what characteristicsdistinguishhigh performing schools from those with low levels of student achievement? (ii) what lessons can be learned fi-om schools that show significantly better scores and that serve populations with similar characteristics?(iii) how can these lessons be applied to improve the performance of schools with poor achievement? Figure 4.16: Eighth-Grade SIMCE Scoresand Vulnerability Index, 2000 I I I 4 I 20 40 60 80 100 Vulnerability Index I5O0 I Source: MINEDUC. SIMCE 4.28 Bellei et al. (2003) analyze the factors that characterize high performing schools in poor areas and found that they are not easy-to-measure characteristics, such as teacher qualifications and experience, class size, etc. Instead, they found that high performing schools have, among other things, clear goals and objectives, committed and highly professional teachers, an instructional leader at the school who evaluates and supports teachers in classroom practice, well-respected principals, clear rules, and multiple approaches to reach a diverse student body.98These factors are difficult to replicate without strong technical capacity in schools as well as in municipalities and private providers (sostenedoves). 97Ths phenomenon is found in all SIMCEtests independent of the grade or subject evaluated. 98Another issue that has not been sufficiently analyzed is the extent to which there is variation in quality between municipalities in Chile. It is possible that some municipalities have, over time, improved student learning outcomes while, in other niunicipalities, average performance has remained stagnant or even declined. Understanding the factors that may explain variations in student performance between municipalities is important for educationpolicy. 4.29 The large extension of education coverage throughout Chile in recent years undoubtedly had an impact on average test scores, especially in secondary school. At present, children who otherwise would have been outside the school system and who tend to have poor educational scores are enrolled in school and taking these assessments. In 2003, around 50,000 more students took the SIMCE test for 10th grade, compared to the previous examination-an increase of approximately 20 percent. The Ministry of Education calculates that half this increase is due to demographic growth and the other half due to increases in coverage and retention rates." 4.30 Comparisons of gross SIMCE results suggest that private fee-paying schools achieve better scores than private subsided schools and they, in turn, achieve better results than municipal schools (Table 4.7). However, these comparisons do not adequately measure quality differences, as differences in the availability of resources between schools are not taken into account. Household characteristics are different between different types of schools, in particular between private fee-paying schools and the rest. According to a survey of parents undertaken together with the SIMCE test in 2003, the average child in a private fee-paying school has the following characteristics: (i) parents with post-secondary or professional education; (ii) annual household income above a million pesos; and (iii) 87 percent have a computer and 73 percent access to the internet. The average pupil attending public (municipal) schools has (i) parents who have only 9 years of schooling; (ii) an average household income of 180 thousand pesos; and (iii) 20 percent have access to a computer and 7 percent to the internet. Further, it is estimated that, on average, private fee-paying schools have three times more resources than schools that only collect the student subsidy paid by the State. The average subsidy at the end of 2004 was around CLP$30,000 for primary and CLP$35,000 for secondary students, while the average monthly fee per student in a private school was around CLP$100,000 (Mizala and Romaguera, 2005b). 99Mizala & Rornaguera (2005a) analyzed the impact of increased coverage on SIMCE test results for 10' grade (secondary) in 2003. By excluding the poorest decile (40 percent of the 50 thousand new students), they found that the remaining scores increased relative to the previous SIMCE test of the 10" grade in 2001. , Table 4.7: Average SIMCE Test Scores by School Type 4thgrade 2002 tIthgrade 2004 lothgrade 2003 Language Mathematics Language Mathematics Language Mathematics Municipal 239 235 240 241 241 230 Private subsidized Private fee- paying National average 251 Source: MINEDUC, SIMCE. 4.31 School performance cannot be assessed on the basis of raw SIMCE scores. An accurate comparison must take account of differences in the financial resources available to schools. Studies show that the gross score differencesare substantiallyreduced when account is taken of the resources available to schools and families, even though private fee paying schools continue to have better results than private subsidized schools, which, in turn, have better results than municipal schools (Mizala and Romaguera, 2001; Sapelli and Vial, 2002; Gallego, 2002). 4.32 Although differences in average SIMCE scores between private and municipal schools are significantly reduced when they are adjusted for student socioeconomic background, the gap is still important for public policy. To guarantee equal educational opportunity, policy makers face the challenge to ensure that all studentshave access to a high quality education independent of their circumstances. Furthermore, if students from low income households are to be given greater access to higher education, that will first require raising the performance of secondary schools serving low income students. It is clear that additional resourceswould need to be fiunished to the poorest students. The present standard per student subsidy discriminates against students from poor socioeconomic backgrounds that require additional attention and educational resources. 4.33 However, increased resources alone will not be sufficient to meet the challenges faced by the Chilean education system. It is important that education reforms in Chile take account of existing market failures and foster competition between schools that may lead to improvements in the quality of education. In that context, it is important to redefine the roles and responsibilities of the players operating within this quasi-market in education and to align incentives to raise quality. A key step in this direction will be to better inform the public about school quality. Finally, to improve cost effectivenessof education reforms and policies, greater efforts could be made to evaluate them on a systematic basis. These issues are examined in the following sections. 4.34 Decentralizationin the early 1980stransferred the administrationof public schools to municipal governments. The reform also opened the way for the private sector to provide publicly financed educationby establishinga voucher-typestudent-basedsubsidy. 4.35 Three types of schools were established: municipal schools, financed by a student- based subsidy granted from the State and run by municipalities; private subsidized schools, financed by the State student-based subsidy and run by the private sector; and private schools, financedby fees and runby the private sector.loOThe amount of the subsidypaid per student is the same for both municipal and subsidizedprivate schools. 4.36 Figure 4.5 shows enrollment in the three types of schools for the period 1981-2002. While enrollment in private subsidized schools was 15percent of total enrollment in 1981, it increased to 38 percent in 2002. Meanwhile, the share of enrollment in municipal schools decreased from 78 percent in 1981to 53 percent in 2002.lo' 4.37 Private schools that charge fees are generally for-profit, whereas subsidized private schools can be either non-profit or for-profit. Non-profit private schools include church schools and those dependent on foundations or private corporations, some of which are linked to industry.For-profit schools operate as firms, generatingreturns for their owners. 4.38 The voucher system together with the private provision of education was expected to promote competition between schools that would need to attract and retain students. Thereby, an "education market" would be created which, through competition, would encourage efficiencyand educationalquality. 4.39 The extent to which school choice leads to better education quality is a hotly debated issue among education economists worldwide. Proponents of privatization favor allowiiig non-governmental groups to provide schooling with the funding partially or totally provided by Government. They argue that private schools are more successful in recruiting the best teachers and in developing their teaching skills. As a result, they tend to generate better student outcomes. Some of the reasons for ths success include greater supervision and mentoring of new teachers by private schools,their ability to recruit better qualified teachers, to attract teachers who exert more effort and independence, and, ultimately, their freedom to dismiss teachers for poor performance (Ballou & Podgursky, 1998; Hoxby, 2000(b)). Opponents to school choice maintain that, because private schools can select students while public schools cannot, privatization leads to increased segregation by raciallethnic, socio- economic, and cultural background, leaving the public sector with the difficult task of serving the most disadvantagedchildrenwhile having less resources (both in terms of finance and supplies) than in a fully public education system (see, for example, Fuller and Elmore, 1996). looA fourth type of school, a concession-type model whereby public schools are managed by private corporations was prohibited by the Contraloria General de la Repliblica. Only 53 municipalities have such schools (Gonzhlez 2003). lo'These types of schools exist in each of the three levels of the Chlean school system: preschool, elementary or primary school, and secondaryor high school. 4.40 Previous research on the relative effectivenessof private and public schools in Chile has led to differing conclusions. Rodriguez (1988), Aedo & LarraSiaga (1994), and Aedo (1997), find that private voucher schools achieve better student outcomes than do municipal schools. In contrast, Mizala and Romaguera (2001) and, in a series of papers, McEwan and Carnoy (1999 and 2000), have found that private voucher schools do not perform any differently from municipal schools. However, when non-religious and Catholic voucher schools are considered separately, McEwan and Carnoy (2000) find that Catholic voucher schools outperform municipal schools. The studies differ mainly in terms of the samples used, the variables included in the analyses, and in the classification of sectors. Given the complicating effects of student background, peer influences and other unobservable variables,it is methodologically challenging to identify empirically the impact of competition on student outcomes. Thus, the findings of research are inconsistentwith regard to the effect of competition between private and public schools and whether competitionimproves student outcomes or exacerbates inequality between different groups of students. 4.41 This review does not take sides in the debate over school choice. Instead, it acknowledges that Chile has made important reforms in providing school choice to families, and that these reforms are widely accepted by the majority of Chileans. A concrete achievement of the reform has been to increase coverage and retention, though its impact on improving school quality is questionable. It is unlikely that Chile will go back to a system where money does not follow the student and it probably does not make sense in terms of improving education quality and equity. Instead of proposing to change the student-based subsidy in Chile, this analysis focuses on policy options to improve the functioning of the school choice system. 4.42 Differences between subsidized private and municipal schools have impeded the creation of an "educational market" that fosters educational efficiency and quality. Public and private schools compete under different conditions, thus limiting the gains in efficiency and quality that may expected from the voucher-type student-based subsidy. As a result, the Chilean education "quasi-market" has been unable to raise average student achievement for the system as a whole, and it has also fallen short of ensuringa high quality education for the elites (Eyzaguirre at al., 2005). 4.43 The most important differences between subsidized private and municipal schools relate to: (i) the student admission process, in which subsidized private schools can select their students while municipal schools are required to admit all students interested in enrolling; (ii) teacher contracting, which enables subsidized private schools (but not municipal schools) to directly hire and fire teachers; and (iii) the ability to have alternative sourcesof financing. 4.44 Student admissions. Private schools (both subsidized and fee-based) have complete freedom to accept, reject, and dismiss students as well as to establish their own selection processes. In contrast, municipal schools are required to accept any student who wishes to enrollunless it can be demonstrated that there are no vacancies at the school. 4.45 Teacher selection and contracts. Teachers in municipal schools are governed by special legislation (the Teacher Statute), which calls for teacher remuneration to be determined by a centralized process of collective-bargaining. Although not stipulated in the Statute, the result has been that teacher wages in the public sector are based on uniform pay- scales with special bonuses for training, experience and working under difficult conditions, as well as restrictions for teacher dismissal. Private schools (both subsidized and fee-based) operate as firms, and their teachers come under the same Labor Code as other private-sector workers.lo2 In addition, private fee-paying and subsidized schools can select, hire, and dismiss their teachers, while teachers in municipal schools are hired and assigned centrally. Because these teachers are governed by the Teacher Statute, their dismissal is also much more difficult than that of teachers in private schools. 4.46 AlternativeJinancing sources. Since 1993, private subsidized schools and secondary municipal schools have been authorized to arrange for co-financing with funds contributed by parents. Elementarymunicipal schoolsmay not charge fees. 4.47 In addition to the differencesbetween public and private schools described above, the education system fails to make information available to the public on the quality of school education. At present, the information on quality that is disseminated to parents and the public in general consists only of the average score of each school in the SIMCE tests. These averages mask important differences-in student achievement as well as in the population of students served. For example, two schools with the same average SIMCE score can have very different levels of quality depending on the extent to which they have the authority to select (or cream-skim) students or the extent to which the range of their test scores differ. School quality is more than student performance in national assessments. Other information, such as the proportion of students who go on to pursue further studies, can help inform parents and improve the functioningof the educationalmarket in Chile. 4.48 Similarly, there is insufficient information available on the quality of higher education institutions in Chile, and, given the expansion of higher education progranis, this creates real constraints to informed decision-making by prospective students and their families (Brunner 2004). A promising initiative to increase and improve the information available on tertiary education programs consists of a follow-up survey of tertiary education graduates. Chile has set up a tracking system to follow the performance of tertiary graduates in the labor market in order to monitor the quality and relevance of the education they have received. The survey draws upon tax informationto comp~~teand update data on the earnings of tertiary graduates in all major career paths.lo3Reliable information on labor market outcomes has enhanced the transparency of Chilean tertiary education and provides potential students a basis for making informed decisions. lo2 There are certain nlinimal labor contract rules from the Teacher Statute that are applied to private sector teachers, such as startingwages, length of working day, holidays and severance pay. lo3The observatory's webpage - www.futurolaboral.cl - has been very successful with more than 4,000 daily visitors. THEROLEOFTHE STATEIN ENSURINGQUALITY AND EQUALITY OF EDUCATIONALOPPORTUNITY 4.49 The previous section emphasizes the need to improve the functioning of the Chilean education market. The State has an iniportant role to play to ease constraints in the education market and guarantee that all children have access to an education of adequate quality even, or perhaps especially, where there is private provision of education and market competition. First, the State has the authority to demand accountability for public resources allocated to schools. At present, schools receive the per student subsidy exclusively on the basis of student enrollment. Greater accountabilityimplies that the State has a role in developing and enforcing standards,norms and incentives for the education system within the framework of free school choice. Similarly, the State has a responsibilityin both supervising and assisting schools to achieve the expected results. Second, there is general agreement that the State has a responsibility to ensure equality of educational opportunity. Third, there is an important role for the State to improve the information available to parents and civil society on school quality. Finally, the State has an important role in ensuring that education policies and programs are subject to rigorous evaluation. The benefits of evaluations are subject to many externalitiesand, as a result, there would be underinvestment in the absenceof State support. Increasing Accountability for Public Resources 4.50 The presence of multiple players - Ministry, municipalities and private providers, head teachers, teachers, pupils and parents - often leads to confusion about who has the primary responsibility for the quality of education provided by a school, resulting in low accountabilityin the education system. 4.51 Publicly financed private and municipal schools in Chile have little accountabilityto Government and the public in general regarding how they use resources from the student- based subsidy. An important step towards increasing acco~mtabilityin the education system would be to raise the requirements for schools to meet in order to access public education subsidies. Indeed, as Eyzaguirre et al. (2005) argues, in an education system as decentralized as Chile's, the freedom of families to choose a school for their children depends on the availability of alternatives and sufficient financial resources. That, in turn, requires a strengtheningof standardsfor student learning - understood as both minimum learning levels as well as a guarantee that school children will at least have command of basic skills. In that regard, the current administrationis designing "maps" of student skills for each grade, which could be used to both guide schools and teachers in their work in classrooms as well as in evaluating school performance. 4.52 Accountability is not only the obligation of schools to parents and Government, however. The State also has an important role to play in supervising and assisting schools and teachers to achieve higher standards of teaching and learning. Those issues are addressedbelow. 4.53 Roles of the Central Governmentvis-A-vis Municipal Governments. Under the current structure of the education system, the Ministry is in charge of technical-pedagogic issues while the municipalities are responsible for administrative operations. In practice, the result of that arrangement is that municipalities assume little responsibility for student learning while the Ministry underestimates problems related to school management.lo4Furthermore, some municipalities lack the capability to manage because of technical weaknesses, an absence of clear agreementsbetween the municipality and the school, the small size of some municipalities,and political issues.lo5 4.54 Differential autonomy conditioned on results. Chile would benefit from central system with the capacity to design and evaluate policies, together with a decentralized apparatus with management capacity to effectively supervise schools. However, given the diversity of the education system, schools and municipalities would benefit from differentiatedlevels of autonomy and supervision.Greaterresponsibility and autonomy could be granted to high performing schools, including, for example, the possibility of making decisiolis regarding the use of financial and human resources. In that way the educational system's decentralization would be deepened and there would be a stronger link between quality and finance, which today is weakened by the intermediary role of the municipality. Greater decentralization and autonomy could depend on the definition and measurement of explicitperformance standards for schools and their teachers.lob 4.55 In contrast, under-performing schools would receive increased pedagogic and administrative assistance and increased supervision from the Ministry and from municipal governments. Intervention in schools whose performance is systematically poor could be considered, including the transfer of students to other schools. Ultimately, private subsidized schoolsthat systematicallyshow poor results could cease to receive State funding. 4.56 There are several options for interveningin the case of ~u~der-performingschools, and they have different implications for the education market in Chile. One option is for municipal governments to close down under-performing schools and relocate students to neighboring municipal schools. Of course, this is not a feasible option in some rural and remote areas. While the proponents of school choice argue that competition would result in the closing of schools that are poorly attended, this hypothesis has not been substantiated by the Chilean experience.In spite of a substantial fall of enrollment in municipal schools (from about 75 percent of total students to less than 60 percent), the total number of municipal schools declined only slightly from 6,370 in 1980 to 6,288 in 1990, a decrease of only 1.3 percent (Beyer at al., 2000). 4.57 A second option is for the central Government to take over the administrationof low- performing schools, at least for a period of time, in order to re-organize the school and improve its performance. This option would go against the decentralization efforts undertaken in the past decades, and it is unlikely that the central Government would have the capacityto successfullyintervenelow-performing schools. lo4 See Traverso (2004). 105The experience of Holland, which has had since 1917 a school choice system similar to Chile, is relevant. The system is very decentralized. Each school has a school board which must comply with highly demanding State regulations.Although subsidized schools can charge fees over and above the subsidy,no private or public school can exclude pupils unable to pay, Karsten (1999), Patrinos (2003). 106An evaluation should be made of the decentralizedadministrationof resources (administracidn delegada) to see how it has worked and to examine the possibility of expanding it to schoolswith good performance. 4.58 A third option, which has proven effective in several countries, is to devolve authority for school management, including the use of resources and personnel, to parents and communities. In Chile, in spite of the decentralization and voucher reforms, parental participation in school-level decisions is relatively low. In contrast, several European countries, including Finland, have school governingboards comprised of a principal, parents and other representatives of the community to manage p~~blicly-financedschools (Box 4.1). In Latin America, several Central American countries have introduced reforms to devolve school management to communities, and these community-managed schools have proven effective in ensuring access to school and good quality results @i Grope110 and Marshall (2005) and Sawada and Ragatz (2005). Box 4.1: How Finland Recruits and Motivates Teachers Faith in education is not just a matter of Government policy in Finland; it is deeply rooted in the national character of a country that has few natural resources beyond fish and timber. Finns came to realize more than a century ago that education was the best antidoteto their endemic poverty. Himanen, the phlosopher said: "The development of Finland has come through investing in the educationsystem." The PISA exam, administered by the Organization for Economic Cooperation and Development in Paris, confirms that Finland does remarkably well in schools serving all levels of this egalitariansociety. "Fmland and Japan stand out for high standards of both quality and equity,"according to the organization. Karkkainen enjoys substantial autonomy as principal of Arabia Comprehensive School in Helsinki. With her governing board -- consisting of five parents, two teachers, a representative of the blue-collar workforce and herself -- she can set the school's curriculum, write its rules and hue its teachers. She explained over lunch how she hires a teacher. She recently announced on the Helslnki school system's computer network a vacancy for a senior Finnish-language teacher. Thirty-seven qualified teachers applied, nearly all with experience. Typically, a new graduate of one of the universities that train teachers here has to work part time or in the provinces for about five years before landing a full-time position in the capital. Karkkainen and the board interviewed five candidates and picked the one who will start at the beginning of the next school year, in mid-August. Arabia Comprehensive is not never-never land; during a day spent in the school, a visitor saw many reminders that luds are kids, school is school, learning is always a combination of rote, fun and angst. But some things were striking. One was the relative maturity of students. The Finns long ago decided that 7 is the right age to begin school, so in every grade the children are a year older than they would be in the United States. Six-year-olds have kindergarten (and a high percentage of Finnish youngsters come to school from State-run day-care centers, which are also generously staffed and supported). But according to Raili Rapila, a kindergarten teacher at Arabia, there is no pressure to begin reading before the first grade. Three of 10in her class are readers, she said, but all 10love to be read to, and are often, every day. "Social skillsand learningto play are more importantthan reading" for the 6-year-olds, she said. In Riitta Severinkangas's eighth-grade English class, one 14-year-old boy conspicuously refused to participate in a recitation of phrases and listened to his iPod instead. "Do you choose not to participate?" Severinkangas f d y but gently asked him, loolung right into h s face. He didn't reply. He looked like the classic tuned-out, bored teenager. But a few minutes later he was conversing in good English with the American visitor, describing h s "personal record" for hours worked at his computer without sleeping -- 76, he claimed. How did he stay awake? "More cola, more coffee.It's easy when you have work to do." And what was the work? Translating Englishpop niusic lyrics into Finnish, for example. Source: Kaiser, 2005. 4.59 Teachers' central role in ensuring school quality. Attracting and retaining highly qualified and motivated teachers is a necessary step to ensure improvements in education quality. This will require a medium-term alliance with teachers and their participation and commitment to educational reform. 4.60 What would be the components of an education system geared toward attracting and retaining qualified teachers and motivating them to perform to the best of their capacity? There are at least seven components that can influence teaching quality, which are summarized in Box 4.2.1°7 First, effective teachers have access to adequate teaching materials and are hrnished with basic infrastructure. As explained above, Chile has made important progress in that direction. 4.61 Second, effective teachers know what is expected of them. Although, in many settings, education standards are a sensitive issue, all education systems have established standardsfor their public school teachers. For example, most systems expect teachers to have completed a minimum number of years of education; many systems expect teachers to have completed teacher training programs andlor teacher certification programs; most systems expect teachers to work a number of dayshours per week and year. A system that is able to recruit and retain highly qualified teachers is one where there are clear expectations for teachers in terins of the subject knowledge and pedagogical skills that teachers need; provides specific guidelines regarding how teachers should behave and what they should do; and makes explicit the specific conipetencies and learning goals that teachers are expected to occasion in their students. Without clarity in what is expected from teachers, it will be very hard to establish incentives for teachers to meet expectations. 4.62 Third, how teachers (and principals) are selected and assigned to schools affects .the quality of teaching and student learning. In many countries, these rules are not clear or transparent, leaving the process open to political influence. When clear rules exist, they are often insufficientlylinked to policies that attempt to target resources to areas of highest need. 4.63 For instance, instead of assigning the best-performing teachers to schools serving studentswho are in most need, some countries give priority of choice in school assignment to experienced teachers, thus leaving schools serving more disadvantaged populations with new, less-experiencedteachers. A purposeful mechanism of teacher assignment may involve devolving authority to school principals or administrators, enabling them to choose the teachers with the best qualifications and characteristicsto serve the students in their schools. Rules for selection and criteria for assignment should be clear, transparent, and rational. Without rules for teacher selection and purposeful assignment to schools, the system is open to political influence and ends up with the least effective teachers teaching the most vulnerable children. 4.64 In Chile, it has been difficult to introduce competitive criteria for the selection of principals (concursabilidad). Fortunately, Parliament recently approved a law to introduce competition in the assignnient of new school director positions. That is a step in the right direction,but competitivecriteria could be used to assign all school directorpositions. lo'This section is based on Vegas and Umansky (2005). 116 4.65 Fourth, on a regular basis, education systems need information about teaching and learning and measures of progress against the expectations that have been set out. This information needs to be used not only for accountability purposes but also as a tool to improve teaching practice. Monitoring and evaluation systems need to address at least three questions: (1) Do existingteachers possess the expected knowledge and skills? (2) Is teacher performance currently meeting the expectations that have been agreed upon? and (3) Are students learning the skills expected of them and making progress in acquiring those skills at the expected rate? Without evaluating teachers' knowledge and skills, teacher behavior and performance, and student learning results, neither policymakers nor teachers themselves can know to what extent teachers are meeting expectations. 4.66 In this area, too, Chile has made important progress in recent years. In 2000, voluntary individual evaluations for teachers were introduced during the collective bargaining round (Asignacidn de Excelencia Pedagdgica, or Teaching Excellence Allowance) that is linked to pay increases. In addition, there is a mandatory individual evaluation requirement for municipal school teachers, though it has encountered enormous opposition and implementationdifficulties.In the near fbture, it will be important to evaluate the impact of the TeachingExcellence Allowance on teaching quality and student learning. 4.67 Fifth, good teaching requires continuous learning; getting all students to learn requires that teachers in each school work together in crafting strategies for every student. Thus, while an effective systeni for attracting and retaining good teachers is one where there are clear expectations of the skills teachers should have, the behaviors they should exhibit, and where these are monitored and evaluated on a continuous basis, it is just as essential to use this informationto provide teachers with the technical and managerial support needed, as well as the professional development opportunitiesnecessary for them to succeed in the task of engendering student learning. This implies having school administrators who are also instructional leaders, working closely with teachers in using the evaluation information to help teachers improve their subject knowledge and pedagogical skills in targeted areas. Also, it implies that school instructional leaders encourage and support teachers in working with each other to produce better results for all students in the school. Finally, it requires that school administrators allow time for professional development and teacher collaboration within the school day and year. Without providing teachers with instructional leadership and professional development opportunities, they cannot be expected to acquire new skills, improve their teaching, or build the teacher professional communitiesneeded to produce high levels of student learning. 4.68 Currently, many Chilean teachers in the second cycle of primary school have general education degrees without subject-specific knowledge. Less than a half of Chilean students have science teachers with degrees in the basic sciences such as physics, chemistry or biology. In contrast, in countries such as Russia and Taiwan, more than 90 percent of science teachers have science degrees and have complemented their trainingwith pedagogic studies. 4.69 Sixth, effective teachers have the authority to use their best professional judgment in determining what is needed to obtain results against the agreed-upon standards. Without authority, for example, to choose among different teaching methods, teachers cannot be held responsible for failing to accomplish learning goals. Such autonomy appears to exist in Finland, a country that has achieved substantial progress in education indicators in recent decades and boasts the highest student test scores in PISA (Box 4.1). 4.70 Similarly, effective school leaders and school management teams have the professional autonomy to provide teachers with the material and technical resources necessary to support teachers in meeting the agreed-upon objectives. Importantly, school leaders and school management teams need to have the authority and resources to reward high-performing teachers and to penalize low-performing teachers. Research on effective schools in Chile highlights the role played by the principal's leadership, his or her expectationsand vision of the future (Bellei et al., 2003). Box 4.2: Components of an Effective System for Attracting, Retaining and Motivating Highly Qualified Teachers Adequate infrastructure and teaching materials;basic resources Clarity in what is expected from teachers What knowledge and skillsteachersneed to have. What behavior and performance teachers should exhibit. What results in terms of student learning we expect teachersto accomplish. Clear, transparent rules for teacher selection and purposeful assignment to schools How teachers are selectedand assigned to schools affectsteaching and learning. Monitoring and evaluating teaching and learning What knowledge and skillsteachershave. What behavior and performance teachers exhibit. What students are learning and at what rate they are making progress. Instructional leadership and professional development for supporting teacher professional communities School administratorsas instructionalleaders. Using information from teacher evaluation to develop professional development opportunities tailored to each teacher and team of teachers. Provide time for teacher professionaldevelopment and teacher collaboration. Professional autonomy and authority Teachers can use their best professionaljudgment in the classroom. School administratorshave professionalautonomy to provide teachers with support,in both material and technical resources. School administratorshave authorityto reward high performing teachers and penalize low performingteachers. Effective teacher incentives Appropriaterelative salaries. Higher salariesfor better performing teachers. Higher salariesfor teachers working in disadvantaged areas. Advancement opportunitiesthroughout the teacher career. Recognitionfor excellent teachers. Real threat of losing one's job for poor performing teachers. Teachershave to convince clientsthat they are performing adequately. I I Source: Vegas and Urnansky (2005). 4.71 Finally, since individuals respond to a variety of incentives, multiple incentive policies exist to attract and retain highly qualified teachers, and to get teachers to work hard to raise student learning. These include: appropriaterelative salaries - that is, salariesthat are at least as high as those available to individualswith similar experience and qualificationsin other professions; higher salaries for better performing teachers; higher salaries for teachers working in more difficult conditions (e.g. remote areas, disadvantaged populations); a well- defined teacher career with advancement opportunities; public recognition and prestige for excellent teachers; a real threat of losing one's job due to inadequate performance; and a relatively close client relationship where teachers have to regularly convince principals, parents, or communities that they are performing adequately. 4.72 Chile has made important reforms to teacher incentives,more so than many countries. Between 1990and 2002, real teachers' salaries increased 156percent. During this period, the Government launched a publicity campaign to encourage students to become teachers and created a scholarship program for o~~tstandingstudents to study pedagogy. Simultaneously, the Government allocated substantial additional resources to schools, thus improving overall working conditions for teachers. While it is difficult to separate out the effect of those initiatives there was, during the same period, a 39 percent increase in the number of education students and the average score for applicants to education programs increased 16 percent. Other degree programs did not experience a comparable increase in the quality of applicants. In engineering, for instance, the average entrance exam score remained more or less constant. The outcomes suggest that the increase in salaries has led to an overall improvement in the quality of teacher applicants. 4.73 However, the structure of teacher salaries in Chile continues to be relatively flat, especially when compared to that of non-teachers (Mizala and Romaguera, 2005(a)). Seniorityis the principal factor that governs increases in teacher pay. There is little room for pay differentiation based on a teacher's activities or effectiveness in the classroom and school. In this system, high-performing teachers are underpaid. Figure 4.17 shows that, in spite of the awards that have been introduced for teaching performance, a high proportion of the teacher's salary supplement is the result of years on thejob rather than Figure 4.17: Allowances and Incentives for Teachers. 2002-2003 IHTenure I Training Difficult conditions I I1IAdministrative positions I1 lilIndividualincentives u I l!d Master teacher HSNED Source: Cox (2005). In each categorythe maximumamount is stated. '08Mizala and Romaguera (forthcoming) provide preliminary evidence that the SNED incentive has had a cumulative positive impact on student performance for those schools that have a relatively good chance of winning the award. 4.74 Beginning in 2006, the school-based performance award for teachers, Sistema Nacional de Evaluacibn del Desempeiio de 10s Establecimientos Educacionales (SNED), will constitute a bigger proportion of a teacher's salary. The size of the award will have increased by 91 percent between 2004 and 2006, and as a result, teachers in a SNED-winning school will earn about double of what they earn today. Ensuring equity of educational opportunity 4.75 Ensuring equity of educational opportunityremains one of the biggest challenges for Chile. Table 4.8 shows the differencesin eighth-gradetest scores by socioeconomicgroup in the 2004 SIMCE. Students from low and medium-low socioeconomic backgrounds have substantially lower test scores than do students who come from more privileged socioeconomicbackgrounds. 4.76 Recent literature on peer effects in education has provided strong empirical evidence that the grouping of students within schools and classrooms has important implications for quality and equity. The underlying argument, as presented in Hoxby (2000(a)), is that the makeup of a classroom - students' average family income, the number of children with disabilities, its racial and gender balance - can affect peer attitudes. For example, children with learning disabilities may draw disproportionately on their teacher's time; racial or gender differences in the classroom may interfere with learning; wealthier parents may purchase learning resources that later become available for the entire classroom. Teachers' expectations may be lower for students from minority backgrounds, and thus teachers might lower their academic standardswhen confronted with a classroom that has a high proportion of them. As a result, other students in the classroom may respond negatively, not because of the minority students' presence but because of the teacher's attitude. In an empirical evaluation of the impact of segregation in classrooms by gender and race, Hoxby (2000(a)) finds that a students' academic achievement is, indeed, affected by the achievement of his or her peers. Table 4.8: SIMCE 8th Grade Test Results,2004 Socio economic group Language Mathematics Low 229 232 Medium low 234 235 Medium 253 253 Medium High 280 282 High 301 311 National Average 251 253 Source: SIMCE, MINEDUC. 4.77 Students from marginalized socioeconomic backgrounds require additional resources and attention to achieve good educational outcomes. The current student-based subsidy is provided equally for all students, independent of the student's socioeconomicbackground. A uniform voucher leads to significant differences in educational results and encourages discrimination against lower socioeconomic students by schools that are able to select their students. Several countries have introduced differential vouchers to compensate for differences among students.For example, in the Netherlands,vouchers are adjusted by socio- economic level and ethnicity; in Colombia and in the U.S., vouchers were introduced only for the poorest segment of the population, either to provide access to the educational system (PACES, Colombia), or to improve the quality of schools (charter schools in the USA). In these latter cases, if there is excess demand, places are allocated by lottery. In addition, several countries expresslyprohibit schools from admitting students on a selectivebasis. (e.g. the Netherlands, Sweden). 4.78 One way that Chile could reduce discrimination against students from low-income households would be to provide a higher student-based subsidy to schools that admit them. This would allow students with less resources to move, if they so wish, to better quality schools, thus contributing to greater social integration and allowing them access to better education facilities. If students decide to continue at their present schools, these would receive additional resources from the differentiated subsidy.lo9The President of Chile announced on May 21st that the Government will propose legislation to Parliament for the introduction of a student-based subsidy differentiated by socioeconomic background. The law would provide for a preferential subsidy to qualified students in pre-kinder through the fourth grade of basic education. 4.79 In a system of subsidies differentiated by student socio-economic background, schools would comply with certain quality standards to receive students with the preferential subsidy and they would sign an Equal Opportunity Agreement (Convenio de Igualdad de Oportunidades)with the Ministry agreeing not to admit students selectively. 4.80 Chile also faces the challenge to improve the equality of access to pre-school education. While most children from the highest income quintile attend pre-school, the proportion of children fi-om low-income families attending pre-school is very low. The low rate of pre-school enrollment among the poor is troubling because pre-school attendance has been shown to have positive effects on subsequent scholastic performance. Raising pre- school enrollment rates overall and especially among the poor would likely lead to better educational outcomes and greater equality of both educational and labor market opportunities. 4.81 Greater equity of access to higher education is also important. Secondaryand tertiary institutions in Chile could work closely together to bridge the gaps in access to tertiary education. To a great extent, inequities in tertiary education stem from problems of quality in secondary schools and low completion rates among poor and indigenous students. These shortcomings are reflected in the university entrance examination results (PSU). Among the highest 20 percent of PSU achievers, only 6 percent come from the lowest income quintile, while 34 percent come from the highest income quintile. The poor would have greater access to tertiary education if remedial and preparatorycourseswere to be made available. 4.82 Financing needs to be made more readily available for students in tertiary education and recent efforts to expand access to financing deserve to be strengthened. In the medium- term, it would be desirable to develop a comprehensive system for financing education in logThe allocationof additional resourcesby school rather than by studentcan lead to problems of exclusion; see Gonzalez, Mizala & Romaguera (2004). tertiary education institutions, covering both traditional and non-traditional universities as well as non-universityinstitutions. 4.83 Finally, it would help improve quality if resources were to be allocated to tertiary education institutions within a results-based framework. Only 5 percent of direct public hnding (AFD) for traditional universities is allocated on a true performance basis. Accountability for results could be enhancedwithout compromisinginstitutional autonomy if financial incentives were to be tied to the achievement of key sector priorities. The planned introduction of performance-basedagreements on a pilot basis would be an important step in that direction. The agreements could gradually be refined and scaled up so that the current fimding system can be supplementedby a performance-based system applied to a significant proportion of universitybase funding. In parallel, efforts could be made to gather information on the performance of tertiary education institutions. The proposed accreditation law would provide a clear mandate to commencethis task (Thorn et al., 2004). ProvidingInformation to Improve the Functioning of the Education Market 4.84 If a competitive education system is to effectively foster quality, parents need to be fully informed about school quality. Well-informed parents can exercise control, denland quality from the schools and, if the school does not respond, transfer their children to other schools. In this sense, information on the quality of basic and higher education provided by schools and universities may be considered a public good. To be used effectively, this information could be comparable among institutions, easy to understand, timely, and accurate. The greater the proportion of well-informed consumers, the more efficient will be the hnctioning of the education market (Gonzklez, 2004). Indeed, the success of market- based education reform depends on the extent to which parents can make informed decisions concerningthe range of school options that are available. 4.85 A recent study by Elacqua and Fabrega (2004) finds that the education reforms in Chile have not generated a critical mass of informed consumers that can exert pressure for school improvement. Parents, they find, consult few sources of information and are not effectively linked to informationnetworks. They look at very few schools, and they choose a school for practical reasons on the basis of very little specific information. The authors also find that the education system does not do a goodjob in disseminatinginformationto a broad public. Parents themselves must take the initiative and the informationnetworks they consult and the quality of informationthey obtain is related to their socioeconomic background. 4.86 It is costly for families to obtain relevant, up-to-date information on what is happening with their schools, and schools are not given incentives to provide informationto parents. Thus far, information on school qualityprovided to parents is limited to the average SIMCE score for the school. This information is not very usehl for parents because, as mentioned above, school averages mask important differences in school quality. More importantly, as Elacqua and Fabrega (2004) document, parents are concerned with school factors other than test scores. In an Appendix to this chapter, an example is given of a "school report card" for an elementaryschool in one of one of the best public school districts in the U.S.A. - Montgomery County, Maryland. The report card is publicly available and is posted on the website of the County. It includes information not only on student test scores by student background, but also on the characteristicsof the student population attending the school, on school facilities, programs offered, teaching staff credentials and experience, and the school's operating expenditures. The Evaluation of Education Policies and Programs 4.87 It is necessary to evaluate education policies and programs in order to adjust them during implementation and in order to design effective reforms for the improvement of education quality and equity. Useful evaluations tend to be expensive, as they require highly skilled staff and accurate information, which is often costly to collect. The lessons from policy and program evaluations have positive externalities - benefits that accrue to others apart from the individuals who bear the costs of the evaluation. As a result, Government support for the evaluation of education policies and programs is important. In Chile, large scale reforms have often been undertaken without much thought to their evaluation. Without a thorough evaluation it is not possible to leam from experience and to use the lessons of experience to design and implement future reforms. 4.88 Program evaluation is appropriately considered at the time of program design. That enables infomiation to be gathered, comparator groups to be identified and experimental projects to be undertaken in the course of program implementation. For example, in great part as a result of the convincing evaluations of the PROGRESA program in Mexico, conditionalcash transfers have become a popular instrument for reducing inequality. 4.89 Similar efforts could be made in education in Chile. For example, the introduction of a preferential subsidy for students from low-income households as well as the introductionof a new student loan scheme for students attending non-traditional universities and non- universityinstitutionscould be accompaniedby comprehensive impact evaluation. 4.90 The Chilean education system has made substantial progress in assembling educational indicators and data that may be used for the evaluation of policies and programs. However, more can be done in that respect. Panel data of students, teachers and schools would enable researchers and analysts to conduct better quality and more conclusive empirical evaluationsof educationpolicies and programs. SUMMARY KEYPOLICYMESSAGES OF AND OPTIONS RAISEEDUCATIONQUALITY TO AND EQUITY CHILE IN 4.91 Notwithstandingsignificant gains in secondaryschool coverageand student retention, two important challenges remain for Chile's education system: improving the quality of education and providing more equitable access to educational opportunity. 4.92 The introduction of a per-student subsidy to private and public schools has enabled Chile to expand substantiallyeducation coverage and completion rates. However, the benefit of that financial arrangement is limited because public and private subsidized schools operate under a different set of rules which constrain competition. Important differences between subsidized private and public schools include the student selection process, teacher contracting, and the availability of financial resources. The main policy messages and recommendationsfrom this review are summarized below. 4.93 First, in Chile, where there is private provision of education in a competitive market context, the State has an important role to play to ease constraints in the education market and guarantee that all children have access to an education. The State has the authority to demand accountability for the public resources received by schools. In that regard, the differential student-based subsidy currentlybeing introduced provides an opportunity to link the eligibility for finance to results. Increasing accountabilityimplies that the Statehas a role in developing and enforcing standards, norms and incentives for the education system. To improve the effectiveness of State supervision and support, under-performing schools could receive greater attention and supervision, while high-performing schools could receive increased autonomy and authorityover financial and human resources. 4.94 Second, the State also has a responsibility to ensure equality of educational opportunity. Recently, the administration announced plans to introduce a student-based subsidy differentiatedby student socioeconon~icbackground. This is a welcome step towards reducing selective school admissions that result in the segregation of students by socioeconomic background. To ensure quality and equity, it will be important for schools that receive the differential subsidy to meet eligibility criteria requiring them to limit or eliminate unfair student selection practices and holding them accountable for the use of additional resources. Similarly, the recent initiative to expand student financing options for tertiary education will likely improve equity in higher education. Greater access to pre- school, especially for the poor, would also do much to improve equity in educational opportunity. 4.95 Third, there is an important role for the State to improve the information available to parents and civil society on school quality. Good information on school quality serves as a public good: the greater the proportion of well-informed parents, the more efficient will be the functioning of the education market. In Chile, studies have documented the poor quality of informationthat reaches parents regarding school performance. The State has an important role to play in ensuring that accurate information, comparable between schools, is made availableto the public in a form that is timely and easy to understand. 4.96 Finally, the evaluation of education policies and programs is an important function to be supported by the State. Without State support, evaluation would not receive the attention or the investment that it requires because the benefits due no accrue to private parties. Program evaluation should be anticipated at the time of program design, thereby building into the process the arrangements necessary to collect information, to build comparator groups and experimental groups (affected by the program). Improved data and systematic evaluations of education policies and programs will enhance the quality of current programs as well as hture reforms. APPENDIX: EXAMPLE SCHOOL OF QUALITY INFORMATIONPROVIDED THE PUBLIC TO Wood Acres Elementary School - #417 Prlnclpal Ms Annette C Hall 5800 Crornwell Dnve Bethesda, MD 20816 m C e Phone (301) 320-6502 ComrnunltySupt Dr Frank H Stetson wvvw m c ~ k12.rnd uslschools~oodacresesl s Fax Number. (301) 32C-6536 Schod Hours 845-3 05 Cluster Name Walt Wnitman FeederSchools: Recelvlng Schods Pyie 2004-2005 Enrollment = 622 1 -- I Students with Dlsabllltlms Least Restrlellve Envlmnment (LRm I I Grade3 1 131 1 21 1 Percentof Instructcnal TomeO~tsldea General Educat~mClass ---- - - I I C r a n e 1 1 9 A 1 1 5 1 I I - I - I Less than 21% I BeWeen 21% and 60% I Mcre than 60% I IGrade5 1 99 1 159 All SPED Students I 70 7 8 6 20 7 cy , V"Uus;, %L &:~%SC~&LIIP~O&&~~~*;~~-! g " = + -+> Leaming and Academn: D~sabllltles(K-5) InclusionSpecial EducatronModelfor LAD R Resource(K-5) Attendance Rate '= 96 6% Suspension Rate *= 0 0% I I Maryland School Assessment Pmflciency Rate* I Grade 3 - --. - I Grane 4 - --. I Graoe 5 -.. Math I Resdlng Math I Readlng Math I Reading All Students 92 1 89 1 97 9 I 96 8 99 0 I 98 1 Afncan American I 1 FARMS I I I I I I I 2003-2004school Year Data Wood Acres Elementary School #417 - B ~ d e n t l l n s t ~ ~ t8affRatlo = 13 4 l ~ a l Average Class S~ze K~ndergarten=25 0 Grades 1 to 2 = 23.6 Grades 3 to 5 = 25 8 lnstruct~onal Support 1000 Paraeducators 1000 Regular Speaal Educabon 000 Teacher Asslotant Medm Assistant lnrtrudlonal Data Assistant 0 750 lnftrumenhl Muslc Other SuppoitTotal Speaal Edcabon Resource Rogmm 1 500 BulldtngSeMces 5. ENHANCING SOCIAL PROTECTION: STRIKING THE BALANCE BETWEEN INCOME SECURITY AND ECONOMIC OPPORTUNITY 'lo "Theappeal of State relief is that it is available to eveiybody but it leaves to eveiy one a strong motive to do without it if he can. The State does not test a person's means, but by offering low wages in returnfor hard work it deters those who can support themselves. In theory, benefits should be close to, but slightly below, the market wage, giving the greatest amount of needful help, with the smallest encouragement to undue reliance on it." John Stuart Mill 5.1 Although social contracts differ widely between countries, social protection policies are generally intended to safeguard the investment of households in human capital. Social protection systems (comprising policies, public institutions, and the regulation of private institutions) aim at helping families manage shocks to their income that can threaten their endowed and accumulated human capital. Given the well established links between human capital and economic wellbeing, social protection policies are a core part of developmentpolicies to fight poverty and increase equity. 5.2 Protecting human capital is also critical to promoting and sustaining economic growth. Where households do not have a sufficiently wide array of tools with which to insure against risks, they are condemned to spend precious resources recovering from shocks that could otherwise be put to productive use. Recovery from certain shocks can even impair a household's human capital or hinder further investment. When there are few options for mitigating the losses from shocks, households may be deterred from taking appropriate risks in their investment decisions-adopting a new, more productive technology, for example-fearing the cost of failure. But since there are many risks which private insurancemarkets cannot handle there is a clear role for the Government to design instrumentsthat will help households adjust to shocks. 5.3 Social insurance-the dominant component of the social protection infrastructure in most countries-should prevent households fkom falling into poverty in the wake of an income shock. Effective social assistance gradually lifts the poor above the poverty line. Thus, country poverty indicators are ultimately the most relevant and objective criteria against which to evaluate socialprotection policies. 5.4 Judged by these criteria, Chile has done extraordinarily well, even if the progress in eliminating poverty is attributable to a broader set of policies than social protection alone. For most of the 1980sand 1990s,poverty in Chile declined. This achievement was 110This chapter was prepared by Truman G. Packard, with the help of Cristian Baeza, Theresa Jones (LCSHD, World Bank), Claudio Montenegro (DECRG, World Bank) and Ximena Quintanilla (Institute for Fiscal Studies, University College London). We are indebted to Monika Queisser, Edward Whtehouse, Alain Deserres and Thomas McGirr at the Organization'for Economic Cooperation and Development (OECD) in Paris, and to Fabio Bertranou at the InternationalLabor Office (ILO) in Santiago, for much of the data that appear in this chapter. mainly due to economic growth-the fruit of market-friendly policies and prudent economic management-as well as to increased, targeted social spending. However, although there are fewer poor households, indigence has remained at stubbornly persistent levels since the mid-1990s and has only recently begun to fall. Further, the distribution of income in Chile is among the most unequal in the developingworld. Many households remain vulnerable to poverty from adverse income shocks from job loss, sickness or injury, and losses associated with old age. , 5.5 This chapter builds on and updates recent analysis in the World Bank's 2004 report "Household Risk Management and Social Protection in Chile" (World Bank, 2004~).That report concluded that Chile largely succeeds in providing households with the instruments that they need to effectively mitigate shocks and protect their human capital. The institutions Chile has put in place to help households lower losses from these shocks-from the new unemployment insurance system initiated in October 2002; the pioneering retirement security system that has served as a model for pension reform around the world since 1981; and the mixed health insurance system that was recently complemented with a guaranteed basic package of services-are appropriately designed to address of the risks to household welfare they are intended to cover. That said, too many Chilean households--even among the non-poor-still do not have access to these sophisticated,state of the art instruments of social protection. 5.6 Lack of coverage is a matter of concern. The evidence presented in the 2004 report shows that the principal shocks to income are highly correlated. Households who earn incomes close to the poverty line are particularly vulnerable to an array of shocks (health, disability, and the costs of unexpected additional children or other dependents)if members have lost employment. If they have lost a job in the informal sector or are among the self-eniployedwhose businesses fail in economicdownturns,this vulnerability is compounded since they face explicit and implicit institutional barriers to even basic forms of social protection. And since many of these uncovered workers are not counted among the poor, the relatively well-targeted social safety net that Chile has in place- including its flagship facility for the most needy, Chile Solidario, launched in May 2002-will not catch them soon enough should they fall. Thus the over-archingchallenge for policy makers in social protection is how to improve coverage among these vulnerable groups without creating moral hazard or dependence, and while fostering opportunity and enterprise. 5.7 The objective of this chapter is to exploit new data that have become available since the last report, to assess how recent changes in the Government's social protection policies have given new options to households managing risks to their welfare. Re- examining the conclusions of the last report with the benefit of new data, provides an opportunity to review the findings and conclusions of later empirical research conducted in the Bank and elsewhere. 5.8 This chapter on social protection has four additional sections. Section A updates the analysis of the principal risks to income throughout life as presented in World Bank (2004~).It discusses changes both in the indicators of risk to household welfare, and in the coverage of programs Chile has put in place to help households manage those risks. The section takes a preliminary look at the progress achieved since 2002 in the fight against indigence with Chile Solidario."' Section B presents the strengths and weaknesses of the instruments designed to cover losses from unemployment, the cost of healthcare, and old age, drawing from the experience of other countries (primarily OECD countries) to identify best practice. Section C poses critical questions for policy makers, and presents options for reform to improve the efficiency and efficacy of the social protection system. Section D concludes with suggestions for htwe social protection policy in Chile. CHANGES THE RISKSTO HOUSEHOLD IN INCOMES AND THE COVERAGE OF SOCIAL PROTECTION 5.9 This sectiondiscussestrends in risks to household earnings-capacityand incomes. It also reviews changes in the coverageof programs that the Government of Chile has put in place to help households manage the risks. The intent is to update the analysis in World Bank (2004~)by showing changes in risk indicators from the surveys of living standards(CASEN) conducted in 2000 and 2003. 5.10 The primary risks to earnings-capacityand income throughout life are discussed in Box 5.1. Table 5.1 and Table 5.2 present the main indicators of risk and the changes over 2000-2003. The first income decile is roughly equivalent to the indigence line.'12 The second income decile is roughly equivalent to the poverty line.'13 Households in deciles 5-7, correspond to households earning average incomes. The risk indicators are presented separately for the urban and the rural population and for men and women where relevant. "' This chapter includes very preliminary material on some of the outcomes of the Chile Solidario initiative, and is not intended as a defitive evaluation. The World Bark has been supportingthe design of a rigorous evaluation protocol for Chile Solidario since the start of the initiative. 112To be more precise, indigence in Chile in 2003 was 3.9%. 113Poverty in 2003 was 15.4percent, and thus lay between the first and second deciles. 129 Table 5.1:Main Indicators of Risk to Earnings Ca~acitvandHousehold Income Urban Households Rural Households Age Group Household Risk Indicator Decile 1 Decile 2 Deciles 5-7 Decile 1 Decile 2 Deciles 5-7 Malnounshed (incidenceundemounshed, underweight or overweight) 10.9 11.5 11.0 12.0 11.2 15.3 From 0 to 5 years Does not attend to pre-school(only 5 years old) 26.6 26.4 22.1 56.4 43.7 43.2 Does not attend pnmary school (6-11years old) 8.2 8.9 6.8 9.0 5.9 8.5 Age does not correspondto grade level (6-11years) (a) 9.1 6.3 4.8 11.4 6.6 5.2 Does not attend primary nor secondaryschool (12-14 years) 8.2 7.6 5.5 5.9 6.4 8.3 From 6 to 14years Age does not correspondto grade level (12-14 years) (a) 23.0 18.1 10.2 30.7 18.2 18.2 Child labor (12-14 years) (b) 1.4 1.3 1.5 1.O 1.3 1.O men 3.8 1.4 0.8 1.6 3.3 1.3 Inactivity-doesnot attend to schooland doesnot work (12-14 years) women 3.2 1.4 0.6 3.0 1.9 4.2 Does not attend secondaryschool(15-17 years) - . 28.6 19.9 12.4 37.6 36.6 32.2 I Does not attend universitv(18-24 Gears) - 1 90.7 86.2 74.6 94.8 96.6 94.8 > - , men 49.3 33.1 12.1 41.4 23.1 8.5 Unemployment (15-24 years) women 70.0 44.4 18.5 62.9 41.4 21.1 From 15to 24 years men 11.9 8.1 3.4 13.9 15.5 6.9 Inactivity-does not attend to schooland doesnot work (15-17 years) women 13.5 8.1 5.9 17.0 16.5 20.2 men 1 45.2 31.2 13.8 47.6 32.5 15.9 women 1 38.9 25.5 8.9 23.1 12.8 5.2 men Labor Force Participation women men Workers withouta contract (c) women men Workers without a indefiniteperiod contract women men Workers that see theirjob as non permanent From 25 to 64 years women men Self-employed women men Non professionalself-employed (% of self employed) women men Part time workers (% of all employed) women Non educated or with incompleteprimary education (25 to 40 years) Non educated or with incompleteprimary education (41 to 64 years) Head of household with dependents(d) men Elderlyadult (% of the total population) women Elderlyadult livingalone or with spouse 65 years and up (e) men 39.2 16.0 5.7 6.7 6.0 4.3 Unemployment women 19.3 10.3 4.5 0.0 2.9 0.1 men 16.2 18.5 26.0 13.9 21.5 31.5 Labor Force Participation women 3.4 4.9 5.4 1.5 4.1 3.9 I Reportshealthproblems (sicknessor accident) 1 26.7 23.9 22.6 24.0 23.5 23.2 Notes: (a) Childrenthat are behind in educationallevel accordingto their age. (b) Includes working and unemployed children. (c) To hold a contractacts as a proxy for having access to tenure based severancepayment,pensions,health insurance,labor health insuranceand unemployment insurance. (d) Heads of householdeligibleto receive family assistancewith dependents of the total head of householdswldependents. (e) Note that the legal retirement age is 65 and 60 years for men and women, respectively. Source: X. Quintanillawith data fromCASEN 2000 and 2003. 130 Table 5.2: Changein the Main Indicatorsof Risk to Earning Capacity and HouseholdIncome Part time workers(% of all employed) Notes: (a) Children that are behind in educational level for their age. (b)Includes working and unemployed children. (c) Hold a contract act as a proxy for having access to tenure-based severance payment, pensions, health insurance, work-injury risk insurance, and unemployment insurance. (d) Heads of household eligible to receive family assistance with dependents of the total head of households wldependents. (e) Note that the legal retirement age is 65 and 60 years for men and women, respectively. Source: X. Quintanilla with data from CASEN 2000 and 2003. Box 5.1: Risks to Income throughout life In the 2004 report, the risks to earnings-capacity and income were classified according to their relevance to different phases of life. For example, risks to future earning capacity for the age group 0 to 5 are malnutrition and failure to attend pre-school. Among chlldren from 6 to 14, risks include failure to attend school and child labor. From age 15 to 24 prior to (full) entry into the labor force, risks to future earnings capacity prinmrily arise from interruptions in the final years of education, and failure to complete secondary school or to enter university, that can later manifest in lower relative productivity and earnings. During working life-usually from ages 25 to 6 L t h e primary risks to income are typically unemployment, ill health and disability. A particular characteristic of Chile-remarkably low numbers of women participating in the labor force-can also be a risk to earnings capacity and income. Furthermore, in most countries, access to social protection is limited to workers who hold a legal employment contract, excluding informal workers without a contract and the self-employed. There is a large and growing literature suggesting that the self-employed often choose to run their own businesses rather than be employed under contract " but, unless they are also taking measures to protect themselves, they and their dependents can be especially vulnerable to adverse shocks to income -above all if they are non- professional self employed. The risks to income in the last stages of the life cycle are more subtle. The loss of earnings ability that comes with age forces individuals and households with an elderly household head to rely on previously made provisions to finance their consumption. Measures to secure an adequate income in old age vary widely from private savings, to national pension systems, and investment in family with the expectation that care will be reciprocated. Thus, it is not agingper se that poses a risk, but reliance on the returns and security of investments made in the past during working life to finance future income. Therefore, households headed by the elderly and those where there are no household members of working age, can be particularly vulnerable. Notes: a. Several recent publications focus on the search for formal employment in Latin America (see Maloney, 2004). Contrary to a large literature on labor markets in developing countries, these studies find little evidence that self employment is the residual sector in a dualistic labor market, or that the self employed are queuing along with the unemployed for formal job contracts. Data on movement in and out of the labor market and across sectors in Mexico, Argentina, Brazil and Chile show that growth in self employment is often greatest during periods of economic growth--a finding more consistent with an entrepreneurial"pull" into self employment, rather than the popular notion that workers are "pushed" out of formaljobs into small enterprises. b. While the self employed may rely relatively more on private resources for protection, many studies show that employees who do not hold an employment contract are vulnerable. Unlike self employment, informal wage employmentis not characterizedby free-entry. Employees of the informal sector are often indistinguishablein their age and education from the unemployed. In Chile,Packard (2002a)findsthat informal employees are more likely to have a greater number of dependentsas parents and heads of household than those still searchingfor ajob, and thus, are more likely to take up informal employment out of necessity. This raises the concern that informal employers may be unwilling to incur the costs of "formalizing" their workers and providing access to the national retirement security system. Source: Arriagada and Hall (2000). 5.11 From 2000 to 2003, there was a noticeable increase in pre-school attendance among children from urban households. This is a significantimprovement that bodes well for future education outcomes, given the known advantages of early school attendance for overall educational achievement. This increase could also lead to more women entering the labor market. While the increase in pre-school attendance has been substantial among five-year-oldsfrom indigent households, the greatest improvementhas been among poor households. Thus the risk of early educational set-backs seems to have fallen considerablysince 2000. 5.12 For children ages 6 to 14,the picture is mixed. In both urban and rural areas there were improvements in school attendance among children of ages 6 to 11, although rural indigent households constituted a worrying exception. From ages 12 to 14, there was a notable increase in school absenteeism, particularly among the urban poor. Although, overall, there were fewer "inactive" children aged 12 to 14 (that is, neither attending school, nor working) in 2003, there was an increase in the number of "inactive" children of both sexes from urban indigent households. 5.13 For youth, the picture in 2003 improved significantly compared to 2000 and the risks to earning-capacityand income shown in Table 5.2were lower. However, while the rate of inactivity among young women fell, a greater proportion of young men from indigent and poor households in rural areas was neither studying nor participating in the labor market in 2003. The share of young men and women who were looking but failed to find employment also rose noticeably, almost doubling aniong indigent households in rural areas. 5.14 For adults well into their working lives (ages 24-65), the picture is also mixed. While labor force participation increased for most, more women from urban households remained inactive in 2003 than in 2000. Furthermore, the fall in unemployment in urban and rural areas was uneven. Unemployment increased among women in urban households, and aniong men from indigent and poor households in rural areas. What is promising is that although the rate of ~memploymentamong men from indigent urban households remained the same, more women from indigent households who were looking for employment actuallyfound ajob. 5.15 Furthermore, in 2003, a greater share of workers aged 24-65 found employment with a labor contract than in 2000, particularly those from indigent and poor households in ~~rbanareas. Again, this is a positive developnlent given that the coverage of almost all of Chile's social protection instruments is limited to employees with a formal labor contract and their households. However, there is no clear trend in the proportion of self- employed (other than an intriguing increase in the number of women who are self- employed). As argued previously, a move into self-employment is cause for undue alarm. However, the extent of non-professional independent workers among the self employed could be a cause for concern. For all income deciles, in urban as well as in rural areas, the share of non-professionalsamong the self employed was well over 90 percent and in some cases 100percent in 2003. 5.16 Among the elderly, there was an increase in the share living alone or with their spouses. The most dramatic increase in the elderly living alone was among the elderly in poor urban households. Labor force participation among the elderly fell, although there was a dramatic increase among elderly women from indigent households in urban areas. More elderly women from the poorest households in urban areas entered the labor market. In most instances, there was an increase in the share of elderly who found work. However, the increase in unemployment among elderly men and women from poor rural households may be cause for concern. 5.17 To summarize, there was noticeable reduction in risks to household income and human capital for every age group. Risks are lower, particularly in the area of education. However, the changes in risks to earnings capacity aid income directly related to the labor market-from low labor force participation; unemployment; non-professional self- employment-has been mixed, with the poorest households in rural and urban areas still showingrelatively greatervulnerability. 5.18 Table 5.3 and Table 5.4 show the coverage and change in coverage over 2000- 2003 of Chile's main social protection programs. Among ages 0 to 5 there was a slight fall in the receipt of food from clinics, particularly among households in the lowest income levels. This could reflect the fall in malnourishrnent observed for this group. Among children ages 6 to 14 the proportion receiving meals at school rose across the board and by a substantially greater amount for children from indigent households. There was a dramatic increase in the coverage of State educational scholarships, particularly among indigent and poor households in urban areas. Similarly positive changes can be seen in the coverage of nutrition and educationprograms among youth. Table 5.3: Coverageof Chile's Principal SocialProtectionProgramsand Interventions Urban Households Rural Households Population Social Protection Indicator Decile Decile Deciles Decile Decile Deciles 1 2 5-7 1 2 5-7 From 0 to Receives food from the primary care clinic 83.6 84.4 64.8 89.7 91.5 87.4 5 years Attends State subsidized school (6 to 11 years) (a) 98.1 98.5 96.0 98.7 98.4 98.9 From 6 to Attends State subsidized school (12 to 14years) 98.1 98.8 95.7 99.0 99.8 99.4 14Years Receives food at school 61.9 51.1 26.6 94.5 90.8 82.6 Has a State's educational scholarship 3.9 4.5 5.5 4.0 3.6 2.2 Attends State subsidized school (I5 to 17years) 98.2 98.3 97.0 98.5 99.7 98.6 Attends State subsidized school (18 to 24 years) 95.9 98.8 91.3 99.7 99.4 98.2 Receives food at school (15 to 17years) 47.6 37.8 25.3 78.2 79.0 70.9 From 15to Receives food at school (18 to 24 years) 29.5 19.5 13.6 56.7 63.5 41.9 24 years Has a State's educational scholarship (15 to 17years) 10.2 11.4 7.4 21.6 13.9 11.7 Has a State's educational scholarship (18 to 24 years) 14.6 14.9 12.8 16.5 20.3 16.0 IReceives credit to pay university's fees (18 to 24 years) (b) 43.4 38.5 32.1 21.1 34.9 31.2 IHas attended some training 1 6.6 9.7 15.6 3.8 4.0 6.1 Has received unemployment subsidy or UI benefit (% of total unemployed) (c) health, employees (d) . pensions, employees (e) 56.6 70.1 81.6 20.2 45.8 65.1 From 25 to Access to social security health, self-employed (d) 87.6 86.9 80.9 92.4 94.1 90.8 64 years pensions, self-employed (e) 6.7 7.0 18.0 1.9 5.7 7.8 Receives food from the primary care clinic (f) 47.9 37.5 45.1 41.6 29.2 40.7 Received care during health emergency (% of those who asked for medical attention) 8.4 12.9 12.2 3.2 4.8 5.7 Receives household allowance (% of total number of people eligible for subsidy) 26.8 51.9 49.7 11.4 25.3 42.9 Receives subsidy from Chile Solidano (% of poor people) (g) 3.1 1.9 0.0 7.6 5.0 0.0 65 years Receives contributory old age pension (h) 18.2 38.8 57.1 1.1 18.7 38.0 and Receives assistance pension 52.3 16.6 5.6 90.3 65.8 31.2 Notes: (a) Does not include pre-school care. (b) Includes Government and private credits. (c) Includes both the old unemployment subsidy and the new unemployment insurance. l'he latter one is mandatory only for new entrants to the labor market but voluntary for workers who hold a job when the law came into force. (d) Includes household dependents. (e) and (f) % of the employees and self employed, respectively. (f) Food is given to pregnant or nursing mothers or in behalf of children younger than 5. (g) The subsidy can be received by any household member. (h) Note that the legal retirement age is 65 for men and 60 for women. Source: X. Quintanilla with data from CASEN 2003 Table 5.4: Changein Coverageof Chile's Principal SocialProtection Programs and Interventions (percentagechange in coverage from2000 to 2003) From 15to 24 sions, self-employed Notes: (a) Does not includepre-school care. (b) IncludesGovernmentand private loans. (c) Includesboth the old unemployment subsidyand pay-outs fromthe new unemployment insurance. The latter is mandatory only for new entrantsto the labor market but voluntaryfor workers who hold ajob when the law came into force. (d) and (e) % of the employees and self employed,respectively. (0Food is given to pregnantornursing mothers or inbehalf of childrenyounger than 5. (g) Note that the legal retirement age is 65 for men and 60 for women. This statistic under-reports coverageof contributorybenefitssubstantially,as survivor benefitsare not reported in the publicly availabledatabasesof the CASEN. Source: X. Quintanillawith data from CASEN 2000 and 2003. 5.19 Among economically active age groups, the proportion of the labor force that received some form of State-sponsored training or skills renewal grew, probably as a result of the Government's active labor programs (discussed in detail, in a later section). While coverage of income protection for the unemployed increased remarkably among the indigent, fewer unemployed workers from poor households were covered. There was an expansion in coverage of the contributory retirement and health-finance systems, particularly among salaried workers in poor and indigent households. There were even substantial gains in the coverage of health insurance and pensions among the self- employed. However, participation in the retirement income security system among the self employed in poor urban households fell. Among the self employed from indigent rural households, the fall in pension participationwas as high as 30 percent. 5.20 By 2003, there was a dramatic fall of in-kind food transfers to mothers with children less than 5 years of age, and to pregnant and nursing women. Furthermore, coverage of Chile's contributory family allowances (asignaciones familiares) program among poor households in urban areas, and among all income groups in rural areas also fell. However, this shortfall may have been made up with an expansion of the non- contributory single family subsidy. ' l4 5.21 Among the elderly, there was a significant increase in the coverage of contributory pensions in urban households of all income deciles, but particularly among the poorest urban households. There was also a significant increase in the receipt of contributory pensions by indigent and poor households in rural areas. The statistics under-report the coverage of contributorybenefits because survivor benefits received by widows are not included (see Rofrnan, 2005). The increase does not reflect any recent change in policy. It simple reflects the expansion of Chile's retirement security system years and even decades earlier.l15Furthermore, while the coverage of non-contributory cash transfers to the elderly remained relatively stable among poorer urban households, there was a substantial reduction in transfers to households of average incomes, indicating a possible improvement in targeting. The receipt of non-contributorypensions among the elderlyfrom indigent and poor householdsin rural areas increased. 5.22 To summarize, households in Chile are relatively well covered by social protection programs, and the beneficiaries tend to be in poorer groups. However, gaps in coverage are still apparent especially among the rural poor. From 2000 to 2003, Chile succeeded in narrowing many of these gaps in coverage, particularly among lower income groups. Progress in extending coverage among lower income groups reflects the Governments concerted efforts to eliminate indigence since 2002 (Box 5.2). Still, many gaps in coverageremain. 114The targeted Single Family Subsidy, SUF, is paid to a wide variety of beneficiaries in need. Because the circumstances for payment vary greatly, it is difficult to determine the appropriate "universe" of potential recipients. " 5Packard (2002) fmds that the rate of participation, as measured by "contribution density", is hgher among those who entered the labor force after 1981 than among those who were already working prior to that date. Participation in the retirement securitysystemis taken up in greater detail in Section111.3. In May 2002 the Lagos Administration announced the Chile Solidario initiative, a package of legislative reforms to enhance Chile's poverty reduction and social protection policies and institutions. The reforms were targeted at the 225 thousand most vulnerable households.The Chile Solidario package included both new legislation and changes in long-standing laws that dramatically altered the Government's strategy for extending socialprotection and confronting indigence. The starting premise of Chile Solidario is that the family is the principal asset of the poorest households. Thus, preferential access to Chile's social protection system is given to the neediest 225 thousand families rather than to the 850 thousand individuals that were classified as indigent in 2000. Furthermore, Chile Solidario aims to be proactive in supplying assistance and protection. Social workers go door to door, and invite indigent families to participate in a two-year program of intensive social support. This proactive supplyof social services is much more characteristicof the approachadoptedby OECD countries. Chile Solidario builds on the Puente ("Bridge") program created by FOSIS in 2002, and whch had expanded to over 218 thousand households by January 2005. Puente acts as the entry point to the social protection system for the 225 thousand households targeted by Chile Solidario. Indigent households are approached by Puente case workers and invited to participate in the program. The Puente case worker helps the household to identify its particular needs by first reviewing whether the family meets 53 minimum conditions that may be grouped into seven potential "problem areas": health; education; employment; housing; income; family life;a and identifi~ation.~The household and the case worker then draw up a strategy for how these needs will be met, and how progress will be monitored over the two year durationof the program. Participants gain preferential access to social transfers and programs. Ths is to say that in the area of health, the households are given priority access to primary health care under FONASA. In the area of employment heads of households are priority beneficiaries of active employment programs. In the area of education, schools attended by members of targeted households receive special retention voucher^.^ Further, participating households receive the newly created Bono de Proteccidn Familiar (BPF), a conditional cash transfer intended as an additional incentive to stay in the Puente program.d Households that graduate from the two-year Puente program receive a Bono de Egreso for 36 months after graduation, and are guaranteedall benefits to which they continueto be entitled. Thanks to Veronica Silva (FOSIS) and Jaime Crispi (DIPRES) for explainingthe objectives and design of Chile Solidario. Notes: a. "Family life" here refers to problems which can range from substanceabuseto physical abuse. b. "Identification" here refers to the routine process of documentationthat applies to most citizens of middle-to-high income countries. Documentation establishes the legal basis by which citizens have access to the social protection system. The "identification" component of the Puente program ensures that all members of targeted households are properlyregistered and documented as citizens. c. MIDEPLAN has been charged with overseeing and coordinating Chile Solidario. In addition to identifying the first households participating in the Puente program, the ministry coordinates efforts with the other line ministries (Labor, Health,Education,Housing, etc.)that deliverpayments and servicesto targeted households. d. The new conditional BPF is set at CLP$10,500 (approximately US$15) per month for the first six months of the Puente program. It decreasesto CLP$8,000in the second six months of the program; then to CLP$5,500;and then, for the last six months to an amount equivalentto the SUF. 5.23 This section closes with a preliminary assessment of Chile S~lidario."~ Beginning in June 2004, the first families who entered Chile Solidario through FOSIS' 116 The Government of Chile has commissioned several studies to examine the initial results of the Chile Solidario program. In addition, other qualitative and quantitativework is in process or planned. The impact Puente program in 2002 began to graduate from the first phase of the program. During this two-year phase a specially trained social worker works closely with beneficiary households to identify their special needs. They also show household members how best to meet those needs by drawing upon the public programs that are available to them by virtue of their participationin Chile Solidario. 5.24 In terms of targeting and participation rates, nearly a quarter (23 percent) of the participants of Chile Solidario are from the poorest 5 percent of the income distribution and 40 percent are from the poorest 10 percent. As of end-2004, participation rates in Chile Solidario ranged from 10percent of the poorest households in the poorest 5 percent of the income distribution to only 1 percent at the mid-point of the income distribution, and none for the richest fifth of the population.117Also as of end-2004, the following characteristicswere typical of many participatinghouseholds: they were likely to be large with a high proportion of young children and possible with a disabled household member; the head of household may be female or someone with a low level of education; they were frequently located in large urban areas; the quality of their housing was poor and they had few other assets. 5.25 In the initial stages of the Puente/Chile Solidario programs, the following seven minimum conditions were identified as program targets of the "highest priority": (i) that all familymembers have Government identification cards; (ii) that women over 35 in the household undergo regular tests for cervical cancer; (iii) that all adults in the family be able to read and write; (iv) that families know their rights and the social programs available to them; (v) that the family have adequate shelter; (vi) that at least one adult member of the household be employed with a stable source of income; and (vii) that household incomebe above the indigenceline. Table 5.2 shows the status of participating households in meeting these "priority" minimum conditions together with selectedothers. analysis would be based on panel surveys from a sample of families participating in Chile Solidario together with a sample of non-participantsselected via propensity score matching techques as the control group. The baseline was drawn at the same time as the 2003 CASEN, the first panel survey was carried out in November 2004, and others are planned for November 2005 and 2006. 'I7In the early implementation stages of Chile Solidario, some municipalities believed that the number of families below the cut-off threshold was inaccurate due to larger margins of error in the municipal calculations. As implementationhas progressed this concern has been addressed.Municipalities are able to increase the number of participants as long as they can demonstrate that they have been able to serve the needs of the first group of participants.As of mid-2005, the number of families to be invited to participate has increased by 4 percent at the national level, with substantially higher increases in some regions. It is likely that the target coverage will increase further during the course of 2005. Table 5.5: Progress Achieved by ChileSolidario Problem Area Minimum Conditions (priority condition in bold) Achieved Unachieved (i) Identification Family members have ID 25.9 74.1 Family members registered 48.9 51.1 Disabled members certified (ii) Health Regular tests for cervical cancer Registered at nearest primary care facility 46.7 53.3 Participating in rehabilitation from a disability 22.1 77.9 (iii) Education All adults can read and write 25.6 74.4 Benefit from school retention voucher 37.4 62.6 Disabled chldren in school 22.7 77.3 (iv) Family Dynamics Knowledge of rights and available services 41.8 58.2 Counseling or other remedy for physical abuse 24.8 75.2 (v) Housing House does not leak nor flood 17.1 82.9 Homestead free of polluting/toxic materials 35.1 64.9 (vi) Employnlent At least one adult member employed 25.9 74.1 If working, any children under 15continue schooling 52.7 47.3 (vii) Income Household income greater than indigence line 15 85 Household has a family budget 30.5 69.5 Notes: The first mininiuniconditionin each of the seven problem areas is that which was found to be apriority (i.e.where need was greatest) amonghouseholdsin the first wave. Performance (i.e.areas of greatest and least achievement)is measured relative to otherminima in same problem area. Source: MIDEPLAN (2005). 5.26 From June 2004 to the end of January 2005, almost 42 thousand families had completed the first phase of the Chile Solidario initiative, and of these 28 thousand families (roughly 67 percent) had met all of the 53 targeted minimum criteria. After two years in the program, the Government interviews participants in order to assess whether or not their compliance with the 53 criteria can be sustained. Figure 5.1 presents rates of progress in sustaining the 53 criteria, as well as rates of "relapse" by monthly cohort of "graduating" families (see MIDEPLAN, 2005). 118 Even after two years of participation, a family is not considered "graduated" until it has responded to a completion questionnaire, the "Fichu Final". In January 2005, 41,439 families had completed the two year phase, but only 25,219 of them had completed the questionnaire. This smaller group is the universe of "graduates" referred to in the indicators cited below. Figure 5.1: Rate of SustainedAchievement of Minimum Conditions, and Relapse 53 mnim sustained -0- Relapse June July August Septerrber October Noverrber Decerrber Month in which graduating family was interviewed I I Source: MIDEPLAN (2005). 5.27 The first cohort of families that graduated from Puente had the lowest rates of sustained progress (and consequently, highest rates of relapse). More than 50 percent of the families graduating from the first phase of Chile Solidario in June and July 2004 successfi~llyachieved the 53 criteria, but were unable to maintain this achievement by the time of their closing interview. However, as shown in Figure 5.1, the percentage of families who relapsed fell from 64 percent in June to 45 percent by December 2004. That may reflect the Government's own progress in program implementation through "learning by doing". Indeed, in January 2005, 53 percent of graduating families had successfully maintained their progress. The three criteria that proved most difficult for participating households to meet andlor sustain were (i) keeping household income above the indigence line; (ii) adequate housing; and (iii) having at least one working-age adult in regular employment. 5.28 These preliminary data should be interpreted with caution. The evaluation methodology is crude, and may over-state outcomes. The Government has devoted substantial resources to a more rigorous, in-depth evaluation of the program that will soon be ready. This said, while the progress made by the initial gro~~pof households that have completed the first phase of the program is promising, there are mounting concerns about the sustainabilityof results. THEPRINCIPALRISKSTO HOUSEHOLDINCOME AND CHILE'SSOCIALPROTECTIONPOLICIES 5.29 The World Bank (2004~)has presented quantitative and qualitative evidence showing which risks most frequently resulted in a loss of household income. Shocks to income from uiiemployment, and the cost of sickness and disability were cited most frequently. Among respondents approaching the age of retirement, the prospect of being unable to count on regular income from work, and the heightened probabilityof suffering costly sicknesses, were cited as major concerns. In light of this evidence, this report focused primarily on the losses associated with job loss, the cost of health treatment, and the risks to welfare that arise with ageing. This section updates the analysis of three principal risks and the Government instruments that are in place to help households cope with them. Job Loss, Employment and the Labor Market 5.30 The labor market is particularly important for social protection since this is where most households come into contact with Government policies and programs to help them manage risks.'19 From the household's perspective, labor is among the most important assets held by most families, particularly for the poor, and employment earnings are typically the largest source of income. Thus, the income losses fromjob loss or extended periods of unemployment are serious for most households and can even be catastrophic, particularly if employment is keeping a household abovethe poverty line. 5.31 If a labor market is relatively free of distortions and operating efficiently, turnover can be very high but extended periods of unemployment are usually rare. Workers are more likely to rely on savings to mitigate the losses from frequent turnover, but they to try and pool risks to protect against the relatively rare but larger losses from extended periods of unemployment. However, the risk of becoming unemployed is generally not considered "insurable" by private agents since it can be highly systemic-- that is, when unemployment strikes, say in a recession, a large number of individuals are affected. Since there are typically not enough "winners" to compensate "losers" from the shock, it becomes too expensive for private insurers to cover losses. For this reason governments step in to provide or mandate unemployment insurance: from risk pooling by the firm in the form of severance programs, to pooling across the entire working population in pay-as-you-go (PAYGO) systems of uneniploymeiit insurance, and even systems based on individual savings accounts backed by guaranteed minimum benefits from pooled funds. Many governments offer non-contributory unemployment assistance benefits and employment creation programs that, although not explicitly designed as an "insurance instrument" for workers in the informal sector, in effect become the public risk-poolingprogram that most succeeds in reaching workers without a contractwho lose employment or the self employed whose businesses fail in a downturn (De Ferranti, et al, 2000, Galasso and Ravallion, 2003, Reinecke, 2005). 119Of course, social protection is closely linked to factor market efficiency and human capital formation. Labor regulations- employment protection legislation,wage bargaining systems, labor taxes, and publicly mandated and provided risk-management structures, llke health, pensions and unemployment insurance - affect work incentives and can also affect investments in human capital. 5.32 From the perspective of social protection policies, the critical issue is whether the risk of unemployment and extended ~~nemploymentin particular is high, and whether social protection policies-including minimum wage and job security provisions-are exacerbating the risk or hindering the effectiveness of other instruments of social protection. Until recently, labor market research in Chile was inconclusive. Most analytical work that focused on the high-growth years of the 1990s suggested that the labor market in Chile was relatively efficient and free of barriers to employment (Gill and Montenegro, 2002). However, in recent years studies by Cowan et al. (2004), and Cowan, Micco, and Pages (2004.) find that Chile's labor market is very rigid in terms of wage adjustments and provisions forjob security, and there is increasing concern that the market is becoming less efficient (see Bergoeing, Morande and Piguillem, 2003). These concerns have been compo~~ndedby the experience of the 1999 recession. At that time, unemployment rose to levels not observed since the mid 1980s and it has been slow to fall in the years of growth since. Since 1998 minimum wages have increased by about 20 percent in real terms, affecting particularly small and medium enterprises, which are the major employers in the private sector. Throughout Latin America and the OECD social protection policies-like the minimum wage and social insurance contributions-raise the cost of labor. Figure 5.2 compares the cost of labor in Chile using the indicator most often cited in OECD countries. Figure 5.2: Minimum Cost of Labor in Chile and OECD Countries,2002" I I I I * The cost of labor indicator shows the legal minimum wage and the corresponding social securitylpay- roll contributions paid by employers as a percentage of the average wage. The data are taken from 2002, the last year for which comparable statistics have been published. Source: OECD (2004c and 2004d). 5.33 The cost of labor in Chile, 43 percent by this indicator, was about two percentage points higher than the OECD average of 41 percent, on par with the cost of labor in the United Kingdom, slightly higher than that in New Zealand, and almost 20 percentage points higher than in South Korea. In Figure 5.3 the labor cost indicator for Chile is updated to December 2004 at which time labor cost stood at almost 44 percent. Comparative research using alternative indicators confirm that the minimum wage in Chile is relatively high and binding (Maloney and Nunez, 2004).120However, despite the rise in labor cost, contracted employment increased slightly from 2000 to 2003. Figure 5.3: Changesin the Minimum Cost of Labor in Chile, 1993-2004 I I Source: The author,using methodology described in OECD (2004) and data from Chile's CentralBank. 5.34 The World Bank (2004~)presented regression analysis that sought to identify the impact of minimum wage and job security legislation on unemployment, the duration of job search, and employment, adjusted for changes in the business cycle.121The estimates have now been updated to 2004 taking account of both changes in the minimum wage and the macroeconomic climate. As in the original analysis, GDP growth strongly increases employment and significantlyreduces both the incidence of unemployment and its average duration. However, the estimated impact of GDP growth on unemployment in the updated regressions has increased dramatically in both size and significance. Afier taking into account the strong effects of the cycle and economic growth, increases in the minimum wage still significantlyincrease unemployment and its duration. However, the effect in the updated regressions is considerably weaker, most likely reflecting the slower 120Maloney and Nunez, (2004), argue for measures that are more sensitive to the distribution of earnings, such as the minimum wage as percentage of the median; as a percentage of the average earned in each income decile; as well as kernal densityplots. 12'The details of these estimatescan be found in Montenegro (2002). increase in the minimum wage since 2002. The effects of changes injob security enjoyed by workers with an employment contract are difficult to determine.122 5.35 The updated analysis suggests that the incidence and duration of unemployment in Chile is due in part to the policy framework. Increases in the minimum wages significantly raise not only the risk of unemployment but the risk of losses from protracted periods ofjob1e~sness.l~~The much weaker significanceof the minimum wage more recently may reflect its slower rate of increase since 2002 (Figure 5.3). In 2001 there was a drop in the duration of unemployment but it has held steady since then and it slightlyincreasedin 2004 despite a lower level of overall unemployment (Figure 5.4). Figure 5.4:Unemployment and Duration of Job Search in Chile, 1964-2004 - I Unemployment Rate (June, University of Chile) ++Unemployment Durafon I 1 I Source: Montenegro using data from the University o f Chile labor survey. 122 On one hand, higher job security can be viewed by firms as a tax on dismissals, reducing turnover of workers and lowering the incidence of unemployment. On the other hand higher severancepayments lower the cost of searching, increasing the duration of unemployment, and raise the unemployment rate. Finally, hgher expected firing costs may deter employers from huing new workers, lowering the supply of jobs in the formal sector and increasing unemployment and informal employment. Given that these two effects have opposite expected impacts on the unemployment rate, the combined impact on overall unemployment of increasesinjob securityfor formal-sector workers is ambiguous. 123This analysis is consistent with Cowan, et a1(2004) who found changes in the minimum wage, and wage rigidity overall, explained a significantportion of unemployment in Chile from 1998 to 2002, particularly in the constructionsector. Further, Cowan, et a1(2004) find that a portion of labor market rigidity in Chile can be explained by a relatively flat wage curve, which may result fromjob security regulations that bias job losses towards younger workers. 5.36 In the 1990s the cost of unemployment protection in Chile was very low compared with other countries in Latin America and the OECD (World Bank, 2004~). Benefits were low and they were paid only to a very small proportion of the unemployed. Chile has been gradually replacing the previous non-contributory unemployment benefit-the subsidiopor cesantia that was financed out of general revenues and paid to all who can present proof of job loss-with a contributory unemployment insurance system financed through a pay-roll tax. The new system combines aspects of savings and market-type risk pooling. Employers and workers in formal employment contribute to privately-managed individual savings accounts-similar to Chile's retirement savings accounts-and, if participant workers exhaust the balance in their accounts, their benefits will be topped-up by the Government from pooled reso~rces.'~~The new system has grown rapidly, from a mere 12.3 percent of eligible workers in December 2002, to 88.3 percent in December 2004.'~~ 5.37 There are five features of the new unemployment insurance system that are particularly attractive. First, the "hybrid" model that combines individual savings with risk pooling, is better able to serve address the needs of workers who change employment frequently as well as the long-term unemployed. Second, the new system provides better levels of compensation than the out-going non-contributory unemployment benefit. Third, under the new system, benefits are indexed to protect their value from inflation and stabilize replacement rates at their starting 1 e ~ e l s . lFourth, the present systemhas a ~ ~ sound financialbasis. Several design featuresprotect funding: benefits are limited to five months at a time; a worker may draw upon the pooled fund only twice every five years; and the overall payout rate from the pooled fund is capped at to one fifth of its current balance in any given month. Finally, to the extent that the new system allows Chile's antiquated severance regulations-an inefficient and costly form of risk pooling-to be phased out, the business environment and the employment prospects of younger job seekerswill be improved. 5.38 However, as pointed out in World Bank (2004~)~the new system also has disadvantages. First, the five month limit on benefits is short, especially if the average duration of unemployment continues to rise. Reducing the limit from one year under the outgoing non-contributory subsidio to five months will reduce the proportion of recipients among the unemployed. Many workers who actively seek work during their five months of entitlement will not be able to find employment. Second, the new system does nothing to extend unemployment benefits to a greater proportion of the employed 124The new system came into effect in October 2002, and applies to all new employment contracts. Workers with existing contracts can choose to participate in the new system, and must join upon negotiating a new contract. The non-contributory subsidio poi- cesantia will be phased out, and the government funds that currently finance subsidio will be diverted to finance the pooled component of the new system. 125Only salaried employed workers in the private sector who hold a labor contract are "eligible", as employees in the public sector have separate arrangements. 126Between 1975 and 1985 and again between 1985 and 2001 the replacement rate of unemployment benefits declined dramatically in the unstable nlacroeconomic climate. From 1975to 1985payments were tied to the minimum wage which did not keep pace with overall wage increases. From 1985 to 2002 payments were adjusted only infrequently, again much less than wage growth. The current system will tie benefits to the CPI as stipulated in the law. labor force. By shifting to pay-roll contributions as the main source of financing, a sharper distinction is drawn between the protection enjoyed by workers with a legal contract, and those without, including the self employed.Nor is the new system likely to be an incentive for greater "formalization". The new accounts are not replacing a PAYGO financed system, as old age savings accounts did in 1981, and they add to non- wage labor costs. However, it is too soon to draw any definitive conclusions about the new system. But, now that has been in operation for three full years, the new unemployment insurance system should soonbe subjectedto more rigorous evaluation. 5.39 Chile has a wide array of public employment programs to assist the unemployed. They are financed through a line-item in the national budget that is triggered when the unemployment rate (calculated by the National Statistics Institute, INE)climbs above a certain thresh01d.l~~These "active" labor programs include "direct" and "indirect" schemes to assist households searching for work.12' Direct employment programs were deployed in response to the sudden increase in unemployment during the deep recession of the late 1990s. However, the Government has chosen to gradually shift resources out of these programs and into indirect forms of employment assistance, specifically time- bound subsidies for employment creation in the private sector (SENCE), skills-renewal, and financial assistanceto would be small-scale entrepreneurs (FOSIS).'~~ 127An unemployment contingency fund in the annual budget law is triggered if unemployment passes 10 percent or when the national rate of unemployment is equal to 9 percent and is likely to grow in a three monthperiod. 128 These are two different modalities. In the case of "active" programs the government assists the unemployed to find jobs, usually through training, job-search shlls and counseling, and even job placement. "Passive" programs, on the other hand, usually take the form of unemployment insurance and non-contributory cash transfers. "Direct" employment programs are those in which the government hires the unemployed directly, usually in public infrastructure works. "Indirect" programs are those in which the government subsidizes the creation of employment in the private sector, usually by paying part of the wages. The direct employment programs that are currently being phased out are Proempleo, run by the Sub- secretariatof Labor; the emergency employment program runby CONAF (PEE); an employment program administered by the Sub-secretariat of the Interior (FOSAC); and the program for urban improvement, administered by the Sub-secretariat of Regional Development (PMU). The indirect programs that have received greater resources since 2001 are the labor re-insertionprogramrun by FOSIS (the PRLE, which is essentially, but not exclusively, financial and techmcal assistance to micro-enterprises); and the employment creationsubsidy administered by SENCE. Table 5.6: Public EmploymentProgramsin Chile PEE PRLE Proempleo FOSAC PMU SENCE CONAF FOSIS Direct or Direct Direct Direct Direct Mixed Indirect Indirect? Includes No Yes No No Yes Yes, training? through one-off payment Amount CLP$139,00 CLP$68,50 CLP$131,00 CLP$127,38 CLP$270,00 Payment to of cash 0 per month 0 per month 0 per month 4 per month 0 one off employer of payment for full for half for full for full payment for 40% of the workday workday workday workday capital and/or minimum inputs wage Other CLP$lO,OOO CLP$30,000 CLP$50,00 benefits for work depositedin 0 one off or materials personal payment for payments savings training account Time- 2 to 6 months 3 months 5 months 4 months limit, if any Targets Yes Yes Yes Yes Yes, but Yes househol priority given d heads? to Puente HHs Targets No No Yes Yes Yes, and No the poor? priority given to Puente HHs Source: Own, based on DIPRES (2005). 5.40 Since the Bank's last report, the Government sponsored a comprehensive, independent evaluation of its public employment programs. The evaluation found that, in 2003, all these programs together only covered 9.7 percent of the unemployed and 30.3 percent of ~memployedheads of households. The evaluation shows that the direct programs seem to do a betterjob of targeting the poor and indigent. For lack of data it is not possible to assess how far the Government's preferred SENCE subsidy favors poor or indigent groups. Figure 5.5: Coverage and Targetingof Public Employment Programsin Chile I 50 10 I Roempleo PEECONAF FOSAC PMU PRLEFOSlS SOVCE Coverage (% of unenployed) m Coverageof HHHeads (% of unemployedHHheads) o Targeting to Poor (% of beneficiaries who are poor) Targeting to Indigent(% of beneficiaries who are indigent) m Average period of beneficiary participation(mnths reportedby beneficiary) 'ource: DIPRES (2005). 5.41 The evaluation indicates that the PEPs have had a positive impact on employment. Individuals who had participated in a direct employment program were between 8 percent and 12 percent more likely to have a job than those who had not participated. Graduates of FOSIS' PRLE were between 11 percent and 14 percent more likely to be employed, and those who had benefited from the SENCE subsidy were 5 percent more likely to be working than those who had not benefited from any program.130, 131 5.42 The programs seem to have increased participants' skills to an extent that varies from one program to another.Participantsin directprograms were 13percent more likely to receive skills training than the unemployed who did not participate in any programs. Beneficiariesof the SENCE employment subsidies and .theFOSIS PRLE were 35 percent 130 The University of Chle, which conducted the evaluation, counts participation in the program as "employment". Thus, although it is clear that beneficiaries would have been unemployed had they not been in the employment programs, it is impossible to assess from the reported results whether the programs had any lasting impact on the employability of beneficiaries-that is, an impact on their future prospects in the labor market. 13'In the case of FOSIS PRLE program, the selectionof the control group may bias the results. Using other beneficiaries of the PEPs or other unemployed as the control group does not take into account the initiative, entrepreneurial instincts, good business plan, etc that may be characterize the self employed. and 51 percent more likely respectively to receive training than non-participating unemployed. These two programs also received the highest praise from beneficiaries for assistingthem to compete and find work ("employability"). 5.43 In an attempt to increase social protection, the Government has required that individuals benefiting from employment programs be hired with a contract (although FOSIS' program aimed primarily at potential self employed is exempt from this requirement). This policy has made "formal" contracted enlployment an objective of the programs. Prior to the new policy in 2001, participants in the direct programs were up to 12 percent more likely to find formal employment. After the policy was introduced, the likelihood that participants in direct PEPs found formal employment increased to 38 percent. Some 20 percent of the beneficiaries of the SENCE subsidy found formal employment. 5.44 But over the medium term, the likelihood of PEP participants eniployed with a contract falls somewhat. Beneficiaries of direct PEPs have a 23 percent greater likelihood of being formally employed. However, neither the participants in FOSIS' PRLE nor the beneficiaries of the SENCE's subsidy are more likely to find formal employment in the medium term. The Impoverishing Cost of Health Care and Financial Protection in Chile 5.45 From a household perspective, there is a clear connection between health and social protection policy. Household spending on treatment for sickness or disability can be impoverishingand may also have a long-termimpact on human capital formation.The cost of care can perpetuate poverty for those who are already poor. And this risk does not necessarily diminish with economic development. In the United States up to half of the people who declared personal bankruptcy in 2001 did so because of health costs (Himmelsteinet al., 2005). 5.46 For these reasons, it is not only important to maintain and improve the health of the population, but also--perhaps equally important fiom a household perspective-to ensure that households are protected financially in the process (Baeza and Packard, 2005). As yet, there is no consensus among health specialists as to what constitutes "financial protection" or what level of spending on healthcare is "excessive". However, there is at least preliminary agreement that households spending on health should be considered excessive or "catastrophic" if the level of spending plunges households into poverty or hinders their capacity to build and sustain human capital. Chile has recently taken significant strides ensure greater financialprotection fiom sicknessor disability. 5.47 A critical indicator is the proportion of health expenditure that is paid out of pocket by households (as opposed to the share that households pay for private health insurance or in contributions to public health financing arrangements). Higher out-of- pocket spending can indicate greater vulnerability to the financial consequences of ill health. Households in Latin America pay on average 85 percent of private health spending out of pocket. In E~~ropeand the OECD the average amount is only 72 percent. Among the wealthier countries in the OECD, the average falls to 65 percent. By this measure, Chilean households may be relatively less vulnerable, paying only 60 percent out of pocket for health care. 5.48 However, the demand for, and spending on, health care rises with income-as people who become better off seek better quality care, including elective and costly procedures. Figure 5.6 shows the relationship between out-of-pocket spending and level of per-capita income. Out of pocket spending in most Latin American countries is high given their level of per-capita GDP. Notable exceptions are Uruguay and Colombia. Chile's out of pocket spending is roughly in line with what would be expected, given it's per capital income, but is still slightlyabove the international norm. Figure 5.6: Out-of-PocketHealth Expenditure as a Percentage of Total Health Expenditures I . DOY. REP. $0 $5.000 $10.000 $15,000 $20,000 $25,000 $30,000 $35,000 $40,000 GDP per caplta (PPP) I I Source; WHO (2004)and UNDP (2004) 5.49 Is out of pocket spending on health impoverishing in Chile? A high portion of out of pocket health spendingdoes not in itself reflect lower welfare or greater vulnerability. The World Bank (2004~) cited household survey data indicating that, after unemployment, treatment for sickness or disability was the second most frequently reported negative shock to income. Raczynski et al. (2002) present qualitative evidence showing that health costs increase the financial vulnerability of households. Bitran et al. (2002) find that the number of people falling into poverty from health care costs rose between1998 and2000. 5.50 Expanding on this analysis, Bitran et al. (2004) estimate the percentage of individuals who are thrown into poverty by the cost of health care. Figure 5.7 shows the percentage of new poor (households that fall below the poverty line who were not previously poor) in each income quintile as a result of out of pocket health expenditures.132In the absence of a consensus in health policy circles on the level of spending that should be considered "catastrophic", the poverty line may be adopted as a threshold. Figure 5.7: Percentageof Households PallingBelow the Poverty Line Due to the Cost of Health Care, by Household Income Quintile I 1 Ql (pwrest) (12 Q3 (14 Q5 (richest) Tdal Source: Bitran, 2004 5.51 Nationally, about 1 percent of households are newly poor as a result of health spending. The vulnerability to impoverishing health expenditure is greatest among households in the third income quintile. About 5 percent of households in this income group are made poor by the cost of health care. Although not shown in Figure 5.7, the further impoverishing impact of health spending for households in the first and second quintile is also substantial.However, the actual poverty effect of health costs are likely to be much larger because the estimates in Figure 5.7 show only the cost of care and do not capture the loss of income due to lower productivity and missed work days. The available data are insufficientto estimatethe combined effect of both (Baeza and Packard, 2005). 5.52 The experienceof Chile with social insurance and a National Health Servicedates back to the early 2othcentury. Coverage of health risks was partially privatized in the early 1980s to give households a greater range of choices in health coverage. All Chileans were able to choose between a publicly run social health insurance system (FONASA) and a privately provided health insurance (ISAPREs), irrespective of their labor status or sector of employment. This dual system has created significant problenis '32TOestimate the number of households thrown into poverty from health expenditures, the authors based their calculationon household disposableincome net of health expendituresin the three months prior to the survey. of segmentationand equity. Coverage under FONASA differs for each income group and can be either fully or partially subsidized for the poor and low income workers. Those who do not qualify for subsidization make co-payments to the public health service according to their income. The premiums for health coverage to both the public and private systems are set at 7 percent of workers7 earnings. However, for the private system, the 7 percent premium is a minimum and the price of coverage can vary greatly between private insurers. That leads to distortions in co-payments and deductibles, and it also leaves a substantial share of participants "under-covered" against costly health care. 5.53 The question of how best to protect households from the financial consequences of sickness and disability has been the subject of lengthy debate in Latin America. The nub of this discussion is a trade-off between the breadth of coverage (how many households can count on some health protection) and the depth of coverage (the package of care they can count on). The constitution of most Latin American countries guarantees the right of citizens to good health and access to health services. In theory, the public health system covers treatment of all health conditions. However, in practice, health systems are financiallyconstrained, and resort to rationing or lowering the quality of care in order to comply with their budget constraints. At least until the mid 1990s, constitutional mandates existed alongside a large unsatisfied demand for healthcare and were never accompanied by effective instruments, especially top serve the poor. Starting in Colombia in 1994, several countries have introduced legislation and sector reforms to transform health benefit packages into explicit entitlements guaranteed to all citizens.'33 These reforms typically include legal mechanisms for households to demand their entitlements from the State. 5.54 Chile has gradually been introducing a new health insurance package (Plan of Universal Access with Explicit Guarantees-AUGE) since 2002. Rather than continuing to offer unrealistic promises to cover all health needs, AUGE establishes a guaranteed minimum basic package of health cover for all Chileans, along with guarantees of attention within specified time periods, set ceilings on co-payments and full subsidies for the poorest households. The new system will eliminate rationing and improve service. AUGE levels out premiums and co-payments and reduces "cream skimming". It is an enforceable patient's "bill of rights" for FONASA affiliates. It reduces waiting lines and other inefficiencies that can increase out-of-pocket costs particularly for poorer households. AUGE was first tried out on a pilot basis with the population covered by FONASA. The pilot experience yielded positive results by increasing access to the services included in the package. Since January 1, 2004, AUGE has been mandatory for all citizens and all public and private providers of health ins~~rers.Households covered by FONASA and the ISAPREs are entitledto the same minimum package of health services. 5.55 It is too early to assess the effect of the Chilean reform on lowering vulnerability to poverty or easing the burden of health expenses for the poor. This said, the "explicit '33Other examples of explicit entitlement reforms include Colombia's mandatory benefit package (Plan Obligatorio de Salud) introduced in 1994, the PMO (PlanMedico Obligatorio) and the Universal Maternal and Child Benefit Package (PlanNacer) of Argentina, and the new Seguro Popular in Mexico. entitlements approach" is already having both intended and unexpected effects in Chile and elsewhere in the region. First, it has forced policy makers to revisit the breath and depth discussion of health coverage. Before the reforms, all health services were theoretically availableto all citizens,but in practicenothing was actually guaranteed. The result was a limited and poor qualitypackage of services,particularly for the poor. Under the new explicit entitlement approach benefit packages become legally binding for governments, throwing open the debate of whether to guarantee a limited package to all or an extensive package (often focused on the most costly health care) to a few. Secoiid, these reforms have lead to a niuch closer dialogue between ministries of finance and ministries of health because the legally binding nature of the packages significantly reduces the space for fiscal adjustment in the health sector. Errors in defining the coverage of the package can have significant fiscal consequencesthat may undermine the sustainability of the reform. Third, although unintended, most reforms have provided governments with a powerful instrument to focus on efficiency enhancing reforms. Guaranteeing a package requires clarity and certainty regarding the quality of delivery, which in turn requires complex monitoring systems, provider payment systems and contracts or quasi-contracts between the public financing agency and health service providers. The Risks to Income for the Elderly and the Multi-PillarPension Systemof Chile 5.56 Old age is frequently accompanied by a loss of earnings and a reduction in income. At the same time the elderly become increasingly vulnerable to sickness and disability which may result in burdensome health care costs. The combined effect of those two factorsmay result in poverty. 5.57 In principle, the financialrisks associated with aging are insurable. All else equal, poverty in old age should be increasingly rare and, therefore, also a risk that can efficiently be pooled. But given the nature of the risk, "p~verty'~,markets cannot be expected to respond with risk-pooling instruments, and the State often steps in to fill this gap with social insurance. 5.58 To help households manage the risks to welfare that arise from ageing, the Government of Chile has an array of instruments, both direct and through the regulation of private provision. Structural reforms in 1981 introduced a shift away fi-om a single- pillar, "defined benefit" public pension system to a "multi-pillar" system in which the private financial sector plays a leading role. These instruments are generally very well matched to efficiently and effectivelycover prospective losses. The Governmentrequires individuals to save for the increasingly predictable loss of earnings ability, in savings accounts managed by the private AFP industry. It has developed strict guidelines for "programmed withdrawals" and it enforces minimum annuity requirements through privately provided insurance to cover the risk of individuals outliving their retirement savings. It guarantees a minimum level of retirement income for individuals who participate in the savings plan. From general budget resources, it provides a modest benefit targeted to the elderly indigent who were too poor or otherwise unable to participate in the system. In addition, the Government offers tax incentives to encourage voluntary retirement savings, and it ensures that the contractual savings instruments provided by the financial sector is reasonably safe and well regulated. 5.59 In the widely accepted pensions lexicon, Chile's multi-pillar pension system has two "first pillar" instruments, the minimum pension guarantee (MPG) and the targeted social assistancepension (PASIS) that pool the risk of poverty. It regulates the provision of obligatory "second pillar" savings (AFP accounts) and annuities to ensure that individuals engage in a minimum level of saving; and it encourages additional,voluntary "third pillar" savings through regulated private instrunients ("ahorro provisional voluntario", or APV). 5.60 Figure 5.8 compares the replacement rate that can be expected from Chile's second pillar plan, with those that can be expected from mandatory retirement income systems in selected countries in the OECD."~The simulated rate of replacement shown here is what a male worker who earns an average wage and has a "full" record of contributions(40 years) can expect from the mandatory, earnings-relatedpension system in each country. 5.61 At first glance the replacement rate that can be expected for the average worker who fully complies with the mandate to participate in Chile, while lower than the OECD average, is not dramatically so, and is in fact higher than the average replacement rate earned from mandatory systems in the United States and the United Kingdom. Concerns have been expressed with respect to the fiscal sustainability of mandatory pension systems in France, Germany, Spain and other OECD countries with an aging population. Chile's relative ranking at the lower end of the scale reflects a difficult but necessary choice to re-define the social contract with respect to old age income - a decision that was taken in the late 1970sthat culminatedwith the 1981reform. 5.62 The data in Figure 5.8 should be treated with caution because Chile's second pillar is fundamentally different from the mandatory pension systems of most OECD co~ntries.'~~Chile's second pillar is based on a defined contribution, individual savings,that places the burden of responsibility squarely on the individual. In contrast, most mandatory retirement systems in the OECD are defined-benefit systems and the State shoulders most of the risk. Further, Chile's system of individual accounts can be invested on the capital markets and they have earned a rate of return that has averaged 10-1 1percent since the system's inception. By way of contrast,benefits financed from a public plan are typically tied to economic growth or to the growth in wages because they are financed from taxes.'36 134The net replacement rate is defined as pension income net of taxes as a percentage of net pre-retirement inconie. The personal income tax systems in OECD countries play an important role in determining total old-age income. 135However, of the countries included in the figure, the second-pillar systems in Australia and Mexico employ a financinginstrument similarto the Chileanpension system. 136The replacement rates presented in the figure were calculated using uniform assumptions including a 2 percent real rate of return. This is indeed low for a system of individual accounts in the long term. Even a conservative 4 percent real rate of return would yield replacement rates above 65 percent. However, the 5.63 Notwithstandingthe different system structures, it is noteworthy that replacement rates from Chile's AFP system are comparable with those from the earnings-related component of pension systems in OECD countries. Hence, what Chile's mandatory retirement security system is designed to do from the pensioners perspective, is broadly in line with what mandatory systems in the OECD are designed to do (from the same perspective). Figure 5.8: Average ReplacementRates from Mandatory Earnings-Related Pension Plans. Chile and SelectedOECD Countries Retirementpension net of taxes as a percentage(%) of pre-retirementsalary net of taxes (Maleworker earningaveragewage, with full contributionhistory of 40 years) Greece Austria Italy Spain Netherlands Portugal Finland Germany OECDAverage France Canada Denmark Chile Australia UnitedStates United Kingdom Korea Mexico 0 10 20 30 40 50 60 70 80 90 100 Source: OECD (2005) and Whitehaus (forthcoming). 5.64 With respect to poverty prevention, Figure 5.9 presents another set of replacement rates, again comparing Chile's mandatory system with those in the OECD. However, this second set of replacement rates compares the minimum pension that a complying worker can expect from the AFP system, with the minimum pensions that complying workers in OECD countries can expect from the mandatory retirement security instrument. This indicator is used to measure the "social adequacyyyof pension benefits (OECD, 2005). figures presented based on very conservative assumptions in the context o f the comparison with PAYGO defined benefit systems inthe OECD. Chile's minimum pension guarantee, currently equal to 28 percent of the average wage, is only a percentage point less than the OECD average. Figure 5.9: Minimum Pensions Guaranteed by the Mandatory Pension System, Chile and Selected OECD Countries Luxemburg Portugal Greece NewZealand Netherlands UK Spa~n Ireland France Korea Canada OECD Average Ch~le Poland Germany Australia Italy F~nland USA Mexico Japan Czech Republ~c -- 0 5 10 15 20 25 30 35 40 45 50 Minimrn pension as percentage (%) of average wage I I Source: OECD (2005) and SAFP web page. 5.65 To conclude, Chile's first and second pillars are generally and favorably comparable to mandatory plans in the OECD. However, policy makers in most OECD countries are increasingly concerned at the prospect of financially unsustainable retirement security systems. In contrast, the pension deficit in Chile is largely a result of the transition-cost of the 1981 reform. It currently averages 5.6 percent of GDP and is gradually declining. 5.66 Despite this positive assessment, the pillars of Chile's old age income security system are once again a hot topic of policy debate. Although policy makers on both sides of this debate are less concerned about the financial sustainability of the pension system, there is increasing concern for its credibility and legitimacy as an effective instrument for smoothing consumption over a lifetime and for preventing poverty. Although the system is designed to yield benefits that meet even OECD standards of protection at a more sustainable fiscal cost, in Chile (as in other Latin American countries) the fill1 participation required for households to benefit from this protection cannot be taken for granted. 5.67 In the past fifteen years, the share of the labor force that contributesto the pension system has rarely been higher than 60 percent (see Arenas and Marcel, 1999, Arenas, 2000, Packard, 2002a, Valdes, 2002, Beyer and Valdes, 2004, and Rofinan and Carranza, 2005). While this participation rate is high by regional standards, surpassed only by Uruguay, it is well below participationrates in the OECD. The international comparisons presented in Figure 5.8 are based on a full lifetime's contribution(4.0years) to the system by participating workers. However, for most Chilean workers, pension contributions are interrupted and irregular so that, upon reaching retirement age, they will not have contributed sufficientlyto merit a full pension. 5.68 In order to receive the minimum pension guaranteed by the State, workers in Chile have to contributefor at least 20 years. Using data from a 1999household surveyin Greater Santiago (PRIEso),'~~Packard (2002a) estimates that as many as 30 percent of contributing men and 40 percent of contributing women do not have the contribution density (months of contribution divided by months of remunerated work) required to qualify for the guaranteed minimum pension. These estimates were recently confirmed with 2002 data from a nation-wide survey (EPS'~~,Quintanilla, 2003), which asked respondents about their contributions during each period of employment/unemployment. When survey data is cross-checked against the record of actual contributions, simulations suggest that up to 50 percent of affiliateswill not be eligible for the guaranteed minimum pension (Bernstein,Larrain and Pino, 2004). 5.69 Many explanations have been suggested for the worryingly low rates of worker participation. They include individual and household preferences139,broader economic factors such as labor market condition^,'^^ the structure of incentives built-in to the first and second pillar as well as the incentives offered by other components of Chile's social protection system.141Although research is far from conclusive, it has also been argued that low rates of participation could be linked to perceived high system costs resulting from insufficient competition between private fund managers (Gill, Packard and Yermo, 2004). 5.70 The Government and the academic community are increasingly critical of the private fund management industry, wary of its degree of concentration and super-normal profits, and skeptical as to whether it is keeping the promise of privatization, namely 13'Encuesta de Previsidn de Riesgos Sociales, Santiago, Chile, December 1999, University of Chile with financing from the World Bank 13'Encuesta de Proteccidn Social 2002, originally, Encuesta de Historias Laboral y Seguridad Social, University of Chile,under the directionof the Sub-secretaria de Previsibn, Ministry of Labor. 13'Including a preference by workers for current over future consumption(Torche and Wagner, 1997,Barr and Packard, 2001); aversion to financial risk (Barr and Packard, 2001); preferences for alternative instruments of savings and insurance (Packard, 2002); as well as credit constraints and liquiditypreference (Beyer and Valdes, 2004). 140Including the business cycle and economic growth (Valdes, 2002, and Packard, 2001); unemployment (Bernstein,et al, 2004); minimum wage legislationand provisions forjob security(Valdes, 2002). 14'Including the availability of non-contributory benefits, the rate of re$m on contributions required for the guaranteedminimum pension (Bernstein,et al, 2004); and health insurance coverage (Valdes, 2002). greater efficiency and cost-savings through market competition.142Although the industry's costs have declined and are becoming comparable to the cost of similar financial services in OECD countries, there is considerable scope for further reductions. The problems of low system coverage, lack of competition and high costs are not only detrimental to the system and its effectiveness, they also threaten the system's credibility. To the extent that the system loses credibility,the important gains from structural reforms in the 1980scould come ~~nderpolitical attack (Valdes, 2004). 5.71 The most recent changes to the pension system in Chile have been to increase investment options in the second pillar, and to increase savings and insurance in the voluntary third pillar. Chile has moved towards making the mandated AFP system more flexible and allowing workers greater choice over their investment portfolios.'43 Government efforts to increase the appeal of the mandated second pillar have been complemented by steps to strengthen the voluntary third pillar and provide greater tax incentives for individuals to participate in it.144Preliminary research indicates that household response has been positive, with a modest increase in participation rates (Medrano, 2004). 5.72 The Government is considering further changes to the structure of the second pillar, with three specific objectives in mind: (i) to increase the rate of participation, primarily by further improving incentives; (ii) to lower administrative costs, mainly by exploiting recognized economies of scale in account management (Marinovic and Valdes, 2004); and (iii) to ensure that cost savings are passed onto workers in the form of lower commissions, by increasing conlpetition in the industry. Among the specific measures being considered by the Government is an "opt-in" default option on income tax returns that wo~~ldlower the transactions costs of participation for the self employed;'45It is also considering measures to consolidate and centralize contributions 142The number of fund managers 5the AFP industry fell from 21 in 1994to 6 in 2003. The fund managers have enjoyed persistentlyhigh profits with annualreturn on equity averaging 30% in recent years, twice the rate earned by banks. These returns on equity have been achieved despite low levels of risk (IMF-World Bank,2004). 143From May 2000, Chilean workers close to retirement were offered the opportunityto swap from the so- called Fondo l (the original pension fund available since 1981) into a fund invested exclusively in fixed income securities(Fondo 2). A law passed early in 2002 extended individual choice over investment even further.A "nlulti-fund" system was approved which permits workers to choose between five funds (A, B, C, D, and E), with varying exposure to equities.Men older than 55 and women over 50 can choose between the four funds with the lowest risk. Pensioners will only be able to choose between the three funds with the lowest risk. Workers who do not select a specific fund are assigned one according to their age: the young are assigned a fund with a higher investment in equities while older workers are assigned a fund with a higher investment in fixed income securities. 144 Since March 2002, Chilean workers have been able to save up to 50 UF of their monthly pre-tax income in any government authorized voluntary pension plan, in additionto any AFP. There are no restrictions on the number of plans or AFPs in which workers may deposit their tax-advantaged, voluntary savings. Workers may also cash out their contribution to these plans at any time before retirement subject to a 10 percent special excise tax (in additionto the applicableincome tax). Employer contributions,however, may still onlybe liquidatedat retirement. 145 The self-employed and employers may choose whether or not to participate in the AFP system. It has been proposed that the self employment be required to participate in the AFP system. However, based on the negative experience in other countries llke Argentina and Brazil, the Government is rightly concerned that such a move could actually increase evasion, and it has chosen instead to improve incentives to and other account management functions,which would allow the AFPs to competeon the basis of their investment services, and would encourage new entrants into the industry.'46 Another option is to introduceregular reverse auctions in which affiliates could chooseto be included in a large group that would be "sold" to the AFP that offers the lowest commissions. "Undecided" affiliates could be allotted to the cheapest fund manager.'47 Other measures being considered would involve a gradual shift in the regulation and supervision of second-pillar investment away from a compliance based approach and toward one that emphasizes effective risk management.'48 Yet another promising proposal-though currently not among those being considered by the Government- would increase incentives to participate by granting workers access to a part of their mandatory retirement savings.14' 5.73 The Govenunent's ideas for changes to the second pillar are very promising, and recent moves to strengthen the third pillar by providing workers a greater degree of choice and incentives are also extremely positive. However, 19 percent of men and 32 percent of women are not even affiliated with the pension system (Beyer and Valdes, participate. The "opt-in'' default is supported by the experience of OECD countrieswhich shows that, when participation in voluntary company pension plans (like the 401k in the US) is made the default option, worker enrolment doubles. Furthermore, by introducing the possibility of a single annual contribution, rather than twelve monthly contributions, the transaction costs of participation for the self employed are reduced almost to zero. Given that the number of permanently self employed is relatively small, and the vast majority are only temporarily self-employed,the opt-in default could prevent them from missing-out on their contributions. 146There are no explicitbarriers of entry to the AFP industry, however, access to the market is not easy due to very high start-up and sunk costs that can put-off potential new entrants.Much of these costs have to do with account management rather than investmentmanagement. Both areas of the AFP business are subject to economies of scale but there is a relatively greater potential for product differentiation in investment management. In OECD countries, Sweden has centralized account management, while allowing fund managers to compete in investment management. The Government is currently contemplating measures allowing AFPs to outsource some of their administrative functions as well as a more radical centralization of the functions. 14' Research shows that the potential savings to affiliates from thls sort of reverse auction would be great, and could lead to greater competition between the fund managers (Valdes,2004). 148 This change would accord well with the "prudent person" financial regulation preferred in the US and the UK. The current compositionof AFP portfoliospartly reflect overly complex and restrictiveinvestment regulation but the scope for liberalizing regulation is constrained by the inherent fiduciary role of the private second pillar. Unlike in other countries in the region where the benefits of private defined- contribution plans are sometimes complemented by public defined-benefit plans, the AFP system is the core and dominant component of old-age income support in Chile. Thus workers are already exposed to a relatively greater degree of market risk. Furthermore, Chile does not yet have the legal, techmcal and regulatory capacity to fully adopt "prudent person" rules. For these reason any liberalization should proceed gradually and cautiously,as suggested in IMF-World Bank (2004). 149Under this proposal, described in Valdes and Beyer, (2004), the affiliated worker would be allowed to "borrow" from their individual account up to a certain amount. The interest on thls "loan" would be set at market rates, and paid back into the worker's account. The affiliate would not be allowed a second "loan" until the first was paid back. Experience in the OECD is mixed. Singapore grants workers access to their mandatory retirement savings for specific investments, such as housing. A large number of affiliates in Singapore exhaust their account balance before reachmg retirement. In the US, individuals can draw on their 401k and IRAs, although taxation thereon dissuade most from doing so. The dilemma for policy makers is to balance individuals' liquidity preference with their long-term consumption smoothing objective. 2004). The tax incentives for voluntary retirement savings, which have been found to be regressive in several countries,150are likely to have only a modest impact on savings given how few Chileans pay income taxes. Thus the iniprovements to household welfare from reforms to the second pillar and from giving greater weight to the third pillar could be suppleniented by consolidating and strengthening the set of instruments intended to prevent poverty in old ageboth the non-contributory PASIS, and the contributory guaranteed minimum pension. 5.74 In a country where all workers contribute to the earnings-related pension system, the current structure of the guaranteed minimum pension is relatively good: it encourages workers to save privately and guarantees a minimum level of retirement income at a minimum cost to tax payers. However, in countries like Chile where many workers will not have a sufficientlylong record of contributionsto the pension system, topping-up on the basis of participation can exclude large segments of the population and also lead to perverse transfers. 5.75 Chile could move toward consolidating and simplifying its poverty prevention pensions (the PASIS and the MPG) into a single (perhaps prorated) public risk pooling device. Ideally a single minimum benefit would be financed from general taxation, made available at a retirement age that is periodically adjusted to reflect changes in overall life- expectancy, targeted to the elderly poor and indexed to prices. However, consideration should be given to abolishing the rationing system which currently leaves some 19 thousand women and 11 thousand men without pensions although they qualify for the PAS IS.'^^ Once a non-contributory old age benefit is in place, a contributory minimum guaranteed pension or matching contribution scheme targeted a lower-income groups would cease to be a first-pillardevice and would instead act as an incentive to participate in the second-pillar. Covering Job-loss 5.76 The World Bank (2004~)has argued that the new PEPs are not well designed to serve as the self-targeting safety net advocated in the best-practice literature (see Ravallion, 2000, De Ferranti, et al, 2000, Galasso and Ravallion, 2003). Shortcomings in the design of PEPs include (i) eniploymeiit that pays the legal minimum wage and requires contributions to the pension/social security system; and (ii) access to all labor programs (whether these are direct employment at the municipal level, or programs of 150Since relatively well off individuals are those most likely to have the information and discretionary income to save for the longer-term, preferential tax treatment for voluntary retirement savings can be regressive. However, various schemes such as matching government contributions for potential savers from poorer groups can be put in place to counter that effect. 151Valdes and Beyer (2004) estimate that the annual cost of covering the current elderly indigent who are not receiving the PASIS, would be US$25 million. They show that the cost could be financed by cutting back benefits that are badly targetedto the non-poor elderly. subsidizing employment creation in the private sector) conditioned on beneficiaries presenting proof of unemployment. While the outcomes shown in the recent evaluationof the PEPSare promising, there is still cause for concern. 5.77 The new programs are not likely to reach the most vulnerable. First, employers are likely to use the SENSE subsidiesto hire the most employableworkers-that is, those who least need the subsidy-or to fornlalize workers already employed informally. Second, the flat structure of payments, and the generosity of the minimum wage paid by the PEPs relative to the market clearing wage, could prove attractive to a segment of the younger population with few skills. The danger is that individuals enticed into the labor force by public employment schemespaying the minimum wage, may cut their education short and impair their future earnings ability. Third, by paying the minimum wage, and offering pensions and social security coverage, the Governnlent may have created relatively attractive jobs (offering wages at above the market-clearing wage, as well as social insurance). To the extent that publicly created or subsidized employment is relatively attractive (especially to workers who otherwisewould not join the labor force), it will be politically very difficult to move workers into private sectorjobs or eliminate job-creation subsidies in periods of economic growth. If the Government cannot easily dismantle what could become costly public employment programs, it may hesitate to establishthem again in the future even if they are needed.152 5.78 When access to programs is conditioned on proof of job loss, it presents an additional barrier between informal workers and what could be the only effective, publicly-provided recourse available. Although direct employment programs administered at the municipal level do not require proof of unemployment, the Government would like to phase them out and move toward indirect employnlent promotion,job training and time-bound subsidiesto private employerswho offerjobs. In these new programs, access will be conditioned on the unemployed person presenting a 'Ifiniquito" (similar to a pink slip in the United States). Thereby, access is restricted to those who lost ajob with a legal contract in the formal sector.153 5.79 As argued in World Bank (2004~)this presents a fundamentalproblem. Workers in the informal sector, and the self-employed do not subscribe to unemployment insurance and public employment programs may be the only recourse they have in an economic down-turn (Reinecke, 2005). The PEPs, particularly FOSIS PRLE and the SENSE enlployment subsidy, seem to be succeeding in placing a portion of the unemployed in jobs and in increasing their skills. But the problem does not lie with the programs' job-placement function, but with their income-protection function. The Government'spolicy of conditioningparticipation on presentation of aJiniquito excludes them. In 2003, 11.6 percent men and 17.3 percent of women in urban areas were employed without a contract. In the same year, 20.8 percent of men and 17.4 percent of 152 Since 2000 direct employment programs have been progressively dismantled, despite substantial political opposition. All new employment program have operated through the private sector. However, indirect employment subsidies can also create a fiscalburden that the government may wish to avoid. 153Much depends on how diligent is the employment office in seeking proof of unemployment. This is likely to vary greatly with the capacity of local employment offices. women in urban areas were self employed.154The number of workers in employment without a contract and in self employment increases at the lower end of the income distribution. 5.80 The Government is right to be concerned that its public employment programs are effectively and efficiently targeted, but requiring proof of unemployment is a poor targeting device.155A more effective way of targeting programs to those who truly need them, would be to pay wages below the legal minimum, and ensure that the work is relatively undesirable. Offering pension coverage (by requiring contribution to the pension system) also makes the jobs created under these programs relatively attractive and difficult to phase out. However, since pension coverage also offers protection against more immediate and potentially catastrophic losses from disability and sudden death, subsidizingprotection against these lossesmay be justified. 5.81 Furthermore, if the minimum wage continues to rise as quickly as it has in recent years, there could be fiscal consequences. Offering above-market wages imposes three separate economic costs. It attracts more workers to public employment programs, pays each of them more than they would otherwise accept, and could crowd out private employment. As Chile's Government has discovered, if public employment programs offer (or subsidize) above-market wages, the fiscal cost can increase unless access is rationed. But this places the Government in the undesirable position of working through quantityrather than through price. 5.82 Indeed, since 2001 the number of hours and work days under the direct employment programs has been significantlyreduced, and the number of beneficiaries of the indirect programs has to be strictly limited to contain costs. In effect, with regard to public employment programs, the Government has forgone un-rationed, self-targeting forms of income protection, in favor of relatively expensive forms of employment assistance at above-market wage cost that is strictly rationed and thus of limited coverage. Should it again be faced with a deep recession, the Government may have to re-consider this choice. 5.83 Notwithstanding from the income-protection function of the PEPS, the best guarantee of social protection is employment creating growth that generates a stable income. For most people, even the best designed social protection prograni is a second- best solution compared to a job. Ensuring that labor markets facilitate employment creation is an important aspect of social protection policy. Notwithstanding a binding minimum wage and labor costs that by some measures are higher than the OECD average, unemployment has fallen in Chile and the proportion of workers employed without a contract also fell from 2000 to 2003 (although it remains substantial). '54Employment without a contract and self employment is even hgher in rural areas. In 2003 19.9percent of rural men and 23 percent of rural women worked without a legal employment contact. Just over 32.5 percent of men and 24.9 percent of women were self employed in rural areas. L55However, requiring proof of unemployment should make it harder for young people leaving school to seek the relativelyhgh subsidized minimum wage in the formal sector. 5.84 Despite higher overall employment, the preliminary assessment of Chile Solidario shows that even families who have graduated from the first phase of the program still find hard to find stable employment. As shown in Table 5.2, despite three years of economic growth and a fall in the overall rate of unemployment, unemployment among men from the poorest urban households did not change, and among men from the poorest households in rural areas unemployment actually increased by 7.3 percent. The concern here is that a binding minimum wage may be limiting the employment opport~mitiesof the poorest-the very groups that Chile Solidario seeks to reintegrate back into society. Reforms could be considered to slow the rise in the minimum wage or to loosen its hold on employment at the lower end of the earnings distribution. However, fbrther empirical work will be required to reach definitive conclusions. 5.85 In the meantime, the Government faces the challenge of finding more effective ways of deploying employment assistance. Public employment programs could be targeted more clearly on low-skilled groups (first and foremost, the graduates of Chile Solidario). Much can be learned from the pro-active approach to training and job-search assistance of the United States and the United Kingdom (Box 5.3). Part of this pro-active approach would involve re-structuring the local unemployment offices (0MILs)-the Government's "front line" in dealing with unemployment at the local level-to more effectively match job-seekers with the needs of local private businesses. Experience in the US and the UK shows how local employment offices can play a critical role in overcoming the labor market's failure to absorb unskilled labor from poorer households. Indeed, the importance of the OMIL was confirmed in the recent evaluation of the PEPS. They have a much greater potential role than they are currently playing in assisting the poorest to find sustainableemployment. Box 5.3:Shifting the Focus of EmploymentAssistance: Recent Trends in OECD Beginning in the 1990's, the United States and several European countries, including the United Kingdom and the Netherlands, placed greater emphasis on work requirements and support for work in the context of social assistance. The United Kingdom completely integrated the administration of social assistancebenefits and employnlentservices.Fortunately,the impact of these changeshas been well evaluated. Some of the major lessons may be relevant for Chile. A strong economy is importantfor the successof work-oriented reforms. Welfare-to-work programs can effectively reduce the need for costly social assistance and they can also increasethe supply of labor. Positive and negative incentives are more effective when introduced jointly rather than separately. Some people respond best to job training and others to work experience. Evidence shows that "mixed" programs, assigning some people to work-fust and others to training, are the most successfulin terms of reintegrating the unemployed into the labor market. Singlemothers have specialneeds. Finding a job isjust one piece of the picture. Most need support in other areas as well, including child care and health care. Others may need help to overcome the barriers they encounter in the workplace. Behavioralpatterns can be changed, as well as attitudesand expectationsabout participation in the labor market. The reforms have discouraged use of social assistance by those who have other options. Sources: Blank (2003); Neubourg, Castonguay, and Roelen (2005). Covering the Cost of Health Care 5.86 Chile has adopted an entitlements approach to health care with a guaranteed minimum health benefit package. That is a bold stride towards ensuring financial protection from the cost of health care in the event of sickness or disability. However, there is still concern for the financial sustainability of AUGE. A cap on co-payments introduced new fiscal costs. The Government chose to meet the costs with a modest increase in VAT and by extending mandatory contributions to the self employed who have previously been outside the health and wider social protection system. There is interest in shifting the financial base for health coverage even further to VAT, which would be more efficient and might even allow a reduction in Chile's pay-roll taxes. 5.87 However, only a small number of countries have succeeded in making such a transition. Among these Spain's experience is particularly relevant (Box 5.4) given the many similarities with Chile. The shift away from pay-roll tax contributions (a "Bismarckian" model) to broader based taxes (the "Beveridgian" model), could not be done in the short term. The shift would require a transition period during which workers currently outside the system would be encouraged to take up coverage along with careful subsidiesto prevent undue lossesto poorer households (Baeza and Packard, 2005). 5.88 Baeza and Copetta (1999) explore the potential fiscal iniplications of shifting from a payroll-tax to risk-rated premiums with Government subsidies for those who can not afford them. The authors suggest that workers be required to purchase a basic package of health insurance, like Chile's new AUGE package, with a minimum set of benefits, and clearly defined standards of quality. At the same time, mechanisms would be established to enforce compliance. For illustrative purposes, the authors show that adopting the package of services offered by FONASA in 1999 could be fiscally neutral. Indeed, they show that shifting from a payroll tax to risk-rated premiums would increase the income tax base and would more than compensate for the cost of subsidizing those who could not afford to pay risk-rated premiums. Box 5.4: :Spain's ShiftFrom Bismarck to Beveridge: De-linking Health Coveragefrom Employment Status The health financing system in Spain underwent radical changes in the 1980s and 1990s. Spain shifted from a social insurance system financed by payroll-taxes to a national health service financed by general taxation. Today, almost 100percent of public expendituresin health are financed from generaltaxation. Spain's transition to democracy and the new constitutionof 1978 gave new impetus to health care reform. A separate organization was established within the social security system for the administrationof health care services. Most health care programs and organizations were consolidated under the umbrella of the Ministry of Health which was established as an independententityin 1981. In 1986the GeneralHealth Care Act was approved following almost four years of public and parliamentary discussion. The act provided a unified legal franlework for many of the previous piecemeal reforms and called for a tax-based financing system. Publicly managed health services were consolidated in a single national organizationand in a small number of regional organizations (Cataluiia and the Basque Country), within the framework of the newly decentralized State. The Spanish National Health System was subsequently devolved to the 17 Autonomous Communities that, since 2001, have fully managed their RegionalHealth Services and together make up the SpanishNHS. Consistent with the reforms introduced by the General Health Care Act, the sources of funding for the health care system were drastically modified in 1989. Beginning in that year, new budgets were financed 70 percent fiom general taxation and only 30 percent frompayroll-taxcontributions. In the mid-1990s, consistent with agreements signed by political parties and trade unions (known as the Pactos de Toledo), 100percent of financing would come from general taxation and individualcontributions were to be progressively phased out by 2000. In 1999, one year ahead of schedule, the entire health care budget in Spain was financed from general taxation. Today, Spanish Regions receive funding for health care as part of the general funding from central Government. Funding is proportional to population adjusted for such factors as age distribution, temporary residents, and services provided to the national system or to neighboring regions. Supplementalfunding is also suppliedfrom fiscal revenues raised in the region. Currently, only workers' compensation for work-related injuries and diseases are financed fiom individual contributionsof employers and employees. Source: Jesus Maria Fernandez (2004). 5.89 P~~blicpolicy in health has two fulldamental objectives. The first is to improve health standards. The second is to afford financialprotection from the cost of sickness or disability. The relative importance attached to those two objectives varies significantly from country to country. Poorer countries, with low life expectancy and very scarce resources, are likely focus on improvinghealth standardsmuch more than middle-income countries that have already achieved high levels of life expectancy (Baeza and Packard, 2005). Faced with evidence showing that the cost of health care can be impoverishing, policy makers in Chile will have to carefully balance the health status and the financial protection objectivesof health systemsin settingtheir policy priorities. Income Security for the Elderly 5.90 The measures currently contemplatedby the Government to improve competition and lower participation costs in Chile's second pillar are very positive. Measures have already been introduced to provide greater investmentchoice with the "multi-fondos" and a wider range of instruments with the "ahorro provisional voluntario". Togetherwith the new reforms to pillars two and three they could very well lead to greater competition, lower costs and better coverage for Chile's pension system. There is an increasing consensusthat the first pillar instrumentsneed to be better coordinated and consolidated. Restrictionsof access to the PASIS could be removed at a minimal fiscal cost. 5.91 However, there is a more hndamental set of policy questions that have not yet been considered. Broad principles such as "Povertyprevention" and "minimum adequate consumption smoothing" are very subjective targets and can only provide very general objectives for policy-makers designing or reforming an old-age income security scheme. With respect to the structure of the pension system, where should the line be drawn between pillar one (collective responsibility underwritten by the State) and pillar two (minimum consumption smoothing, mandated-b~~t not guaranteed or ~~nderwritten-by the State)? 5.92 There is increasing consensus in Chile and across Latin America on the importance of increasing coverage. However, there is still no consensus with respect to the question "What do we need to cover?" In the context of "multi-pillar" systems, the answer is relatively easy for the first pillar, i.e. some level of consumptionclosely related to the poverty line.156The answer for pillar two is much more difficult, and raises still more questions. What should pillar two cover? What is the minimum level of consumption smoothing society should demand from individuals?In Chile, the difficulty is compounded because the second-pillar does not explicitly define a benefit, but rather defines a minimum amount of savingsfor individualsto set aside for the AFPs to invest. 5.93 In setting the rate of contribution andlor the ceiling on earnings subject to mandatory savings, the architects of a pension system implicitly adopt a benefit package based on plausible assumptions of salary growth and the rate of return from investment. Prior to reforms in January 2004, the requirements for "earlyretirement" were that from the age of 50 an affiliate had to have sufficient savings in their individual account to finance an annuity equal to 50 percent of their average annual salary during the previous 10 years or 110 percent of the guaranteed minimum pension, whichever was higher. These parameters were designed to ensure a significant cushion between the pension that is fully financed from the affiliate's account and the guaranteed minimum benefit. Once the January 2004 reforms are fully iniplemented, savings accounts will be sufficient to finance an annuity of 70 percent of the last 10 years average annual wages, and 150 percent of the guaranteed minimum pension, whichever is higher. 5.94 The current minimum guaranteed pension is roughly CLP$77,000 up to the age of 70 and about CLP$84,000 thereafter. The Government is seekingto increaseparticipation among the self employed and employees without a contract but they frequently under- '56For example, Chile's non-contributory PASIS is approximately equal to the poverty line, and the contributoryminimum pension guarantee is about two times the poverty line. report their earnings. The only reliable assumption is that affiliates to the system are at least earning the legal minimum wage.lS7 5.95 If this minimum contributory second-pillar benefit objective is already defined in the law that created Chile's system, are any further reforms required? That depends on whether Chile's second-pillar defined benefit really satisfies households. The fact that many workers opt for early retirement from the AFP system seems to indicate that it does. Although low by Latin American standards, the proportion of earnings subject to compulsory second-pillar contributions in Chile is much higher than the OECD average. Other authors have suggested that lowering the contribution ceiling to 1 or 2 times average earnings would be more consistent (Valdes, 2002) and might prove a more acceptableto Chileanhouseholds. 5.96 The Government confrontedthese same questions in the early 1990swhen it tried to increase participation in the pension system among household domestic workers. The Government created a special account-the CAI (Cuenta de Ahorvo de Indemnizacidn)- that responded to the "demand" for coverage estimated on the basis of a survey. Employers are required to deposit 4.11 percent of domestic workers' earnings in individual accounts which are managed by the AFPs, and which the account holder can access should they lose their employment. In the year that followed the change, participation among domestic workers increased dramatically. While not specifically geared to old age income security, the CAI helps workers to smooth their consumption, and may have even acted as the foundation for unemployment savings accounts. More importantly, the Government's experience with the CAI is instructive in posing the question "What is it we want to cover?" and adjusting the parameters to the system accordingly. 15'Valdes (2002) citesreports that up to 500,000 contributing affiliates were under-reporting their incomes and making contributionson the basis of the legal minimum wage. Figure 5.10: Earnings Subject to Mandatory Contributions in Chile and Selected OECD and Latin American Countries I I Source: OECD (2005) and SAFP web page. OPTIONS FOR POLICY IMPROVEMENTS IN CHILE 5.97 The objective of this chapter, using new data that has become available since the last Bank report on social protection in Chile, has been to assess how Government social protection policies in the last three years have helped households manage the principal risks to their welfare. It has drawn upon the findings and conclusions of empirical research conducted in the Bank and elsewhere since the last report. 5.98 There are no reasons to change the conclusions in World Bank (2004~).Chile largely succeeds in providing households with the instruments that they need to effectively mitigate shocks to their income and to protect their human capital. The institutions Chile has put in place are generally appropriately designed to match the risks they are intended to cover. However, despite some improvements from 2000 to 2003, too many Chilean households-even among the non poor-still do not have access to social protection. Vulnerability is greatest where households depend on employment without a contract or on self employment. A household head or primary earner who loses a job in the informal sector, or a self employed person whose business fails, faces explicit and implicit institutional barriers to even basic forms of social protection. And since a significant number of uncovered workers are not counted among the poor, Chile's relatively well targeted social safety net will not catch them soon enough should they suffer a fall. 5.99 Between 2000 and 2003 there was a reduction of informal salaried employment among vulnerable groups of men and women from the poorest households. However, employment without a contract rose from 15.8 percent to 16.1 percent for men from urban households in the 3'd and 4thincome deciles. Among women in the same income group, employment without a contract fell two percentage points, but was still a high 29 percent in 2003. And while a move into self employment should not be automatically interpreted as increased vulnerability (followingMaloney, 2004), the prominence of non- professional self employment across all income groups in urban and rural areas may be cause for concern. Thus the over-arching challenge for policy makers in socialprotection is how to improve coverageamong these potentiallyvulnerable groups,primarily through the eliminationof explicitbarriers to minimum levels of incomeprotection. 5.100 In many middle income countries-notably in Latin America and especially in Chile-there is no lack of formal institutions to help workers manage risks to income. However, these are only available to the shrinking minority of workers with legal, regulated forms of employment. In order to extend protection to the majority, governments can lower the costs to workers and employers of complying with labor regulations, and increase the enforcement of regulations to encourage legal employment. They can also design instruments of social protection that are financed from general taxation and introduce "self targeting" instruments that are not based on labor contracts. Chile has already done much of the former and has thereby shifted away from the typical Latin American context. In its new context Chile can go yet fbrther by first, removing explicit barriers and quotas on minimum forms of income protection as it improves its targeting and enforcement; and second, further improving incentives for households to take up more than minimum coverage. 5.101 But whatever the context, as mentioned previously, it is importantto acknowledge that the best guarantee of social protection is good macroeconomic management and employment creating growth. The best guarantee of risk management is employment that generates stable income. For most households, even the best designed social protection system is only a stop-gap support for their consumption and welfare. Ensuring that the economy is managed so as to generate adequate employment and that labor market policies facilitaterather than impede employment creation is the most important aspect of sound socialprotectionpolicy. This is a lesson that few in Chile will ever forget. 6. FINANCING DEVELOPMENT IN CHILE: ISSUES AND POLICY OPTIONS 15' 6.1 As income grows and equity concerns are heightened, demographic trends and social preferences are bound to raise the demand for public goods. How far is it desirable or acceptable to accommodate such demand pressures? Tradeoffs in efficiency and growth are inherent in the need to finance expenditures. Taxes tend to distort resource allocation and debt financing can crowd out private investment. However, public expenditure itself may be pro-growth, e.g. by fostering human capital accumulation andlor innovation. Furthermore, if increased public expenditure is associated with increased expenditure efficiency, it may lead to falling public expenditure-GDP ratios. Thus, it is difficult - on conceptual grounds alone - to ascertain whether higher public expenditure is good for growth or not. 6.2 A first objective of this chapter is to assess whether Chile is likely to face expenditure pressures. Specifically, given the trend of public expenditure in comparable fast-growth developing countries, how fast is public expenditure likely to grow in Chile? Also, given Cliile's current structure of public expenditure, which categories of expenditure are more likely to grow? The experience of OECD countries suggests that demography, equity-driven expenditure pressures and a demand for new public goods are likely to drive public expenditure growth in the development process. At least as relevant is the scope for efficiency gains in public spending. Ultimately, the quality and efficiency of public expenditures -rather than their size - may matter most for growth and equity. 6.3 Whether additional financing is required or not, there will always be the issue of how best to finance public expenditure, be it through taxes or debt. A second objective of this chapter is to assess whether changes in the tax structure can increase the efficiency, equity, or compliance/administrativecosts of the tax system. Chile has a fairly efficient tax system and tax administration. However, the growing complexity of the system tends to foster tax avoidance and increases collection costs. In addition, income inequality is a key characteristic of the Chilean economy and it is relevant to ask what role if any -the - tax system plays in determining income distribution. In particular, what is the impact of numerous deductions, exemptions and tax credits on tax progressiveness? 6.4 Over the last 15 years Chile financed higher public expenditure primarily out of tax resources. Rising tax revenues and fast economic growth account for the sharp drop in the public debt-to-GDP ratio. Lower public debt contributes to reducing Chile's vulnerability to terms of trade volatility - highly relevant in view of the dependence of budget and export revenues on copper. The application of the fiscal rule (under a positive growth scenario) would lead to further reductions of the debt-to-GDP ratio over the long term. However, whether this is an efficient outcome or not is an issue that deserves to be '58The chapter was written by D. Oks based on background papers by A. Vivanco (public expenditure), C. Agostini (taxes),R. Perotti and F. Giavazzi (fiscalrule). examined. For example, the benefit in terms of improved creditworthiness of further reductions of debt may be less than the cost of foregoing public investments with high social rates of return. Also, the eventual elimination of public debt could imply losing a useful benchmark for capital markets such as the term structureof interest rates. 6.5 A third key objective of this chapter is to assess current financing arrangements particularly as they concern the application of the current fiscal rule. The fiscal rule has held up remarkablywell and has contributed to establishingthe credibility of fiscalpolicy in Chile. Any major change to the rule after such a short period could be a blow to the credibility of economic policy, which in all probability would adversely affect its perceived creditworthiiiess. The analysis in the chapter thus is primarily focused on options to improve the implementationof the existing rule. The section concludes with a discussion of longer term options to facilitate the financing of high social return public investments while consolidating creditworthinessand ensuring the application of an anti- cyclical fiscal policy. PUBLICEXPENDITURE: STRUCTURE, EFFICIENCY AND LONG TERM SCENARIOS Public Expenditure and Growth 6.6 Fiscal prudence in Chile has been a key factor behind gradual disinflation and macroeconomic stability, which, in turn, had a positive impact on growth and poverty red~cti0n.l~~For several decades Chile has kept its fiscal accounts under strict control. Sincethe return of democracy,the central Government public debt-to-GDPratio declined fkom 45 percent in 1990 to 12 percent in 2004. In 2001 the Government adopted a structural fiscal rule by which the central Government is required to maintain a surplus equivalent to 1 percent of GDP where fiscal revenues are calculated on the basis of estimated long-term copper prices and trend GDP. The rule allowed the central Government to run small deficits during years of slow growth - reaching a maximum deficit of 1.4percent in 2002 but then helped to curb expendit~resduring the upswing - leading to a 2.2 percent fiscal surplus in 2004. Chile's prudent fiscal behavior compares favorablywith OECD countries (IMF, 2004). 6.7 Within this prudent fiscal framework, Chile has been able to double public expenditure in real terms between 1990 and 2004. Yet, Chile's Government expenditures as a share of GDP are relatively small compared to other countries at the same level of developnieiit: with a per capita income of around US$8,500 (in PPP 1995 dollars), the Government of Chile spends close to 22 percent of GDP, whereas other countries at similar levels of development spend around 27 percent of GDP. Based on cross-country econometric comparisons,Barro (1999) estimates that the low Government consumption 159There is a relatively large literature on the negative effects of inflation on growth, including Fischer (1993) and Barro (1997). The positive relationship between economic growth and poverty reduction has ample empiricalbaclung, as in Dollar and Kraay (2002). ratio in Chile accounts for 0.8 percent of the growth rate differential with a sample of other countries (De Gregorio,2004; Barro, 1999). 6.8 However, the composition of Government spending and its efficiency are potentiallymore important than size when explaining growth, poverty and inequality. An approach taken in the literature is to distinguish between "productive" and "unproductive" public spendingand to measure their different effects on growth (Landau, 1983; Aschauer, 1989). Furthermore, Government expenditures enter the aggregate production function directly in endogenous growth models that include human capital as a production factor. Hence, since public investment in education and health improves the quality of h~lmancapital, expenditures in these areas should have a positive effect on economic growth. 6.9 The supposed tradeoff between growth and equality through taxes and transfers implicitly assumes that public expenditure is inefficient. This is not supported by the evidenke, at least not in OECD countries. Lindert (2004), for example, found that welfare programs such as basic social assistance, public health and pensions tend to have pro- growth effects in OECD countries. Moreover, when social public goods are perceived to be a net benefit by households, taxes on labor may be neutral in the sense that they do not distort the labor-leisure choice of workers. In that event, public spending need not be a drag on growth. 6.10 The literature also shows that well-executed public spending in infrastructure complements the efforts of the private sector, and in some cases it might be an essential element of economic growth. The seminal paper by Aschauer (1989) found that Government capital had a significant effect on economic output in the United States. Kamps (2004), using new capital stock estimates for 22 OECD econonlies found a significant and positive elasticity of output with respect to public capital, thereby demonstrating that Government capital is generally productive. Easterly and Rebelo (1993) found that investment in communication and transportation can have a positive impact on growth. However, the empirical link has not been clearly established,due to a variety of factors, including the possibility that inefficient and costly public projects may have a negative impact on growth, as well as the discouragement of private participation by public provision. Indeed, Devarajan et al. (1996) find that for developing countries investments that are generally considered productive, such as capital, health and education, can have a negative or insignificant relationship with growth, while current expenditure has a positive correlation. They interpret this result as an indication that otherwiseproductive incentivescould become unproductive beyond a certain threshold. 6.11 In short, Chile's prudent fiscal policy has created a stable macroeconomic environment conducive to growth. In spite of fast rising public expenditures, the expenditure-GDP ratio is below the average for countries with a comparable per capita income. To that extent, it is unlikely to be a factor retarding growth. In any case, there is no simple link between public expenditure and growth. Much will depend on the efficiency and composition of public expenditure - productive vs. unproductive expenditures -and not just on the overallmagnitude. These issues are examinedbelow. Public Expenditure -Structure and Efficiency The Changing Structure of Chile's Public Expenditure 6.12 Chile has undergone a process of rapid growth in public social expenditure since 1990. Expenditure in health, education, justice, and other social sectors increased as a proportion of total public expenditures. The share of education rose from around 11.1 percent of central Government expenditures in 1990 to 17.8 percent in 2003. Over the same period the share of health increased from 9.1 percent to 14percent. Likewise, as a share of GDP, educationrose from 2.3 percent to 3.9 percent and health from 1.9percent to 3 percent (Figure 6.1). Figure 6.1: Share of Expendituresin GDP (percent) ~ ! ! ~ ~ 8 i ~ 4 8 ~ ~ 8 / 8 IGeneralPublicSenices IEconomicActivities Defense IPoliceandJustice Health IEducation Social Protection Source: Ministerio de Hacienda, DIPRES. 6.13 The combination of higher social spending and almost continuous fiscal surpluses was feasible partly due to tax revenue performance and partly due to declining shares in expenditure of defense, debt service and social security. The share of expenditures for national defense decreased fiom around 10percent in the early 1990sto 7 percent over 2000-03. Debt service declined from 9 percent to 2.2 percent. These trends are depicted in Figure 6.2, which plots selected real expenditures between 1990 and 2003. Figure 6.2: Real Expenditures: 1990 -2003 (2002uesos) 4 . . . . . . . . . . . . . 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 20W 2001 2002 ZW? +&blicDebtTransactwns -+Defense -Heakh -6Iucaton Source: Ministerio de Hacienda, DIPRES. Chile Ranks High by International Standards in Terms of Social Spending 6.14 Chile currently ranks among the countries with the largest share of social expenditures relative to total expenditures - 70 percent compared to an average of 45 percent in all countries for which data was available. This places Chile at the same level as Uruguay, Switzerland and New Zealand. At about 15 percent of GDP, social expenditures in Chile are above the value predicted from a cross-country comparison based on income (Figure 6.3). "Social expenditures" include social security expenditures, which are low in Chile because the system was privatized in the early 1990s and the Government only pays for those individuals who choose not to participate in the new system (De Gregorio, 2004). Hence, the data in Figure 6.3 probably underestimates the degree to which Chile is an upward .outlier in international comparisons of social expenditures other than social security expenditures. Figure 6.3: Total Expenditures and Social Expendituresfor Selected Countries Total GovernmentExpendituresto GDP Social Expendituresto Total Expenditures --c*.--. -- -- .- "-. ---..-.- "A + I - 5.000 10,000 15.000 20.000 25.000 30.000 5.000 10,000 15,000 20,000 25.000 30.000 GDP per Capna(PPP. 1995 US$) GDP per Capna (PPP. 1995 US$) Source: World Bank World Development Indicators. 6.15 To assess the impact of higher expenditures on growth it is convenient to focus on expendit~~resthat improve human capital, namely health and education. In terms of health expenditures,the public sector in Chile ranks low compared to other countries,but it ranks much better when private expenditure is included. As is shown in Figure 6.4, Chile is above the predicted level of health expenditurebased on income per capita. The same pattern can be seen in terms of education expenditure:public expenditureas a share of GDP is low compared to other countries, but once private expenditure is included, Chile is one of the highest spenders in education (among countries for which data is compiled by the OECD). Figure 6.4: Expendituresin Health and Education for Selected Countries EducationalExpendituresand GDP HealthExpendituresand GDP 0 01 5,000 10,000 15,000 ZO.000 25,OW 30.000 35,000 GDP percapita. PPP GDP oer caoita. PPP Source: World Bank World Development Indicators,World Health Organization and OECD. 6.16 Notwithstanding the need for improvements in targeting and in the quality of social services, the expansion of social expenditures has been associated with a period of fast economic growth and poverty alleviation. Several studies have concluded that much of the decline in poverty that took place in the early 1990s can be at least partially attributed to the implementation of social policies starting in 1990 (Foxley, 2004 and Meller, 1999). Meller claims that around 60 percent of the reduction of poverty can be traced to higher economicgrowth, while the rest is due to socialpolicies. Chile has Improved the Targeting of Social Expenditures, but there is Scopefor Further Zmprovement 6.17 Health, education and cash subsidies provided by the Govemment are targeted towards the lowest income quintiles, as can be seen in the figures below. There has been an improvement in targeting in tem~sof the share of subsidiesthat go towards the lowest three quintiles. The inclusion of these transfers in measures of income inequality shows that they had a positive effect in diminishinginequality (World Bank, 1997). 6.18 However, progress has been uneven. The targeting of educational subsidies has not improved much since 1990 and a large proportion of those resources go to the wealthiest two quintiles. In 2003, only 81.2percent of education subsidies were received by the poorest 60 percent of the population, compared with almost 90 percent for cash transfers, and nearly 100percent for health transfers (Figure 6.5). Hence, an improvement in the targeting of educational subsidies, which currently do not differentiatebetween the income level of recipients, would help to diminish income inequality.Also, better access to finance for tertiary education will have a significant effect on inequality - private returns to university education are particularlyhigh. Figure 6.5: Targetingof Social Expenditures Share of Public Subsidies Received, by Income Quintile, 1990,2000and 2003 I I 1 1 Subridiesl MonetarySubsidies Health Subsidies EduMional 420 -" "--- Source:Ministerio de Hacienda, 2004. ThereIs Also Scopefor Reducing Inefficiencies in Health and Justice Expenditures 6.19 In the 1990s, Chile increased public expenditure in health 191 percent in per capita real terms. As shown by Rodriguez & Tokman (2000), increased spending was directed towards physical and human capital development in the sector. However, indicators of efficiency, especiallywhen compared with total wages in the health sector, show a negative trend. Like in other countries (e.g. Uruguay), private health insurers tend to focus on high-income, low-risk individuals,leaving the public sector with low-income, high-risk individuals. Thus, rather than increasing public health expenditures, Chile's health authorities could consider options to improve the system design, e.g. the competition framework for private and public providers; enforce greater accountability on specific subsidies, e.g. labor disabilities; and in general create incentives to improve efficiencyin public sector units, e.g. hospitals (Aguirre, Albagli & Rappoport,2004). 6.20 Similarly,while Chile has already undergone importantjudicial reforms, there is still room for greater efficiency in the judicial system. For example, the overloaded system (the case load is 278 percent the number of cases resolved, almost twice the levels of Colombia and Peru) could be relieved through the creation of specialized courts to deal with bankruptcy,tax or regulatoryissues (Buscagliaand Dakolias, 1999). Research and Development (R&D) is One Sector Where ThereIs Scopefor Greater Private Sector Involvement -Primarily ThroughPublic-Private Partnerships 6.21 Chile spent on average 0.56 percent of GDP on R&D between 1999 and 2002. Although this share has increased recently, it is still well below the 1 percent average of other countries. Indeed, most of the countries that have been mentioned as models for Chile's development,such as Finland, Korea and Ireland, have more than twice the R&D expenditurebudget than Chile. Furthermore, Chile's expenditure in this area is below the expected value based on its level of GDP (Figure 6.8). There is empirical evidence of a positive link between R&D and economic growth through the domestic development or adoption and adaptation of technologies from abroad. However, the empirical evidence is much weaker when it comes to assessing whether public expenditures on R&D complementor crowd out private R&D. 6.22 R&D funding and expendituresin Chile are overwhelminglyconcentrated in the public sector, in sharp contrast with the composition of R&D expenditures in OECD countries (Figure 6.6). In Chile more than 70 percent of the funding for innovation comes from the public sector. That compares with only 40 percent for the majority of OECD members, where the bulk of funding is provided by business enterprises. The distribution of expenditures is also very different in Chile than in OECD countries. In countries such as Finland and Korea, the majority of the public spending on R&D takes place through business enterprises, while in Chile it is academic institutions that spend the majority of the funds available. In part as a result of the low level of spending, Chile lagsbehind in innovation indicators. Figure 6.6: Expenditures in Research and Development in selected Countries Researchand development expenditure (% of GDP) 1888- I 1 R&D ~xpendituresand GDP per capita 0 5.000 10,000 15.000 20,000 25,000 30.000 35.000 GDP Der caDlta (PPP. 1995 US$. 2000) Source: World Bank World Development Indicators 6.23 One way to increase the funding going into R&D is to develop public-private partnerships (PPP) whereby public efforts can be combined with those of the private sector. Indeed, PPP in R&D have been used in several countries, following their success in the infrastructure sector. These arrangements increase efficiency by complementing the interests of public and private actors, and by filling in the gaps in innovation technology and addressing new social needs. Examples include the creation of grant schemes with matching funding from the private sector, the commercialization of publicly funded R&D, and the development of human resources. The OECD has carried out research on good practices for R&D PPPs in Australia, Austria, France and the Netherlands. Public expenditure Long-term Scenarios - Public Expenditure will come under Pressurefrom Demographic Factors, Greater Demandsfor Equity and New Public Goods 6.24 Total Government expenditures as a share of GDP have grown substantially during the last four decades in all OECD countries. This has been ascribed to diverse factors, including the following: (i) an increase in the efficiency of tax collection; (ii) new demands for the provision of Government services; (iii) demographic shifts accompanied by a gain in the political power of the groups affected; (iv) changes in the relative political power of individuals affected (taxpayers and recipients of transfers); and (v) incentives for politicians and interests groups to increase the resources under their control (Meltzer and Scott, 1981). In general, the literature on public choice highlights the critical inlportaince of voter's preferences, institutional and electoral arrangements, and actions by elected officials, voters and interest groups in determining the size and growth of the State. However, it has proven difficult to distinguish between these competing theories and quantify their effects. 6.25 As Chile continues on its path of economic development, new expenditure pressures are likely to emerge from: (i) demographic factors, (ii) social development and equity needs and (iii) demand for new public goods as income rises. 6.26 Demorzraphics. In terms of demographic changes, the share of the population over 65 years old is projected to increase from 7.8 percent at present to 11 percent by 2020, while the share of those over 60 is forecast to escalate from 11.2 percent to 16 percent during the same period (Figure 6.7). Figure 6.7: Age Distribution of Populationin Chile (percent) Source: Statistical forecasts by INE. 6.27 An aging population will increase the public costs of health provision in Chile. First, the elderly have a higher probability of being ill and of seeking more expensive medical attention. The use they make of medical facilities is three times greater than for younger individuals. Currently the majority of older people belong to the public insurance system (FONASA), and hence the burden of the demographic shift will fall on the Government. Second, the pricing system of the public and private insurance provides incentives for individuals at high-risk to shift from the private to the public provision of health as they become older The public sector charges for services on the basis of a flat fee, while private insmances price their services according to the risk profile of the insured (gender, age and family members). Hence once the risk of the individual crosses a certain threshold, the incentive is to move to FONASA (Sapelli, 2003). The structure of incentives, however, is being modified with the implementation of Plan Auge (see chapter 5). 6.28 Demographic factors are also likely to exert pressure on other public expenditures. In particular, the share of the population which is expected to require supplementarypublic pensions -in those cases where contributionsto the funded scheme do not reach the State-guaranteed minimum pension - is likely to rise over time as the age profile of the population increases. Minimum pensions cover old age, disability and widowhood; according to the Ministry of Finance, contingent State liabilities stemming from insufficient contributionsto cover the minimum pension guaranteed by the State are expected to quadruple over the next 15 years to over US$300 million by 2020 ("Informe de Finanzas Pziblicas", 2004). 6.29 Social development needs. The quest for equity and - especially - equality of opportunity in Chile is linked to: social demands which were not sufficiently satisfied under the dictatorship; Chile's unequal income distribution; the growing consensus that public expenditures are the main vehicle for redistributing income; and a growing awarenessthat equalityof opportunity is a key responsibility of the State as well as a key factor underlying sustained growth. Hence, Chile's population is likely to demand increased coverage of servicesprovided by the Government. 6.30 In particular, as primary and secondary education coverage reaches significant levels, society will demand a higher quality of instruction, which is particularly expensive, as it involves teachers' training and other complementary services. Also, the share of population attending tertiary education will increase, thereby increasing the resource requirements of the universities and grants for students of low-income households. Finally, greater political participation of women in the labor force and in the political process will require an expansion of the current system of childcare. Also, as the Government acknowledges, greater spending in pre-school education is desirable. This can level the playing field for students of diverse socio-economic backgrounds at an early stage - in primary school, where performance may already be excessively dependent upon family backgrounds (impairing opportunitiesfor students from low socio-economic households). 6.31 Demandfor new public goods. In addition, as the income of the representative voter reaches a level comparable to that of more advanced economies, and the hndamental needs of the population are satisfied, it is likely that society will demand new public goods appropriateto a higher standard of living. Areas of public expenditure that have an elasticity with respect to income greater than one include citizen's rights, environmental protection, urban development, transport, pollution, justice and security. Public expendituresin these areas are likely to increase in Chile as well. 6.32 It is difficult to know which functional types of Government expenditures are most likely to rise over time. Data constraints substantially limit this analysis but some lessons may be drawn from developed countries. As Figure 6.8 shows, in EU and OECD countries the trend has been for public consumption, investment and subsidies to decline as percentage of GDP. Only social transfers show - over the last 25 years- an upward trend. Ireland is an exception as the public investment-GDP ratio also rose significantly over that period. Figure 6.8: InternationalCom~arisons-Distribution of Public Ex~enditure T&l spending* 35 - - 35 W , " , , . . """'..,'""' 30 2 1W5 1 1- 2XO 2.D&a arebasedon SMS3fESM. 2 GNPisused clforIhe ratioskr Idand. 3. WWiM awerage (I= 5WamWPkfofawsilabkcountries. 4. %%a1 spending is M da~larreniautlays @usnetqihid j y s Development Financing Needs of Chile -An International Perspective 6.33 The economic literature has studied the absolute as well as relative increase in public expenditures that accompany growth. Wagner's Law has been tested empirically by applying different methodologies ranging from simple regressions to causality testing and more recently to cointegration analysis (for example, Oxley, 1994; Ansari et al., 1997; and Chlestsos and Kollias, 1997). The results vary dramaticallybetween countries and over time. As Peacock and Scott (2000) explain, the increasingly complex empirical analyses of Wagner's Law miss the point of the simple relationship that Wagner formulated as an empirical fact based on observationduring in the 1 9 century, instead of ~ ~ an articulated model. 6.34 Hence, rather than an empirical analysis based on Chilean time-series, the attempt here is to examine the experience of countries that have experienced rapid growth. The objective is to evaluate public expenditure performance and draw possible implications for Chile. As mentioned, as countries reach different levels of development, the needs and demands of the population change accordingly. Hence, it is important to examine countries that started their growth process at roughly the same level of development as Chile now has. According to the World Development Indicators, in 2003 Chile had a GDP per capita of US$9,706 (in dollars of 2000). Assuming that the economy grows at between 5 percent and 6 percent per annum, and given population growth of around 1 percent per year, the GDP per capita of Chile could be around US$20,000 in.2020. 6.35 Since 1970 five countries have experienced an increase in income from around US$10,000 to more than US$ 20,000 in PPP terms over periods of 20 years (Figure 6.9). The fastest growing economy was Hong Kong, where the per capita income went from US$9,043 in 1979 to US$23,863 twenty years later. Singapore followed a very similar path, while Ireland lagged in development in the first couple of years, but eventually reached the same level by the end of the period. Portugal experienced relatively lower growth, reaching a GDP per capita of around US$17,000. Figure 6.9: GDP growth in selected comparatorcountries Source: World Bank's WDI database. 6.36 These countries'60 also experienced a remarkable expansion of total public expenditures during two decades of high economic growth (Figure 6.10). Indeed, Cyprus, Ireland, and Portugal saw their real expenditures more than double during the period (Cyprus in half the time), while Singapore saw its expenditures grow fourfold. As Figure 6.11 illustrates, there is a diversity of experiences in terms of the evolution of the expenditure-GDP ratio. While both Cyprus and Portugal show an upward trend 160Note that comparable data for Hong Kong was not available. throughout the period, in the cases of Ireland and Singapore an initial upward trend is later reversed and, in the case of Ireland, the final ratio is below the initial one. 6.37 It is important to note that these increases in public expenditure took place even though the ratio to GDP at the beginning of the period was higher than that of Chile, with the exception of Singapore. Indeed, the Government of Ireland spent around 38 percent of GDP in 1978, while Cyprus and Portugal spent around 30 percent at the beginning of their 20 year expansion. Their experiences suggest that the Chilean Government may expect higher expendit~~resin the long run and would benefit from exploring financing options accordingly. Figure 6.10: Total Public Expenditures of Selected Countries(indexat time t=100) Source: World Bank's WDI database. Figure 6.11: ExpenditureIGDPratios Selectedcountries - Source: World Bank's WDI database. FINANCING DEVELOPMENT -OPTIONS TO IMPROVE THE EFFICIENCY AND EQUITY OF THE TAX SYSTEM 6.38 Given Chile's development needs, should expenditure be financed by taxes or debt? The tradeoffs between tax and debt financing are complex to assess. Taxes may be distorting. Income taxes may discourage labor. Taxes on saving may discourage saving. Taxes that discriminate between sectors or regions may result in resource misallocation. Tax policy can also affect equity and economic efficiency - through collection/compliance costs, tax evasion or avoidance. On the other hand, debt financing - to the extent that it is perceived as future discounted taxes may also discourage work, - saving or investment or can simply crowd out private investment through higher interest rates. In a country vulnerable to terms of trade volatility like Chile, debt management can also play a critical role in helping to smooth out adverse shocks and, particularly when creditworthiness is a concern, it can prevent an adverse shock fiom turning into a financial or solvencycrisis. 6.39 Rather than answer the question of whether taxes or debt should be used to finance development, this section steps back and addresses the issue of how to improve tax policy so as to reduce the potential burden of taxation. Specifically, it evaluates options to reduce incentives for tax avoidance and evasion (horizontal equity), improve tax progressiveness (vertical equity), and - more generally - improve the efficiencyof the tax system. Chile's Tax System 6.40 As can be seen in Figure 6.12, there is a positive correlation between the per capita income level of a country and the ratio of tax revenue to G D P . ' ~Regardless of~ whether correlation denotes causation, it seems clear that the tax ratio in Chile is, given its income per capita, below the averageof comparable countries. 6.41 Tax revenues increased 46 percent in real terms between 1995 and 2004. Figure 6.13 shows the trend of real revenues by tax category. During the period 1995-2004 income taxes revenue increased by 65.7 percent, despite a tax reform that reduced the top marginal rate and increased the minimum income exemption. Over the same period, revenues from the Value Added Tax (VAT) and excise taxes increased by 55.6 percent and 64.1 percent, respectively.In part this was the result of an increase in the VAT rate in 2003 (to offset import tariff reductions under the FTAs with the US, EU and Korea) and several gasoline and tobacco tax increases during the period. 161Sample of 60 countries based on 1995 dollars at PPP. It includes all sources of revenues in the case of Chile. In 2000-03 tax revenue in Chile averaged 16.5% of GDP whereas total revenue of the central government averaged21.4%of GDP. 185 Figure 6.12: Tax Ratios and GDP per capita Tax Ratio versus GDP per capita (1995 US$) 0 10000 20000 30000 40000 50000 60000 GDP per capita Source: C. Agostini (2004). Figure 6.13: Revenuesby Tax I Real Revenuesby Tax 1 1 +Income ~ Taxes - t V A T Excise Taxes 1 +Transaction Taxes +Tariffs +Others Source: C . Agostini (2004). 6.42 The Chileantax structureis dominatedby the VAT, which represents 50 percent of the total tax revenue. The two other main sources of tax revenue are the income tax (personal and corporate) and the excise taxes on cigarettes, alcohol and gasoline which represent 25 and 13percent of federal tax revenues, respectively. The shares of the three main sources of tax revenue have been relatively stable over the last ten years. 6.43 As pointed out by Auerbach and Kotlikoff (1987), based on statistical simulations, the most distortionarytaxes are taxes on capital (savings) followed by taxes on income and, last of all, consumption taxes. Thus, Chile's tax structure - which leans toward consumption taxes - is relatively non-distorting because it only affects the consumption/leisurechoice. 6.44 Chile's income tax includes 4 categories: (i) the first category (corporate) consists of a 17percent uniform rate on an accrual basis to business income and to capital gains when assets are sold162- about half the rate prevailing in most Latin American and OECD cowitries; (ii) the second category (payroll) is a progressive tax with 8 brackets and a marginal tax rate of 40 percent applied on salaries, wages and any other remuneration for personal services; (iii) the third category (surtax) is a progressive tax assessed on the worldwide income of residents from all sources - 8 brackets and 40 percent top marginal tax rate; and (iv) the fourth category (additional) is a 35 percent tax levied on income withdrawn or remitted abroad by non-resident individuals or corporations. 6.45 First and second category taxes paid can be used as a credit against the s~lrtax; implying that the dividends of Chilean corporations are not subject to double taxation. Similarly, first category tax can be used as a credit against any additional tax. Social security contributions (19.7 percent) are deducted from the payroll tax. There are numerous other exemptions and deductions from the personal income tax. Exemptions - with caps in the absolute amount - include income from investments in mutual funds, investment funds, rental properties, withdrawals from vol~~ntarysavings in the pension system and income from interest and capital gains. Deductions - also capped- include voluntary savings in the pension funds, mortgage payments, scholarships, and contributionsto universities. 6.46 Chile created a special regime to provide economic stability for foreign investment through Decree Law 600 (D.L. 600). Under D.L. 600, investors are guaranteed a fixed tax liability in exchange for accepting an additional (fourth category) tax rate of 42 percent instead of 35 percent. FDI under D.L. has been substantial and rising (Figure 6.14). Usually investors get up to 3 years to make their investments but in the case of large mining projects it can be up to 8 years (and if previous exploration is required up to 12years). Figure 6.14: Foreign Direct Investment under D.L. 600 Materialized Foreign Direct Investment (DL 600) I I Source: C. Agostini (2004). 16' Sales of Chlean corporations held for over 1 year are subject to a 15 percent flat tax rate and sales of real estate by individualson a non-recurrent basis are exempt. 187 6.47 The VAT, a tax on the sale of goods and services,was increased from 16percent to 18percent in 1990 and then to 19percent in 2003. There are exemptions for transport, financial services, education and health, professionals, and real estate rentals. There is a special regime for constructionwhereby companies are allowed to use against the VAT a credit of 65 percent of the VAT on their sales. The Internal Revenue Service (SII) has estimated that VAT revenues would increase by 10.6 percent if the exemptions and credits to the VAT were eliminated. Luxury goods, tobacco and alcoholic beverages are subjectto higher rates. 6.48 Chile has a stamp tax levied on credit operations. Among the excise taxes, it is noteworthy that the tax rate on (more polluting) diesel is only about one quarter the rate on gasoline. 6.49 Chile runs special regimes for specific geographic zones. Investors are given tax incentives in: Arica and Parinacota (Region I), Eastern Island, Regions XI and XI1 and Tierra del Fuego. In some of these areas up to 40 percent of investment costs can be offset against payments of the corporateincome tax. Chile has, in addition,two free trade zones: one in the port of Iquique (Region I) and the other in Punta Arenas (Region XII). Merchants and manufacturers are exenipt from the corporate tax and VAT from imports (except for goods sold within Chile). 6.50 Several indicators suggest that the cost to taxpayers of complying with the tax code has been rising over recent years, although there is no comprehensiveestimate. The number of pages of the tax supplement published by the internal revenue service (SII) each year increased from around 60 in 1995-98 to around 110 in 2004. The average number of lines used in the s~~rtaxform increased 50 percent during 1997-2004,while the number of tax payers filing tax returns has doubled in the past 15years. 6.51 On the other hand, the budget of SII has more than tripled in real terms over 1990-2004(Figure 6.15). This is partly explained by a 66 percent increase in personnel. Figure 6.15: SII Budget Budgetof Sll Source: SII. Tax Avoidance and Tax Evasion 6.52 Tax evasion and avoidance is a source of horizontal inequity because individuals with similar income can end up paying different tax rates. It is also a source of market inefficiencybecause it is difficult to have a competitivemarket when some of the sellers avoid or evade taxes whereas others don't. Tax evasion can be reduced by increasing penalties andlor by increasing the probability that the tax evader will be caught - although a very high penalty can reduce the probability that it will be enforced. In general there is a positive relationship between tax evasion and the cost of compliance; and evasionis lower when there is withholding at source. More generally, tax evasion and tax avoidance varies accordingto the type of tax and taxpayer, institutionalarrangements and cultural factors. The following features of the Chilean tax system constitute a potential sourceof tax avoidance. 6.53 Foreign direct investment (FDI).Multinational firms can avoid taxes through tax planning by shifting operationsto low tax countries, transfer pricing aniong related firms, taking advantage of differences between tax rules of different countries, or negotiating tax subsidies with host countries. The financing of foreign subsidiaries also offers opportunities for tax avoidance. For example, if a parent company finances its investments with equity, then its profits are taxable in the host country and no taxes are due in its home country until profits are repatriated. If investment is debt financed, the subsidiary can deduct interest paid and generate taxable receipts for the parent company. Therefore there is an incentive to finance investment with equity (debt) in low (high) tax host countries. 6.54 In the case of Chile, the 35 percent additional tax rate levied on profits distributed or remitted to foreign shareholders is likely to encourage tax avoidance and may also discourage FDI. While first category income taxes paid can be credited towards fourth category liabilities, the effectivetax on distributedprofits is still 35 percent. When evaluating incentives for tax avoidance, though, it is also important to consider the tax liabilityin the home country. 6.55 In order to avoid double taxation, some countries offer a tax credit for income taxes paid to foreign governments. However, in the case of a US corporation, for example, a company only gets full credit for its foreign taxes when the average foreign tax rate is less than the domestictax rate. The difference in the tax rate between the home country and the foreign country becomes then a relevant parameter in deciding the location of investment. Thus, a high additional tax rate can - even in the presence of tax credits- discourage FDI. A different situationmay arise in countrieswhich tax income at source and exempt foreign income from corporate income taxes. In this case, the relevant tax rate for deciding investment is the tax rate in the host country and, ceteris paribus, a high additional tax rate will deter FDI in favor of other countries with a lower tax rate. 6.56 VAT. Figure 6.16 shows trends in the estimated rate of VAT evasion. The evasion rate dropped fast during the early 1990s, then increased moderately until 1998 and since then it steadily decreased to 15percent in 2004. Figure 6.16: VAT Evasion Rate VAT Evasion Rate 35 30 25 s 20 15 10 5 0 s8s9\'9 s s s s s \ 9 s + + + + + & 8 9 b 8 9 6 4 8 8 8 6 9 8 ~ b Source: SII. 6.57 There are several mechanisms for VAT evasion: (i) failure to issue receipts for sales to final consumers; (ii) presenting counterfeit receipts to increase the amount of credit against VAT payments; and (iii) registering purchases for personal cons~mptionas purchases for the firm (owners of a firm can buy a personal computer or a car, especially a truck or a pick up, and then use the receipts as a credit for VAT paid). Firms which relocate to regions subject to special tax regimes with VAT exemptions provide an example of tax avoidance. 6.58 Personal income taxes. Tax avoidance is made possible by the fact that income taxes are levied on a narrow residual income base after allowing for deductions, exemptions and credits. Income tax avoidance may result from a taxpayer assigning income into a lower tax category or increasing "tax-deductible consumption" such as mortgagepayments, charitable contributions, and voluntarypension contributions. 6.59 The empirical evidence shows that evaders respond to higher tax rates by increasing evasion. Crane and Nouzard (1990), for example, find that individuals with higher levels of income tend to evade more and that evasion is generally inelastic with respect to changes in both marginal tax rates and income. The reported income tax evasion elasticity is 0.2 with respect to income and 0.65 with respect to the tax rate. The results regarding avoidance are quite similar. Clotfelter (1983), for example, estimates a reported income tax rate elasticity ranging from -0.5 to -0.3 and Feinstein (1991) finds similar results. 6.60 It is difficult to measure the extent of tax avoidance and evasion in Chile, particularly for income taxes. However, the data suggest that taxpayers receiving income from more than one source are better able to reduce their tax liabilities compared to taxpayers earning similarincome but only from labor. Figure 6.17 shows the gap between the effective average tax rates of the second category tax and the surtax. Taxpayers for whom labor is the only source of income consistentlypay more taxes than taxpayers with similar income fkom multiple sources. On average, individuals paying only second category taxes pay an effective rate 0.5 percent points higher than individuals with similar income paying surtax taxes. The example in Box 6.1 shows how independent professionals - who would normally be subject to second category income tax - can benefit from a lower joint tax bill by incorporating themselves into a single conipany subject to category 1income tax. Figure 6.17:Average Tax Rates Gap Between Second Category Tax and Surtax Average Tax RatesGap o y ~ " l n 4 ~ $ ~ " 4 l n " 0 - - ? - Z ? R % & UTA Source: C. Agostini (2004). Box 6.1: Tax Avoidancethrough Incorporation Consider two individuals, both independent professionals. The first earns a gross annual income of CLP$12,000,000 from which a 30 percent of presumptive expenses can be deducted (professionals can either deduct actual expenses or a presumed expense equivalent to 30 percent of their annual gross income). He or she therefore has a taxable income of CLP$8,400,000 and pays CLP$174,505 in income taxes with an average rate of 2.1 percent. The second earns CLP$36,000,000 annually, deducts CLP$5,455,440 from hislher gross income (the maximum allowed), and pays CLP$3,389,989 in income taxes with an average rate of 11.1percent. Now, the two individuals decide to incorporate themselves into a firm in which each owns stock. The gross income of the firm will be CLP$48,000,000, if their actual expenses are equal to their presumptive expenditures the taxable income will be CLP$33,000,000 and the corporate income taxes will be CLP$5,610,000. If the firm distributes 100 percent of its profits, each individual's taxable income will be CLP$16,500,000 (10 percent marginal tax rate bracket) and each of them will pay CLP$858,961 in personal income taxes. As a result, they will pay a total of CLP$1,717,922 in income taxes, while before they paid CLP$3,564,494. The average effective rate will be 5.2. Stock ownership and profits can be distributed in a way such that both individuals have a lower average effectivetax rate. Avoidance and Tax Progressiveness 6.61 Figure 6.18 shows the average tax rate for the integrated Chilean tax income (second category plus surtax). It shows that the Chilean income tax system is quite progressive and particularly so for the highest bracket; progressiveness stems from the progressiveness of both the second category (pure wage income earners) and the surtax category. A progressive income tax does not only serve as an instrument for income redistribution, it also acts as a partial insurance mechanism: when income falls, tax liabilitiestend to fall more than proportionately. Figure 6.18: Tax Pro ressiveness (SII, 2003)'3 Average and Marginal Income Tax Rates 50.00% -7;mv-m-' ., .*mu. -! 5:. , '-= 40.00% 30.00% 20.00% 4- Marginal 10.00% 0.00% O ~ ~ s ? g $ ~ ~ ~ g s J ~ UTA Source: C . Agostini (2004). 6.62 There are, however, legal loopholes (such as exemptions, deductions, and tax credits) through which taxpayers can reduce their tax liability. The fact that tax credits, exemptions and deductions are mainly used by high income taxpayers implies that the effective progressiveness of the income tax system is less than it would otherwise be. In addition, high income taxpayers may be able to disguise part of their personal income as corporate income, thereby fiu-therreducing the progressivenessof the system. Figure 6.19 shows the effective average tax rates by bracket in 2004. The calculation is based on aggregated data by bracket and it reflects the effective average tax rate paid by the average individualwithin each bracket. 164 Figure 6.19: Average Effective Tax Rate, 2004 I EffectiveAverageTax Ratefor 2004 I I 1 1 2 3 4 5 6 7 Bracket Source: C . Agostini (2004). 6.63 To assess the impact of loopholes on progressiveness, all tax credits (except for tax credits originatingin first or second categorytaxes), exemptions and deductions were ignored in order to simulatethe would-be average effectivetax rate by tax bracket. Figure '63Data was provided by SII. Two sources of data were used. The fust (for second category) contains the number of taxpayers, taxable income, and tax payments, for each 5 UTM (monthlytax unit - equivalent to CLP$30,308in December, 2004) of income. The second (surtax)contains the number of tax payers, taxable income, deductions,credits, and tax payments by income bracket. 164It is not feasible to do the calculation with more disaggregated levels of income as with the previous figure because SII data contains credits and deductions only by bracket. 6.21 shows the actual average effective tax rate and the simulated effective rate when credits, deductions and exemptions are ignored.'65 6.64 Eliminating all credits that are not tax-related does not have dramatic effects on the average tax rates. However, the tax rates for higher brackets increase proportionately more than the ones for low income brackets. For example, for the second bracket the change is from 1.65 percent to 1.67 percent, but for the last bracket the change is from 28.46 percent to 28.97 percent, i.e. more than half a percentage point. On the other hand, when no exemptions and deductions are allowed against the personal income tax, the change in the average tax rates is more significant (Figure 6.20). The increase is proportionately larger for higher income brackets. The rate for the second bracket increases from 1.65 percent to 1.84 percent, while the average rate for the top bracket changes from 28.46 percent to 29.79 percent. Figure 6.20: Average Effective Tax Rates (with and without credits, deductions and exemptions) EffectiveAverage Tax Rates I 1 2 3 4 5 6 7 8 Bracket Source: C. Agostini (2004). 6.65 Given the empirical evidence concerning the impact of tax rate reductions on evasion and avoidance,166it seems reasonable to consider the following alternatives: (i) reducing the progressiveness of tax rates and the number of brackets; and (ii) eliminating exemptions and deductions in order to compensate for the loss in revenue and 165There are three important caveats that must be taken into account when interpreting this table. First, it has implicitly been assumed that the number of taxpayers within the same bracket remains constant. That assumption imparts a downward bias. Second, it has also been assumed that there are no behavioral responses to the change in the tax structure. That assumption imparts an upward bias. Thlrd, only deductions and credits on taxable income are considered and exempted income is not taken into account. That restriction imparts a downwardbias. Erard (1993), for example, finds that an increase in tax complexity leads to a greater use of tax consultants ,and the average level of non-compliance is higher for tax returns filed with paid assistance. progressiveness.167Tax avoidance and evasion are likely to be significantly reduced, particularly, if many exemptions and deductions are eliminated, and compliance and administrative costs could drop. An example of tax reform along the lines proposed is described in Box 6.2. Box 6.2: Reforming the PersonalIncome Tax-An Example Considerthe following income tax rates and brackets: Taxable Income in annual tax Marginal Tax rate units (UTA) UTA Yo 0 to 14 0 14to 50 5 50 to 80 15 80 to 120 25 Over 120 35 Compared to the actual income tax structure, the number of brackets has been reduced from 8 to 5, the exempted level increased from 13.5 to 14 UTA, and the top marginal rate reduced from 40 percent to 35 percent. If there were no behavioral responses and no general equilibriumconsiderations, the revenue loss resulting from ths new income tax structure would be around 16percent of income tax revenue in the short run,which is significant,and the progressiveness of the income tax would be marginally reduced. However, a decrease in tax avoidance and evasion could compensate for a loss in revenue. The net outcome it becomes an empirical question. Furthermore, if many exemptions and deductions are eliminated, the net effects on both the revenue and progressiveness are likely to be almost fully compensated. The Difference between Corporate and Personal Income Tax Rates 6.66 Optimal tax models indicate the disadvantage for a small open economy of corporate income taxes. This is because capital is mobile and the tax burden falls completely on the immobile factors, mainly labor and land, in which case it would be more efficient to tax them directly. The policy recommendation of these models is to set corporate tax rates equal to zero. 6.67 If tax avoidance is taken into account, however, there is a role for corporate income taxes. Gordon and Mackie-Mason (1994) show that the main role of the corporate income tax is to discourage income shifting between the personal and the corporate tax bases, or between domestic and foreign subsidiaries. Unless corporate tax rates are roughly comparable to personal tax rates, business owners would be able to avoid taxes by retaining earningswithin their firm. The policy recommendationof this model is to set corporate income tax rates equal to personal income tax rates. 6.68 The consequences of a gap between corporate and personal income tax rates go beyond these considerations, however. Gordon (1998) shows that the option to incorporate means that a low corporate tax rate relative to personal income tax rates encourages risk-taking. Therefore, given the positive effects of entrepreneurship on 167See Messere (1993) for a review of the effects of tax simplificationin OECD countries. economic growth, the policy recommendation is to set corporate income tax rates much lower than personal income tax rates. Consistent with this hypothesis, Lee and Gordon (2003) find a significant negative correlation between the corporate tax rate and economic growth in a cross-section of 70 countries during 1970-1997. The estimates suggest that cutting the corporatetax rate by 10percentage points can increase the annual growth rate by around 0.7 percent. 6.69 Given the theory and empirical evidence in the literature, when deciding how close to set corporate and personal income tax rates, there seems to be a trade-off between reducing tax avoidance (implying a low personal-corporate rate gap) and encouraging entrepreneurship (implying a large gap). 6.70 It is not clear what the optimal gap is for Chile, but the evidence available suggests that avoidance and evasion of income taxes are quite important. The SII estimates that evasion in the first category of income tax (corporate tax) was around 44 percent on average in 1990-97. The relatively large gap between corporate and personal tax rates might have played an important role in explaining evasion. In view of the likely impact of avoidance and evasion on horizontal equity, reducing the gap between corporateand personal income tax rates is an option to be explored. Policy Options to Improve the Efficiency and Equity of the Tax System 6.71 The Chilean tax system has several attractive features: an overall moderate level of taxation, a relatively non-distortingtax structuretilted towards consumptiontaxes and a low rate of corporate income tax, the avoidance of duplications, and price indexationto avoid erosion in the value of tax brackets. While the personal income tax is fairly progressive, the dominant role of the VAT in tax revenue, the tax credit on dividends,and various loopholes in the income tax tend to make the overall system less progressive, costly to administrate and comply with, and less equitable. The following options could help to addressthese issues. a) Income Taxes. 6.72 Chile could consider reducing the number of tax brackets and the marginal tax rates; this will reduce incentives for tax avoidance and evasion and can also reduce administrative costs of the tax system. Eliminating some tax credits, deductions and exemptions could also help to improve horizontal and vertical equity while reducing administrative costs. Chile could also consider reducing the additional tax rate (on foreign company dividends) of 35 percent. The rate is too high and it encouragesplanned tax avoidance. In the case of foreign firms fiom countries that use a credit system, it might create excess credits,whch cannot be recovered later. 6.73 The net impact of local and regional tax incentives on employment and investment is questionable. Their elimination could improve the allocation of resources and reduce incentives for tax avoidance and evasion. It is estimated that income tax revenues could increase by around US$50 millions and VAT collections could increase by about US$63 million if these specialregimes are eliminated. 6.74 Similarly, consideration could be given to the elimination of exemptions for capital gains, investments in national mutual filnds, income from interest, and income from rent of DFL2 housing properties. There are no reasons on efficiency grounds to allow them, and eliminating them would improve horizontal equity. In 2002, the revenue loss due to the exemption of rents from DFL2 housing properties was estimated by the SII at US$150 million. 6.75 Presumptive taxes deserve to be exanlined carefully. Enforcement is easier now than when these taxes were first designed. If they are not eliminated, the Government could at least consider a reduction of the limit under which a firm pays presumptive taxes instead of corporate income taxes. The estimated revenue loss stemming from presumptive taxes was US$41 millions in 2002. b) VAT 6.76 The Government could consider eliminating the following exemptions: life insurances, professional services, and real estate rentals. There are no externalities or any other efficiency reason for these exemptions. The exemptions for insurance and real estate rentals reduced VAT revenues by around US$87 million in 2002. The special credit for the construction sector cost US$191 million in 2002. If the Govenment wants to promote this sector it could provide a direct subsidy. On the other hand, the special rates on luxury items and non-alcoholic beverages could be reduced to the standard rate of 19percent. The direct impact on revenue would be a loss of only US$7.8 million and the administrativeand compliancecosts of the specialrates are high. c) Other 6.77 Diesel is a more polluting fuel and the preferential tax rate on diesel carries negative externalities. The gap between diesel and gasoline taxes could be balanced out with a special tax on diesel car's fuel consumption. If the tax is set to be equivalent to gasoline taxes on average, the increase in revenue would be around US$120 million, considering that around 16 percent of the cars in Chile have a diesel engine. The stamp tax distorts financing decisions and increases switching costs for consumers in the financial sector, which might have negative effects on the degree of the competition in these markets. The Government could consider gradually reducing or eliminating the stamp tax, identifying first less distorting alternative sources of revenue to compensate for losses. Finally, the SII co~~ldconsider undertaking a survey of tax compliance. Annex 6.1 summarizes policy options and their expected impact on equity, efficiency, and revenue. FINANCING DEVELOPMENT -FISCAL RULE, DEBT, PUBLIC INVESTMENT AND INTERNATIONAL RISK DIVERSIFICATION 6.78 Fiscal policy in Latin American countries tends to be pro-cyclical: deficits increase during good times and fall during bad times. This pattern is obviously inconsistentwith the use of fiscal policy as an instrumentof stabilization.Possible causes of pro-cyclical behavior are credit constraints and political incentives to "grab revenues" and raise Government spending in good times. In fact, the two explanations are likely related: markets might be unwilling to lend in bad times if they know that extra revenues in good times will be used, not to repay debt, but to hnd extra Government spending. 6.79 The adoption of a fiscal rule by Chile in 2000 (implementationbegun in 2001) should be seen against this background. The rule was conceived as a measure to break the pro-cyclical behavior of fiscal policy. When the rule was first introduced, the common expectation was of buoyant copper prices and an expanding economy and the rule was designed as a means of constraining the pressure to spend the expected high revenues. The rule16' mandates an annual structural surplus of 1 percent of GDP for the central Government. To calculate the structural fiscal surplus, actual revenues are adjusted to reflect the medium-term "trend"169in GDP growth and the long-term reference price of copper - Chile's main export."O Thus, when output is below trend and/or the copperprice is below the long term reference price, a surplus of less than 1 percent of GDP or indeed a fiscal deficit can be accommodated thereby permitting the budget to play a stabilizing role during the cycle. 6.80 Apart from its stabilizing role, the fiscal rule has helped to establish prudent fiscal management consistent with improved creditworthiness.As long as Chile's o~~tput fluctuates around its (positive) trend level and copper prices oscillate around the long term reference price set by the fiscal rule, the applicationof the rule leads to falling debt- to-GDP ratios and eventually the elimination of public debt. Application of the rule has held up quite well so far: in the 4 year period 2001-2004, the average cyclically adjusted surplus has been about 0.8 percent of GDP, which compares closely with the 1 percent target. 6.81 Public acceptance of the rule - linked in part to its relative simplicity and transparency - has held up remarkably well. Most criticism of the fiscal rule has to do with its implementation, particularly as it may prevent fiscal policy from fully exercising a stabilizing role. These criticisms are examined below and options for improvement are 168Strictlyspeaking, the structural surplus rule in Chile is not law. It is a self-imposedmeasure adopted by the by the government in the context of its fiscal policy from 2001 to 2005. 16'The methodology used to project GDP for purposes of the Structural Balance Rule does not reflect the potential output of an economy functioning at full capacity. Rather, it is an extrapolationof the actual trend GDP in previous years, enrichedby additionalinformation on domestic and externalvariables. 170TOincrease transparency, the Governmenthas delegated the estimation and projection of trend GDP and the long-term copper price to two committees of external experts. Estimates employ a standard production function approach.The estimationof GDP is then "filtered" to separate the trend and cyclical components. A Hodrick-Prescott (H-P) filter is used; but since the H-P techmque gives too much weight to the end points of the time series, these are eliminated before projections are made. suggested. One major theme over which there appears to be consensus is that any major change to the rule (including scrapping it) after such a short period would be a major blow to the credibility of Chilean economic policy, which in all probability would reflect severelyon its creditworthiness. 6.82 Over the longer term, though, several options could be considered.For any given level of expenditure, the fiscal rule constrains action with respect to the optimal mix of budget financing through taxation or through borrowing. Whether or not the present fiscal rule is an optimal arrangement exceeds the scope of this chapter. However, there are circumstances under which it may be preferable to avoid such constraints. For example, given the structural surplus target and prevailing tax revenue limitations, an application of the fiscal rule may require foregoing high return public investments in education or innovation with potentially adverse social consequences. The second part of this section thus looks at a long-term alternative that would help to address such constraints while preserving the creditworthiness and anti-cyclical objectives of the current fiscal rule. It also considers the introduction of financial strategies aimed specifically at addressing terms of trade volatility, which has been a driving factor of economic cycles in the country. Chile's Fiscal Rule - Performance, Shortcomings and Possible Technical Improvements 6.83 The rule held up well during the slowdown of 2002, and through some of the lowest copper prices in 100 years, withstanding considerable political pressure. In July 2003 the "long-term price" of copper for 2004 (which is used to compute cyclically adjusted Government revenues, and hence the maximum Government spending) was set at 88 US#/lb. But by the end of 2003, while the budget was being discussed, the price of copper increased to 120US#/lb. However, despitepolitical pressure to raise the long term estimate, the Finance Ministry resisted, and Government spendingwas set on the basis of a price for 2004 of 88 US#/lb. 6.84 As mentioned, there is in general no strong opposition to the rule on conceptual grounds. Most objections refer to implementation and they come with suggestions for technical improvements.171The main objections to the rule, and - where applicable - some possible modifications are summarizedbelow: 6.85 Objection # 1 The rule prevents fiscal policy from exercising a stabilizing - role. Since Chile's fiscal rule is based on the output gap, it is compatiblewith any degree of counter-cyclicalityof fiscal policy; the latter depends on the extent to which automatic stabilizers are built into the Government budget. At present, the counter-cyclical effectiveness of Chile's budget is limited because there are virtually no automatic stabilizers on the expenditure side. To compute the structural surplus only revenues are adjusted to obtain structural revenues as a function of the output gap. Increasing the l7'This conclusion emerged from many interviews, spanning a wide spectrum of views, during a visit to Chile in May, 2005 of Roberto Perotti (consultant) and Daniel Oks (staff). 198 share of automatic stabilizers to include some categories of expenditure (e.g. social programs) would increase the stabilizing role ofjiscalpolicy. 6.86 Objection # 2 - The rule is easy to manipulate, and forecasts may not be consistent. The forecasts of copper prices and of the output gap are produced by two panels which in 2004 were composed of 12 and 14 experts respectively. The broad spectrum of backgrounds and interests of panel members as well as their technical reputation and integrity make it unlikely that they would provide biased forecasts. The methodology adopted calls for outlying results to be discarded. Thus, in general, the presumption that the rule is easy to manipulateis not justified. 6.87 One source of concern, however, is that while there is a high historic correlation between Chile's GDP and the price of copper, GDP forecasts and the forecasts of copper prices are independent of one another. A useful improvement would imply making sure that the twoforecasts are mutually consistent. 6.88 Objection # 3 The rule is not transparent, and therefore it is not easy to - communicate. The presumption is that the notion of cyclically adjusted revenues is hard to communicate to the media and the politicians; and the construction of the output gap and the long run copper price is convoluted. In practice, though, exactly the opposite appears to be the case.17' The cyclicallyadjusted rule is an easy and effectivebenchmark to communicate to politicians and the media. 6.89 Objection # 4 The implementation of the rule is incorrect. The long-run - forecast of the copper price is extrapolated from a moving average of the last few years. However, a significant amount of literature that argues that commodity prices are essentially a random walk, or that, in other words, their change is unpredictable. If the copper price is indeed a random walk, forecaststhat are based on a moving averageof the last few years will be consistentlywrong. They will under-predict the price and impart a contractionary bias to the budget during a period of rising copper prices. Conversely, when copper prices are falling, the forecasts will over-predict the price, thus exerting an expansionary bias. All this in itself will exaggerate the counter-cyclical effect of the budget more than intended. 6.90 Similarly, there is evidence that the forecast of potential GDP is essentially an extrapolation from a moving average of actual GDP. Forecasts from year t onward are then pooled with the actual time series from initial year to t-1 and subject to Hodrick- Prescott filtering. However, in order to assess whether this imparts any systematicbias to the cyclical adjustment of fiscal policy one would need to distinguish between the trend vs. cycle decompositionof real GDP in Chile, which is a notoriouslydifficult exercise. 6.91 Chilean policymakers seem to be well aware of these issues. In the case of copper prices, the correct approach would depend on the characterization of the stochastic process- a highly controversial issue. Thus, in the end, the current approach 17'Idem footnote 13. might be just right: an alternativethat puts more weight on recent copperprices to predict the futurewould, among other things, imply much more variability of fiscal policy. 6.92 Objection # 5 - The rule imparts excessive intra-year variability to fiscal policy. Under the rule, any intra-year revision in forecasts or, more generally, any shock that would affect the annual surplus, would have to be corrected within the year. For instance, in 2003, when the trade agreement with the European Union led to a permanent downward revision in tariff revenues, spendingwas cut accordinglyduring the year. This could impart excessivevolatility to fiscal policy. An alternative that has beenproposed is to correct the effects of any shock over a period oJ say, 3 years. More generally, it has been argued that, if one goal of the rule is to provide more stability tofiscal policy, it should really beformulated as "1percent surplus on average over a period of x years". Perhaps permanent, and presumably larger, shocks could be adjusted over a longer period than temporary ones. 6.93 While the merits of this proposal are obvious, the possible costs must also be considered: lengthening the horizon increases the incentive to postpone any adjustment, and could lead to problems of time-inconsistency. The experiences of European fiscal rules over the 2000-2001 cycle provide a clear example. The rules implied by the EU Stability and Growth Pact require the budget to be balanced on average during the cycle. 2000-01 was a period of output growth above potential, but the fiscal adjustments necessary to bring the budget into a surplus -a necessary condition if the budget is to be balanced on average during the cycle- were postponed, hoping that the high growth would continue. Eventually, in 2002, the euro zone economies entered a slowdown starting from a negative budget position. 6.94 Another aspect of the rule that increases fiscal variability is the normalization of the cyclically adjusted surplus. Currently, actual GDP is used as the denominator of the surplus-GDP ratio. If GDP turns out to be higher than forecast, the cyclically adjusted surplus, as a share of actual GDP, is lower than would result if the GDP forecast were used as the denominator, and an extra-adjustment will be required. Usingpotential GDP as the denominator would eliminate this source of intra-year variability infiscal policy. However, the difference is likely to be small. And Chile'spractice of using actual GDP as the denominator may reduce the likelihoodof mis-measurement ofpotential GDP. 6.95 In short, the Government could consider improvements to the fiscal rule by extending the role of automatic stabilizers to some expenditures, improving the consistency between GDP and copper price forecasts, and reducing the intra-year variability of fiscal policy. More substantial modifications to the fiscal rule should only be considered once the rule has acquired enough credibility, so that it could withstand some modification without major consequences. Five years-the time elapsed since the rule was introduced-may still be too soon. Longer Term Issues and Options 6.96 If an important goal of the rule is to enhance country creditworthiness,it should probably target Government debt rather than the deficit. International capital markets are more concerned about the stock of debt outstanding rather than the flow. Participants in the debate were aware of this issue when the rule was being discussed. However, there were probably sound practical reasons to focus on the surplus. Conceptually, the best indicator to focus on is net Governmentliabilities or, perhaps, liabilities less liquid assets. Conceptuallywhat matters is net Government liabilities relative to the net asset position of the private sector, i.e. the net asset position of the whole country. The asset side of the net liabilityposition is always difficult to measure and somewhat arbitrary (what types of assets should be counted?). Examples are financial operations whereby the Government sells some of its property and then leases it back-a practice common with Government buildings and some military equipment. Hence, focusing on the deficit was probably a more practical and better option. 6.97 The 1 percent surplus target for the cyclically adjusted surplus may also be considered arbitrary. Importantly, at the interest rates and average rates of growth of the last few years, it implies a falling debt to GDP ratio, which at about 35 percent (consolidating the central bank) is already low. While the 1 percent figure is certainly arbitrary,it did have a rationale: the estimated quasi-fiscal deficit of the central bank was also around 1 percent of GDP. Hence, a 1 percent surplus of the central Government would imply a balanced structurally adjusted surplus of the consolidated public sector. Even so, even a balanced budget is an arbitrarytarget as it would imply- as long as GDP grows - a falling debt-to-GDPratio. 6.98 As mentioned, Chile's fiscal rule is too new to warrant major changes. Rules work when they enjoy credibilityand any change to a relatively new rule, even ifjustified and explained, runs the risk of destroying credibility. This, however, does not rule out thinking about possible improvements for the longer term in order to accommodate the criticisms described above. The adoption of a flow (the budget surplus) rather than a stock (public debt) as the operational target and a reexamination of the arbitrary 1percent level of the target are issues that may be considered. 6.99 One possible alternative which could be examined is a rule under which the debt-to-GDP ratio eventuallyconverges to the ratio of the stock of public capital to GDP. This has the benefit of avoiding a scenario in which all public debt is eliminated at the cost, for example, of foregoing potentially high return public investments or expenditures. By the same token, it avoids constraining high social return public investment projects. 6.100 Although there is no unambiguous definition of what an "optimal" level of public debt should be, a stock of debt equal to a country's stock of public capital is certainly one possibility. As is demonstrated in Box 6.3, a fiscal rule which over time implies the convergence of the stock of debt to the stock of public capital is one which aims for a budget which is balanced over the cycle. In the steady state, this implies that public investment net of depreciation and maintenance costs must equal the net increase in debt. 6.101 The application of this rule in practice could encounter several obstacles. First, from a practical point of view the rule would create incentives to reclassify current spending. Well developed public institutions and methodologies to ensure proper accounting of public expenditures and public investment would be a key requirement. Second, as mentioned earlier, the measure of public debt employed is not trivial. Conceptually, the most appropriate measure to adopt would be the net liabilities of the Government. Box 6.3: The Arithmetic of Public Investmenta Let r be the real rate of interest, n the growth rate of GDP, 6 the rate of depreciation and the cost of maintaining public capital and 8 the gross financialrate of return on public capital. In general 8 < r + 6: Public investment is worthwhile from a social point of view although its net financial rate of return, 8 - 6, may be lower than the financing cost, which in turn we expect to be smaller than the social rate of return on Government projects. Let k be the stock of public capital, i public investment, so i =dkk + (n + 6 )k, and b the stock of public debt, each as a fraction of GDP. Also assume that there is no inflation. The Government'sbudget constraint is (ignoringinflationand its role on the budget constraint.} d b / b = g - t + i - 8 k + ( r - n ) b where t and (g + i) denote taxes and Government spending. Let's assume that the Stability and Growth Pact (SGP), as currently applied, requires countries to run a zero budget deficitwhen output is close to potential (it is straightforwardto extend our argumentto the case where the SGP is taken to allow a deficitof x% of gdp): g-t+i-Bk+rb=O so that dbh= - nb the debt ratio will eventuallygo to zero. Supposenow, as is usual for firms, that only capital depreciationand maintenance costs are included in current spending,and impose the rule that current spending be balanced.This leads to g-t+(6-B)k+rb=O so that dbh -dMk =- n (b-k) which indicates that eventually b-k -+0, no matter what the initial level of b. Eventually the entire stock of public debt is backed by public capital. If the stock of public capital, as a fraction of gdp, is constant,this rule allows the Government to run a deficit which is equal to nk. a. This Box draws upon Blanchard and Giavazzi (2004). 6.102 Demographic factors and social preferences are bound to raise demand for public goods that need to be financed by Government. Chile is likely to experience demand pressures in social transfers and, possibly, public investment. Demand for new public goods are likely in the areas of citizen's rights, environmental protection, urban development, education, innovation, justice and security. However, the pressure of demand for more public goods and services should not obscure the need for iniprovements in the efficiency of public expenditure, e.g. better targeting of educational stlbsidiesto address inequality. 6.103 Chile has a fairly efficient tax system - largely consumption-based - and tax administration. However, while the system includes a progressive personal income tax, its bias towards consumption-based taxes and the many deductions and exemptions imply that the system is not as progressive as it could be. In addition, the growing complexityof the system favors tax avoidance and makes it difficult to enforce compliance. There is scope for improvement in the tax structure so as to simultaneouslyincrease its efficiency, improve equity, simplifyits administration and increase overall compliance.In particular, reducing exen~ptionsand loopholes in the income tax which benefit primarily the upper - quintiles of the income distribution - could help to improve progressiveness of the system, reduce tax avoidance and lower collection costs. Reducing the tax on dividends paid to non-residents would simultaneouslyhelp to lower avoidance and encourage FDI. 6.104 Chile's falling public debt and, more recently, the adoption of a structural fiscal surplus rule, which is likely to fixther reduce the debt-to-GDP ratio of the central Government, have helped to establish the credibility of fiscal policy, facilitate anti- cyclical fiscal policy and enhance the perception of sovereign solvency. While the rule has held up remarkably well, there is scope to strengthen its implementation. For example, the Government could consider improvements to the fiscal rule by extending the role of automatic stabilizersto some expenditures, improving the consistencybetween GDP and copper price forecasts, and reducing the intra-yearvariability of fiscal policy. 6.105 More substantial modifications to the fiscal rule should only be considered once the rule has acquired enough credibility, so that it could withstand some modification without major consequences. Five years-the time elapsed since the rule was introduced-may still be too soon. Consolidatingthe credibilityof the existing fiscal rule remains the most important challenge. 6.106 Taking a longer-term perspective, though, there is scope to consider alternative fiscal rules which could help overcome some of the objections to the existing rule. In particular, a rule whereby debt converges to the public capital stock will ensure that the budget is balanced over the cycle and will avoid pre-empting high return public investment projects. Also taking a longer-term view, a strategy for international risk diversification through swaps of equity returns is proposed. While the possibility of equity swaps has existed for a long time, the more recent development of an adequate legal infrastructurenow makes this option more feasibleon a large scale. Income: Reduce numberof Ambiguos, brackets and probably Positive Negative marginaltax rates negative Allow recoup of excess credits for positive None Decrease Second Category taxpayers Increase: Eliminate Positive Positive US$50 geographictax millions at incentives most Eliminate exemptionsfor Increase: capital gains, Positive Positive US$160 income from millions renting DFL2 housing Eliminatetax Positive Positive Increase: 10~000 credits for forestry Reduce limits for- Increase: presumptivetaxes US$ 41 Positive Positive millions at most ReduceTax Rate None Positive Decrease on Additional Tax VAT: Reduce rates on Decrease: luxury items Negative, US$7.8 but small Positive millions at effect most Eliminate Increase: US$87 exemptionsfor Positive Positive millions at insurance and real estate most Eliminatethe Increase: special credit for US$192 constructionsector Positive millions at most Other Taxes: Enact a tax for Increase: diesel cars around None Positive US$120 millions Eliminatestamps Decrease: tax None Positive US$ 660 rr~illions 7. PROSPECTS FOR ACHIEVING CONVERGENCEAND REDUCING VULNERABILITY 7.1. Chile has an impressive record of rapid growth and poverty reduction which is due, in large part, to a wide-ranging and soundly-administered reforni effort. The Government has enjoyed broad domestic support for its twin agenda of growth and social equity. Chile is an outward-lookingcountry, and it appears ready to strengthenthe policy framework to further increase total factor productivity, boost efficiency and growth, cushion the domestic economy from shocks and combat poverty. Yet, some risks do remain. 7.2. The medium-term macroeconomic objective of the Chilean authorities is for GDP to grow at some 5 percent annually in real terms over the period 2005-2008 (Table 7.1). That would be equivalent to an annual real growth in GDP per capita of about 4 percent. At that rate of growth Chile, by 2023, would attain a level of GDP per capita equivalent to that of Spaintoday in terms of PPP. That growth rate is achievable, provided that Chile continues with prudent macroeconomic policies and barring any radical change in the world economic environment. Growth is expected to be driven both by exports and by internal demand. That projection is based on the assumption that the price of copper will gradually adjust downwards until it reaches the long-run equilibrium price estimated by the committee of experts which manages the fiscal rule in Chile. It is anticipated that stabilization policies, including a diligent application of the fiscal rule, will reduce fluctuations of GDP about the trend as compared with the past. Table 7.1 Chile: ProjectedMacroeconomicIndicators, 2005-2008 2005 2006 2007 2008 GDP Real Growth (%) 5.2 5.3 5.3 5.3 Internal Demand Real Growth (%) 6.6 5.7 5.4 5.3 Real Growth of Imports (%) 12.7 7.4 7.2 7.6 Consumer Price Index Change (%) 2.5 2.3 2.7 3.0 Nominal US$ ExchangeRate LIBOR 180-daysInterest Rate (%) Cooper Price (US cents per pound) 110.0 100.0 93.0 93.0 GDP Trend Real Growth 4.2 4.4 4.5 4.5 DeviationGDP Trend - Effective GDP 2.5 1.6 0.8 0.1 Long Term Cooper Price 93.0 93.0 93.0 93.0 Source: Informe de Finanzas Publicas 2005, Ministerio de Hacienda. 7.3. The baseline scenario is predicated upon a sound macroeconomic environment including a stable nominal exchange rate and a rate of inflation converging to the center of the target range. The Chilean authorities have demonstrated a steady and firm commitmentto sound macroeconomic policies. Suboptimal Growth and/or Volatile Growth 7.4. For purposes of projecting the convergence of Chile with the lower tier of OECD economies,the GDP of Spain is used as a benchmark. Lower-than-expectedGDP growth and/or volatile growth would threaten convergence. It is instructive to estimate the number of years that it would take Chile to reach the same level of GDP per capita as Spainusing different assumptions for the growth rate of each country (Table 7.2). Table 7.2 Chile: Projected Year of Convergencewith Spain (GDP per capita in PPP terms) Chile (% GDP growth) 3 4 5 6 0 2030 2023 2019 2017 Spain (% GDP 1 2043 2030 2024 2020 growth) 2 2083 2044 2030 2024 Note. Calculated using 2003 GDP per capita expressed in US$ PPP. Source: World Bank, Development Data Platform and World Bank staff estimates. 7.5. First, this analysis shows that, although Chile has enjoyed relatively healthy growth rates for the last two decades, there still remains a large disparity with OECD economies. Indeed, Spain has an income per capita that is more than twice the Chilean level, although the ratio has declined from 3.3 since the early 1970s. Consequently, even if Chile continues to increaseper capita income by 4 percent per annum -- the average of the 1990s -- full convergencewith Spain will take at least 20 years. Obviously, if Spain grows faster than expected, or Chile slower than expected, the process would take even longer. Hence, a second important conclusion from the analysis is that a couple of percentage points difference in the growth rate makes a big difference to the outcome. Basically, a one or two percentage point decrease of the Chilean growth rate adds 20 years to convergence.Notwithstandingthe progress that Chile has achieved so far, it will need to continue growing relatively fast in the coming two decades in order to reach its convergence goals. Figure 7.1: AlternativePossible Growth paths of GDP per capitain Chile, 2004-2030 (PPP in constant 2000 US dollars) 1-Alternative Scenario 2 ---+--Alternative Scenario 3 Source: World Bank staff estimates based on World Bank data. 7.6. Indeed, a sustained growth rate of 4 percent for more than 20 years is a very ambitious target given the erratic performance of most emerging market economies.Even short periods of divergence from a 4 percent trend would put the economy on a path that requires significantlymore time for convergence. If 4 percent annual growth is adopted as a baseline leading to convergence with the current per capita income of Spain in about 20 years, a number of alternativescenariosare possible (Figure 7.1): Baseline: If Chile continued to achieve 4 percent growth for the next two decades, per capita GDP would amount to more than US$21,000 by 2023 (Spain's GDP per capita PPP$ in 2003). That would be consistent with the medium-term projectionsof the Chilean Government. Historic trend: If Chile were to repeat the growth pattern it experienced since the early 1970s - an average growth rate of 4 percent but subject to volatility-it would take until 2030 to reach a per capita income of US$26,000. Alternative Scenario 1: If the growth rate in Chile were to decline gradually to a level closer to the experience of advanced economies, the outcome would be quite different.By 2030, GDP per capitawould amount to only US$17,000. Alternative Scenario 2: Even a brief setback similar to the slowdown in 1998 caused by the Asia/Russia/Brazilcrises could significantly affect convergence. If the economy were to grow at 4 percent for the first five years and then suffer a slowdown to 1.5 percent in the sixth year and to 2.5 percent in the seventh year before again resuming a steady 4 percent growth, the economy would be put on a markedly slowerpath to convergence. Alternative Scenario 3: If growth were to follow a track similar to the slowdown of 1998-2003 followed by a lower growth trend of 3 percent, there would again be a large divergence from the baseline case. 7.7. This analysis of alternative possible growth paths shows the importance of maintaininghigh growth rates and avoiding severe contractions if Chile is to attain a level of income comparable to more advanced economies in the next two decades. That will be a difficult endeavor, given the degree to which Chile remains susceptible to external shocks. Adverse External Conditions 7.8. Chile, a small open economy, remains sensitive to regional and global developments, and shifts in demand from its major trading partners. There is a strong association between GDP growth and external conditions (Figure 7.2). A slump in the market for primary products such as copper, or an interruption in capital flows can substantially reduce short-term growth. Over the past 100 years, external shocks have played a major role affecting output and incomes. Out of the six major economic recessions in Chile, five were precipitated by exogenous shocks: the crisis of the nitrate industry in the 1910s, the collapse in mining income that was precipitated by the Great Depression in the 1930s, the oil and copper crisis of 1975,the debt and banking crisis of 1982,and the slump in capital flows and commodity prices that followed in the aftermath of the AsiaJRussiacrises. Figure 7.2: GDP Growth and External Conditions, 1971-2003 (index) ' +Index of external conditions 11 UGDP growth, left 1 1 Sources: Chilean Central Bank, IFS,US Federal Reserve. 11Index of external conditions is a weighted average of the average growth rate of the 10 top trade partners, the change in the terms of trade, and the change in LIBOR. 7.9. Worldeconomicprospects. The exceptionalgrowth in world output seen in 2003- 04 is expected to moderate in 2005 and 2006 due to a slowdown in global trade and industrial production, an increase in interest rates, and a loss of liquidity. The fiscal stimulus that supported global growth is waning, and rising oil prices are also expected to slow down growth. The slowdown in global growth has the potential to turn into a more substantial downturn. Since the world economy is extremely dependent on the United States and China, a higher-than-envisaged fall in US demand or a sharp decline in China's rapid rate of growth or continued high oil prices would cause the world economy to slow down. Chile-a country wholly dependent on energy imports-would be especially affected by a sustained iise in oil prices. Chilean exports are also particularly vulnerable to a slowdown in Asia (and specifically China). However, in the short-run strong copper demand from Asia is likely to keep prices at a high level, thereby ensuring high export revenues for the Government. 7.10. In 2004, exports accounted for about 41 percent of Chile's GDP (up from 34 percent in 1990). Export earnings were particularly strong in 2004 due to high prices for Chile's main exports, particularly copper. Copper accounted for 45 percent of export earnings in 2004. Copper prices continue to be very high in mid-2005 and world copper inventories are at a 31-year low. Nevertheless, the record-high copper prices seen during 2005 are expected to decline in the near future as the global economy slows. That (togetherwith lower volume growth) would reduce copper export earnings. 7.11. Latin America. Chile is not very dependent on Latin America as an export market. Only a relatively small 16 percent of Chile's exports go to Latin America. Therefore, a decline in demand from the region is not likely to affect the Chilean economy seriously. The main risk confronted by Chile relate to the same external financing conditions that affect the entire region. Chile has built-up credibility through sound macroeconomic management; the reduction in sovereign spreads for Chilean sovereign debt has been remarkable and it places Chile in a class apart from its Latin American neighbors. In fact, Chile's average EMBI spreads since 2000 have been only one fifth of the composite for Latin America and they have declined to less than 100basis points in the last few years. 7.12. Notwithstanding this achievement, sovereign risk and hence the vulnerability of the capital account remains sensitive to the performance of other emerging markets, and particularly to the Latin composite. The correlation with the Latin composite has increased during periods of turbulence in emerging markets, indicating that Chile is still vulnerable to contagion shocks, although the low risk premiums that it currently enjoys provide a valuable cushion (Figure 7.3). Figure 7.3: SovereignRisk: EMBI Spreads2000 -2005 Overall Spreads and Correlationwith Latin Composite - --c--EM41Composite - EMBl Latin EMBlChile -c- Correlation right scale - Source: Bloomberg. 7.13. Export concentration. Chile has diversified towards non-traditional exports, which have doubled as a proportion of total exports since 1970.The most significantnon- mineral exports are forestry and wood products, fresh fruit, processed food, fishmeal, seafood, and wine. Also, there has been a diversification of export markets, with Asia being the fastest-growingexport market in recent years. Chile's major export partners are Asia (36 percent), the European Union (27.1 percent) and the United States (15.7 percent). Within the fast growing Asian market for Chilean exports the most important destinations are China (12.1 percent), Japan (11.5 percent), Korea (5.5 percent) and Taiwan (3.3 percent). However, Chile still remains heavily dependent on copper exports. As the world's largest copper producer, Chile accounts for one third of the global supply. It supplies 20 percent of China's copper imports and copper represents about 70 percent of the total exports of Chile to China. Over the period 1990-2004, the share of copper in total exports did not fall, but it fluctuated accordingto internationalprices, accountingon average for around 10percent of GDP and 35-47 percent of merchandize exports. 7.14. The pace of economic activity in Chile is thus linked to the price of copper. Since 1996 there have been six separate periods of growth and contraction of the Chilean economy, which in most cases have been preceded by a change in the price of copper in the same direction a couple of months earlier. The only period when that close prior relationship was not immediately apparent relates to the expansion in the second half of 1999. On that occasion, it appears that the change in copper prices affected the economy almost immediately and without the customary lag. 7.15. When changes in copper prices are plotted against GDP growth over 1960-2004, the association is clear (Figure 7.4). The close association represents a significantrisk for Chile, since copper prices are currently at extremely high levels. While there might be structural reasons (such as continuouslygrowing demand from China) to expect a higher equilibrium price, the risk of a price reversalremains a strongpossibility. Figure 7.4: Real Copper Prices and GDP growth 1960-2004 +GDPannualgrowth C h a n g e in realcooper price right - Sources: Central Bank of Chile, the Federal Reserve Bank of St.Louis and WDI. 7.16. To further illustrate the impact of changes in copper prices on GDP growth, a Vector Autoregressive model with exogenous variables (VARX) or near-VAR is employed. In a framework similar to that used by Deaton and Miller (1995), the VARX considers changes in aggregate demand, with output and its components-broken down into investment and consumption-all endogenous to the system. Copper prices are exogenous since Chile is assumed to be a price taker in the internationalcoppermarket. 7.17. Impulse response analysis graphically shows the effects on growth of a shock to copper prices. The responses of aggregate demand growth and its components to a one- off shock in copper prices are shown in Figure 7.5. The shock is equivalent to one standard deviation in year-on-year price changes, which is equivalent to about a 30 percent increase in copper prices. The standard deviation was calculated on the basis of quarterlydata from quarter 1, 1987to quarter 1,2005. Figure 7.5: Chile: Response to a Copper Price Shock -1.50 Time period: shock is in period 1 I I Source: Bai~coCentral de Chile, Federal Reserve Bank of St. Louis and World Bank staff estimates. 7.18. When changes in copper prices are subjected to a one-off shock equivalent to a single standard deviation, the response of growth in output is instantaneous and sizeable. There is an immediate increase in the growth of GDP equivalent to 0.22 percentage points of its initial value and, two quarters later, the increase amounts to 1.48 percentage points. Growth in consumption follows a similar path to growth in output. The one-off price shock has an even greater impact on growth in investment which increases immediately by 0.43 percentage points and by 2.84 percentage points four quarters later. The effects begin to die away slowly after 16 quarters. It should be noted that, in the earlier years from which data was drawn, copper prices changed by as much as 93 percent in a singleyear, equivalentto a shock of 3 standard deviations. 7.19. This section has reviewed a number of seriousrisks to Chile's economy including a slowdown in the demand for Chilean exports, a decline in the price of copper or the terms of trade more broadly (caused by an increase in the price of oil, for instance), and a reversal of capital inflows from abroad. Given the strong correlation between copper prices and economic activity, the Chilean economy could experience a contraction in the short term if copper prices revert towards their historical average. In terms of sovereign risk, Chile remains vulnerable to shocks in other Latin American economies but, in the last couple of years, it has made important strides to lowering the cost of borrowing from the international financial markets. Strict enforcement of the fiscal rule has been conducive to large fiscal surpluses and debt cancellation thereby paving the road for a less rough road ahead in response to futurecopper price volatility. 7.20. Financial instrumentsfor risk reduction. Chile has a large and well-diversified financial system. The mandatory pension funds established in the 1980shave grown fast and have provided a source of growth for other financial institutions includingbanks, life insurance companies, the mortgage industry and corporate bonds. Integration with the internationalfinancialcommunity was advanced yet further when the capital account was liberalized in 1998and, in particular, when pension funds were allowed to diversifytheir assets abroad. The foreign assets and liabilitiesof Chile as a proportion of GDP are about 50 percent greater than those of the average emerging country. Fiscal and financial sector strengths are the main pillars of Chile's strategy to reduce its vulnerability to terms of trade shocks and capital flow volatility. 7.21. However, there still remains scope to design specific instruments that could further shelter Chile from adverse changes in the terms of trade and, in particular, from fluctuations in the price of copper which weigh heavily on the budget. In emerging- market economies the domestic financial markets typically provide little scope for risk diversification. Risk diversification through international capital instruments is the obvious alternative. The cross-border transfer of real and financial assets -- e.g. via privatization-- is one way to diversifyrisk but it may be subject to political constraints. 7.22. Over-the-counter (OTC) derivatives such as swaps are another way of diversifying risk that merits consideration. Since OTC derivatives are not traded in a formally organized exchange, market participants are not subject to the constraints and restrictions that ordinarily govern a structured system of clearance and settlement. OTC derivatives offer a flexible and low-cost way of reducing risk provided they are skillfully structured and provided the contingent liabilities associated with all derivative contracts are carefullymanaged. 7.23. In the case of equity swaps, for example, the total return per dollar on the stock market of a small country may be exchanged annually for the total return per dollar on a market-value weighted-average of world-wide stock markets. An equity swap would enable a small country to diversify internationally without violating restrictions on investing capital abroad because there are no initial payments for the acquisition of a contingent liability. 7.24. Similarly, the Government might consider a swap with a global pension intermediary in which it offers to exchange annually part of its per-dollar "copper dividend" in CODELCO for the per-dollar return on a market-valueweighted-average of world-wide stock markets (see Box 7.1). The swap effectively transfers the risk of the "copper dividend" to foreign investors and provides the Government with the risk-return pattern of a well-diversified world-wide portfolio. An equity basket which is negatively correlated with copper prices may do even better in terms of risk diversification. However, there can be no guarantee that the correlationswould never go the wrong way, and an episode of "wrong" correlations could be problematic. That underscores the need for a careful management of contingent liabilities through a diverse and well-hedged portfolio. Box 7.1: International Risk Diversification throughEquity Return Swaps- An Example Suppose that the pension funds of a small-country, already invested in domestic equity, were to enter into swaps with a global pension intermediary (GPI). In the swap, the total return per dollar on the small country's stock market is exchanged annually for the total return per dollar on a market-value weighted- average of the world stock markets. This exchange of returns could be in a common currency, dollars, or adjusted to different currencies along similar lines to currency swaps. The magnitudes of the dollar exchanges are determined by the "notional" or principal amount of the swap to which per dollar return differencesapply. The swap effectively transfers the risk of the small-countrystock market to foreign investors and provides the domestic investors with the risk-return pattern of a well-diversified world portfolio. Since there are no initial payments between parties, there are no initial capital flows in or out of the country. Subsequent payments, which may be either inflows or outflows, involve only the dzfierence between the returns on the two stock market indices, and no "principal" amount. Consider, for example, on a notional or principal amount of US$1 billion. If, ex post, the world stock market earns 10percent and the small-country market earns 12percent, there is only a flow of (.12 - .lo) x US$1 billion or US$20 million out of the country. Furthermore, the small-country investors make net payments out precisely when they can "best" afford it: namely, when their local market has outperformed the world markets. In those years in which the domestic market under-performsthe world stock markets, the swap generates net cash flows into the country to its domestic investors. Hence, in our hypothetical example, if the small-country market earns 8 percent and the world stock market earns 11 percent, then domestic investors receive (.11 - .08) x US$1 billion =US$30 million, a net cash inflow for the small country. Moreover, with this swap arrangement, the ownership and trading of actual shares remain with domestic investors. 7.25. Another possible swap could involve exchanging the copper dividend for a risk- fkee interest rate denominated in a "strong" currency or in units of constant purchasing power. As with all swaps, the counter-party must have a very good credit rating, or it must be guaranteedby a third party with a strong credit rating, or by a two way market- to-market collateral. Guarantees of that kind can be costly. In economic terms, this swap would not be very different from what the Government is currently doing with excess copper revenues over and above the reference price set under the fiscal rule, i.e. accumulatingreserves or buyingback debt. 7.26. While the international financialmarkets can readily accommodate such financial transactions, the political economy of Chile may not yet be as well-prepared. The public may wish to be protected against falling copper prices but may be sceptical of the Government "giving away" part of the copper dividend when copper prices are rising. Moreover, the public may not like the idea of paying a premium to buy into the contract. Over time, some of these concerns could be allayed however. In this respect it is encouraging that the public seems prepared to accept a fiscal rule that forgoes the potential short run benefit of high copperprices whenever these are estimated to be above the long run price. Policy Options for Rapid Growth with Equity An integrated framework - 7.27. Chile is well placed to achieve high growth with increased equity, and to move towards convergence with the advanced OECD economies. Aside from consolidating macroeconomic and financial stability, achieving the objective of strong and equitable growth will require a combined effort to foster a dynamic and innovative enterprise sector and to raise the capabilities of the population, particularly the poorer members of society, by increasing access to high-quality education and by enhancing social protection. The biggest challenge for the future will be to reduce the still high degree of inequality as well as the vulnerability of many Chileans to poverty. The country cannot attain its growth potential if a large share of the population is ~lnableto develop their capabilities to the full. None of the advanced econon~ieshave a degree of inequality approaching that of Chile. Inequality of opportunity cannot but penalize Chile's growth prospects. 7.28. A dynamic and innovative enterprise sector. Growth has already made possible a large reduction in poverty. A continuationof rapid growth in the future will be necessary in order to further reduce poverty and vulnerabilityof low income population.A dynamic enterprise sector, in an economythat fosters competition and innovation,can generate the growth and employment needed for Chile to reach its goal of convergencewith advanced countries in terms of income and social equity. Chile faces the challenge of generating further productivity growth in the traditional export sectors and additional efforts to develop nontraditional exports if the country is to benefit fully from the new opportunitiesopened up by free-trade arrangements.Policies which are conduciveto such end include: strengthening of creditor rights, more efficient bankruptcy proceedings, tighter competition and increased labor market flexibility. Higher productivity will also stem from R&D investment - particularlyby the private sector - and in general improved knowledge flows across units of the National Innovation System. Suchpolicies have been the subject matter of chapters 2 and 3. 7.29. Increased education access and qualiq. Increasing education is necessary both for high growth and for greater equity. The educational reforms begun in 1990 have already led to a substantialexpansion of the quantity and quality of education.Additional investment has also brought significant gains in terms educational coverage. However, two important challenges remain: improving the quality of education and guaranteeing equal access to educational opport~lnity.The State can play an important role in ensuring that all children have access to a good quality of education. The State also faces the challenge to ensure equality of educational opportunity, particularly by promoting pre- school education and higher education enrollment among the poor. Chapter 4 focuses on policies to improve both the quality and the equalityof access to education. 7.30. Another substantial challenge is to redress existing shortfalls in the educational attainment of adults. A majority (53 percent) of the adult population in Chile has not completed secondary education, and low schooling levels are particularly prevalent among the poor. This is a huge barrier to equity and to innovation. But adult education alone will not solve the problem. On-the-job training also has an important role to play and that, in turn, will depend on the extent to which a greater part of the labor force can be brought into formal employment. 7.31. Enhanced social protection. Some 18.8 percent of the population is below the official poverty line. In addition, vulnerabilityto poverty remains high: none of the lower five deciles of the income distribution have a level of income very far above the poverty line. Many households remain vulnerable to poverty in the even of income shocks provoked by job loss, costly health hazards, and loss of earning-ability in old age. That not only poses a threat to equity, but it also compromisesthe outlook for growth. Where households do not have a sufficiently wide array of tools with which to insure against risks, resources that could otherwise be put to productive use may be preempted in adjusting to shocks. In some households, the effort needed to recover from shocks can even impair human capital or hinder investment. Chapter 5 discusses specific strategiesto enhance the socialprotection system. 7.32. Financing development. The Government is unlikely to depart from the supportive macroeconomic policy environment that has been put in place. However, the quest for higher growth is likely to put pressures on Chilean Government finances. Chapter 6 looks at the likely pressures on Government expenditures that will arise from fbrther growth and the financing options that are available to the Government. In particular, it assesses options to improve the efficiency and equity of the tax system and identifies alternatives to improve the effectiveness of fiscal arrangements to simultaneously enhance anti-cyclical policy, improve creditworthiness and avoid impairing growth prospects. AGENDA FOR FUTURE RESEARCH 7.33. This DPR has been selective in reviewing issues of concern to the development agenda in Chile. It was not intended to give comprehensive coverage to all the development issues in the country. A number of key policy cliallenges have not been covered and they could be considered as potential areas for future research. They include labor market participation, informality, pre-school educationand regional development. 7.34. Labor market participation. Differences in labor income have a direct bearing upon overall income inequality. Disparities of labor income can be explained both by differences in education and also by differences in labor market participation. Greater labor market participation among the poorer groups in society is central to increasing equity. Bringing the under-employed into the labor market would also increase growth. In Chile, the participationrates of women and young people are particularly low. 7.35. Female participation in the labor market stood at 33.7 percent in 2002. That is less than half the participation rate for men and also less than half the rate for women in most advanced economies. Data from the CASEN household survey shows a much higher participation rate for women from the highest income decile compared with those from the lowest income decile. The survey also points to the importance of education for labor market participation. Better educated women were much more likely to participate in the labor market (Mizala, Romaguera and Henriquez, 1999).As the majority of women in the poorest income deciles have only an elementary education, this has implications for poverty. Family size and childcare opportunitiesare other important factors affecting the participation rate. An extension of pre-school facilities may be one means of assisting women. 7.36. The scarcityof flexible employment arrangements, such as temporary or part-time jobs, impact on the decision of women with children to seek work. It is important to ensure that institutional rigidities do not play a role in discouraging female or youth employment. Increased access to temporary and part-timejobs for the young who are still in school or college may be important for raising their participationrates. 7.37. Informality. Employment in the informal sector entails special problems with respect to poverty and inequity. Participants in the infom~alsector typically have lower incomes and they may not have access to programs of social protection. Also, opportunities for education, training and other modalities of human capital improvement may not be availableto those employed in the informal sector.Mobilitybetween informal labor markets and formal labor markets appears to be low - a better understanding of outstanding barriers to such transition is needed. A dual labor market is an obstacle to the reduction of inequalityin general, and tends to deepen inequity in difficulttimes. Table 7.3: Structureof employment:Chile, 1990-2003 Definition 1 Definition 2 Formal Informal Formal Informal Source: World Bank staff estimates based on microdata fromthe CASEN. 7.38. The report touches on informality in chapters 2 and 5. It can be seen from the summary data in Table 1.3that the scale and trend of informal employment depends very much on the definition used. A more comprehensive analysis is warranted. Such an analysis would examine the main causes of informality; including its relationship to the business cycle and to rigidities in the formal labor market. It would also look at policy options for incorporating micro- and medium-sized enterprisesinto the formal sector. 7.39. Pre-school education. Although Chile has made substantial progress in educational coveragein basic and secondaryeducation, enrollment in pre-school is below internationalstandards.In addition,there is regional and socio-economicvariation in pre- school enrollment, which may translate into inequality in student outcomes in basic and secondary education. Low pre-school enrollment has been highlighted as one of the major policy areas in education for the incoming administration. Higher pre-school enrollment can impact positively on learning performance and thus constitute a key policy for growth. 7.40. Regional development. There is a wide disparity in GDP per capita between the ten different regions of Chile. The richest region has a level of income almost six times higher than that of the poorest region and convergence between the regions has been slow, especially when compared with the experience of developed countries. Recent estimates of the rate of convergence have ranged from 0.5 percent (Soto and Torche, 2004) to 1percent (Diaz and Meller, 2003). At this rate of progress it could take almost a century for Chile to close the gap between the rich and the poor regions (Duncan and Rodrigo, 2005). 7.41. The structure of the regional economies is diverse. In the north of the country mining has accounted for much of the growth since the 1980s.In the metropolitanregion of Santiago, industry and servicesprevail. Agriculture is the dominant sector in the ~0~1th and some 74 percent of the rural poor are concentrated there. 7.42. The predicament of the lagging regions may well be exacerbated by the progressive integration of Chile into the world economy because many of the agricultural products of the south face increasing competition from imports. In particular, wheat, which is typically cultivated by low-income smallholders, cannot be produced competitively in Chile. There has already been a 28 percent reduction in the land under wheat cultivation siiice 1989 and many farmers have abandoned wheat cultivation altogether. 7.43. In 1982, Chile introduced a price support system for wheat, wheat flour and sugar in order to protect the vulnerable agriculture sector. However, after concluding a number of free trade agreements and afler accession to the WTO, those price supports must now be phased out. The southern regions that specialized in import-competing crops such as wheat may not easily diversify towards nontraditional exports, particularly since they are not well-endowed with human capital and infrastructure.One way of tacklingthis may be through integratedregional or territorialpolicies. 7.44. The World Bark has studied regional issues in Chile on a selective basis. Most recently, there has been a study of rural infrastructure, and a look at Government policy in the extreme northern and southern zones. However, an evaluation of the overall regional developnlent picture is missing. It would be instructive to undertake a more complete analysis of regional issues such as regional income disparities and regional development policy including Government strategies with respect to the distributive impact of social expendituresin the regions. STATISTICALANNEX Table 1. Republic of Chile Gross Domestic Expenditure - (shares based on current price data) 1996 1997 1998 1999 Total consumption 74.3 74.4 76.4 76.8 Privateconsumption 63.3 63.3 64.9 64.4 Durable goods 5.2 5.2 5.1 3.9 Non-durablegoods 30.7 30.0 30.7 30.8 Services 27.4 28.0 29.1 29.7 Public consumption 11.0 11.I 11.5 12.4 Change in inventories 1.O 0.6 0.8 0.1 Gross fixed capital formation 26.4 27.1 26.1 20.8 20.7 21.7 21.3 21.2 20.6 Construction 16.3 16.6 16.2 13.8 13.0 13.4 13.4 13.7 13.3 Machineryand equipment 10.1 10.5 9.9 7.1 7.8 8.3 7.9 7.6 7.3 Exports of goods and services Goods Agriculture Fishing Mining Copper Other lndustry Services Importsof goods and services 29.0 29.2 29.6 27.3 29.7 31.8 31.6 32.6 31.9 Goods 25.2 25.1 24.9 22.0 24.6 25.9 25.5 26.4 26.4 Agriculture 0.6 0.5 0.5 0.6 0.5 0.5 0.5 0.6 0.5 Mining 2.0 1.8 1.6 2.4 3.3 3.3 3.2 3.8 4.2 Industry 22.6 22.8 22.9 19.0 20.8 22.1 21.8 22.0 21.7 Services 3.8 4.1 4.6 5.3 5.1 5.9 6.1 6.2 5.6 GDP 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source: Central Bank of Chile Table 2. Republic of Chile Gross Domestic Product by EconomicActivity - (shares based on current price data) 1986 1996 1997 1998 1999 2000 2001 2002 2003 2004 Agriculture 7.8 4.5 4.4 4.6 4.5 4.6 4.0 4.2 4.1 3.7 Fishing 1.2 1.3 1.3 1.3 1.5 1.3 1.2 1.3 1.6 2.2 Mining 10.5 7.1 6.6 4.9 6.2 7.5 7.0 7.3 8.9 14.1 Manufacturing industry 18.7 18.7 18.5 18.2 18.4 18.7 19.5 19.5 18.9 18.1 Electricity, gas and water 2.8 3.0 3.0 2.9 2.8 3.3 3.3 3.3 3.2 3.0 Construction 5.0 10.0 10.4 10.0 8.5 7.5 7.8 7.9 8.3 8.0 Retail, restaurantsand hotel 14.8 11.9 11.7 12.0 11.3 11.I 11.I 10.0 9.8 8.9 Transportand Communications 6.6 6.9 6.8 7.6 7.5 7.4 7.4 7.5 7.5 7.1 Financialservices 12.8 12.9 13.2 14.6 14.6 14.4 14.6 14.6 14.1 13.0 Housing 6.0 8.0 8.0 7.1 6.5 6.1 5.9 5.8 5.5 5.0 Personal services 9.3 11.3 11.6 12.4 13.3 13.4 13.5 13.9 13.6 12.9 PublicAdministration 4.6 4.3 4.4 4.5 4.9 4.8 4.6 4.6 4.4 4.1 GDP at market prices 100 100 100 100 100 100 100 100 100 100 Source:Central Bank of Chile Table 3. Republic of Chile Gross Domestic Expenditure - (Current Prices) (millions of pesos) Internaldemand Other Total consumption Privateconsumption Durable goods Non-durable goods Services Publicconsumption Change in inventories Gross fixed capitalformation Construction Machinery and equipment Exportsof goods and services Goods Agriculture Fishing Mining Copper Other Industry Services Importsof goods and services Goods Agriculture Mining Industry Services GDP Source: CentralBank of Chile Table 4. Republic of Chile Gross Domestic Product by Economic Activity - (Current Prices) (millions of pesos) Agriculture Fishing Mining Manufacturing industry Electricity, gas and water Construction Retail, restaurants and hotel Transport and Communications Financial services Housing Personal services PublicAdministration Subtotal 32,664,638 34,294,698 34,964,305 38,146,076 41,177,875 43,627,773 48,062,411 54363209 Minus: banking imputations 1,169,591 1,366,000 1,322,770 1,408,189 1,663,371 1,666,614 1,754,193 1750391 GDP at factor cost 31,495,047 32,928,698 33,641,535 36,737,887 39,514,504 41,961,I 46,308,218 59 52612818 Plus: VAT 2,456,888 2,750,512 2,849,997 3,128,292 3,353,528 3,716,616 3,899,440 4262166 Plus: importduties 770,701 855,663 647,010 709,140 668,720 664,053 523,045 481980 GDP at market prices 34,722,636 36,534,873 37,138,542 40,575,319 43,536,752 46,341,827 50,730,703 57356964 Source: Central Bank of Chile Table 5. Republic of Chile Annual Growth Rates of National Income and Product at Constant Prices - (percentages) 1997 1998 1999 2000 2001 2002 2003 2004 Internaldemand 7.2 3.7 -5.8 6.0 2.4 2.4 4.8 7.9 Other 6.1 4.4 -1.3 5.1 1.8 2.7 4.5 6.4 Total consumption 6.5 4.3 -0.4 3.6 2.9 2.5 3.9 5.2 Privateconsumption 6.6 4.7 -1.O 3.7 2.9 2.4 4.1 5.6 Durablegoods 11.3 2.4 -26.1 7.1 -2.7 -0.4 8.9 13.4 Non-durablegoods 5.6 3.5 -0.4 3.6 2.5 1.9 4.4 7.2 Services 6.9 6.4 3.2 3.3 4.0 3.2 3.2 3.0 Public consumption 5.8 2.2 2.7 3.0 2.9 3.1 2.4 3.0 Gross fixed capitalformation 10.5 1.9 -18.2 8.9 4.3 1.5 5.7 12.7 Construction 6.5 2.6 -12.4 0.8 4.5 2.2 6.0 6.0 Machinery and equipment 17.0 0.9 -26.9 23.2 4.1 0.6 5.4 22.8 Exportsof goods and services Goods Agriculture Fishing Mining Copper Other Industry Services Importsof goods and services 13.2 6.7 -9.5 10.1 4.1 2.3 9.5 18.6 Goods 12.5 6.2 -13.3 12.4 2.4 1.4 10.4 20.8 Agriculture -1.3 7.1 20.3 3.8 -11.5 4.1 15.9 3.2 Mining 5.8 17.8 16.7 -6.4 6.0 -5.1 12.0 8.1 Industry 13.4 5.2 -16.9 15.1 2.5 2.0 10.1 22.6 Services 18.0 9.9 13.4 -0.8 13.0 6.4 5.3 8.2 GDP 6.6 3.2 -0.8 4.5 3.4 2.2 3.7 6.1 Source: CentralBank of Chile Table 6. Republic of Chile Annual Growth Rates of National Income and Product at Constant Prices - (percentages) 1997 1998 1999 2000 2001 2002 2003 2004 Agriculture 1.7 5.0 -0.8 6.0 6.1 4.5 5.2 7.0 Fishing 9.5 -6.2 6.4 8.5 12.3 13.9 -11.9 21.3 Mining 11.3 8.3 10.6 3.2 5.7 -4.2 5.0 6.9 Manufacturingindustry 4.7 -2.3 -0.5 4.9 0.6 1.9 3.1 6.9 Electricity,gas and water 8.3 4.4 -4.7 9.5 1.5 3.3 3.8 3.8 Construction 6.3 1.9 -9.9 -0.7 4.1 2.5 4.5 5.0 Retail, restaurantsand hotel 7.6 3.5 -4.4 4.4 2.6 0.9 4.8 6.8 Transport and Communications 10.9 6.6 0.8 8.6 7.4 5.3 5.1 4.6 Financialservices 7.1 6.0 -1.O 5.0 3.6 3.1 3.2 5.9 Housing 3.9 3.4 3.0 2.3 2.1 2.2 2.1 2.2 Personal services 6.1 3.2 1.9 3.5 3.2 2.7 3.4 4.1 PublicAdministration 1.5 1.5 1.5 1.5 1.7 1.9 1.9 2.0 GDP at market prices 6.6 3.2 -0.8 4.5 3.4 2.2 3.7 6.1 Source: CentralBank of Chile Table 7. Republic of Chile Balance of Payments - (Millions of US Dollars) 1996 1997 I.CurrentAccount -3,083 -3,660 A. Goods and services -1,072 -1,563 Goods -1,072 -1,428 Exports 16,627 17,870 Imports -17,699 -19,298 Services -1 -136 Credits 3,588 3,892 Debits -3,589 -4,028 B. Income -2,518 -2,617 Wages -14 -21 Investment income -2,504 -2,597 Incomefrom direct investment -1,760 -1,943 Abroad 132 243 In Chile -1,892 -2,185 Portfolio income -281 -355 Dividends -219 -252 Interest -62 -104 Other investment -462 -299 Credits 710 923 Debits -1,172 -1,222 C. Current transfers 508 520 Credits 665 835 Debits -158 -315 Table 7 (cont). Republic of Chile Balance of Payments - (Millions of US Dollars) II.Capitalandfinancial account A. Capitalaccount B. Financialaccount Direct investment Abroad Stocks and other capital participation Reinvested profits Other capital In Chile Stocks and other capital participation Reinvestedprofits Other capital Portfolio investment Assets Liabilities Derivativefinancial instruments Other investment Assets Commercialcredits Loans Currencyand deposits Other assets Liabilities Commercialcredits Loans Currencyand deposits Other liabilities Reserve assets Table 7 (cont). Republic of Chile Balance of Payments - (Millions of US Dollars) 1996 1997 1998 1999 2000 2001 2002 2003 2004 Ill.Errorsandomissions 19. 238 -242 -1,074 447 -964 -939 -925 -1,061 Memorandumitems Balanceof Payments 1,122 3,320 -2,194 -737 337 -873 199 -366 -191 Current account excluding reserve assets 4,186 6,742 1,966 237 787 671 1,717 1,661 -519 (1) Includesinterest: To abroad 0 3 9 7 17 45 12 2 12 From abroad -95 -149 -161 -228 -276 -239 -262 -123 -193 (2) Net short term flows -869 -1,908 -2,304 -4,054 -126 -2,451 2,172 1,746 -2,541 Assets -856 -416 -1,946 -3,373 -1,922 -2,411 1,073 -256 -3,078 Liabilities -13 -1,491 -358 -681 1,795 -40 1,099 2,002 536 (3) Net liabilityflows due to medium term loans Payments 5,314 6,349 6,221 4,278 4,252 4,409 5,404 5,022 5,975 Amortizations -4,756 -2,979 -2,428 -3,354 -4,229 -4,723 -6,190 -5,796 -7,148 (Prepayments) -2,713 -1,027 -232 -754 -1,112 -1,831 -2,600 -1,539 -3,707 Source: CentralBank of Chile Table 8. Republic of Chile Trade - (Millions of US dollars) Exports, fob I.General merchandise A. General regime 1. Mining 2. Agriculture and fishing 3. Industry Food Beverage Forestry Paper and cellulose Chemicals Basic metal industry Machineryand equipment Other B. Trade free zone (zona franca) Imports I.General merchandise A. General regime 1. Consumer goods Durables Semi durables Other 2. Intermediategoods Combustibles Oil Non-oil Others 3. Capitalgoods B. Trade free zone (zonafranca) Total goods imports Minus insuranceand fright Total goods imports, (fob) Source: Central Bank of Chile Table 9. Republic of Chile - Prices and Exchange Rates Consumer price indeces (Dec 1998= 100) Consumer price index Excludingfruits, vegetables and energy Excludingenergy, utilities and others Consumer trabables price index Consumer non-tradable price index PercentageAnnual Variation Consumer price index (period average) Consumer price index (end of period) Exchange rates US Dollar nominal exchange rate Multilateralexchange rate Exchange rate for selected currencies 11 Multilateralreal exchange rate Real exchange rate for selected countries 11 Terms of trade (2000=100) Cooper price (US$) 11 United States, Japan, Euro, UK and Canada Source: Central Bank of Chile, International Financial Statisticsand DevelopmentData Platform. Table 10. Republic of Chile Monetary Survey and Interest Rates - (Averages in billions of pesos) 1,996 1,997 1,998 1,999 2,000 Monetary base 1,204 1,432 1,483 1,652 1,683 Currencyin circulation 748 880 853 997 1,001 Depositsin current account 1,530 1,923 1,677 1,951 2,046 M1 2,278 2,803 2,530 2,949 3,047 Sight deposits 311 305 321 478 471 MIA 2,589 3,108 2,851 3,426 3,518 Time deposits 7,994 9,829 11,650 12,605 13,856 M2A 10,583 12,936 14,501 16,031 17,374 Time savings deposits 1,683 1,875 1,867 1,982 2,149 M3 12,266 14,812 16,368 18,014 19,524 Central Bank paper 6,865 8,336 8,133 8,345 9,273 M4 19,131 23,148 24,501 26,359 28,797 Treasury's lOUs 0 0 0 0 0 M5 19,131 23,148 24,501 26,359 28,797 Letters of credit 3,187 4,089 4,287 4,841 5,218 M6 22,318 27,237 28,789 31,200 34,015 Deposits in foreign currency 547 553 1,091 1,742 2,097 M7 22,865 27,790 29,880 32,942 36,113 Interest rates 30 to 90 days deposits 12.7 11.4 13.9 8.2 8.8 31 to 90 days loans 16.1 14.6 18.4 11.9 13.9 90 days to 1 year deposits 14.2 12.7 16.3 8.9 8.7 90 days to 1 year loans 21.6 20.2 27.4 17.6 18.7 1 to 3 years deposits 14.5 13.1 20.3 14.4 10.9 1 to 3 years loans 26.3 23.1 26.0 23.7 23.3 Source: CentralBank of Chile Table 11. Republic of Chile Financial Sector Indicators - Credit Market Domestic credit provided by bankingsector (% of GDP) 68.59 71.53 69.61 72.60 76.31 73.86 70.59 Domestic credit to privatesector (% of GDP) 62.04 63.38 61.51 63.69 65.06 64.81 63.26 Stock Market Stocks traded, total value (% of GDP) Stocks traded, turnover ratio (%) Listeddomestic companies, total Source:Development Data Platform. Table 12. Republic of Chile External Debt - (Millions of dollars) 1996 1997 1998 1999 2000 2001 2002 2003 2004 I.Centralgovernment 2,669 2,284 2,183 2,594 2,420 2,887 3,565 4,551 4,720 1. Short term Moneymarket instruments Loans Commercial credits Other liabilities 2. Long term 2,669 2,284 2,183 2,594 2,420 2,887 3,565 4,551 4,720 Bonds and lOUs 500 500 1,145 2,098 3,159 3,433 Loans 2,669 2,284 2,183 2,094 1,920 1,742 1,467 1,392 1,287 Commercialcredits Other liabilities II.Monetaryauthorities 189 165 92 26 62 15 12 10 17 1. Short term 186 162 89 24 60 13 11 9 16 Moneymarket instruments Loans Deposits Other liabilities 186 162 89 24 60 13 11 9 16 2. Long term 3 3 3 2 2 2 1 1 1 Bonds and lOUs Loans 3 3 3 2 2 2 1 1 1 Deposits Other liabilities Table 12 (cont). Republic of Chile External Debt - (Millions of dollars) 1996 1997 1998 Ill.Banks 3,619 2,506 2,785 1. Short term 2,392 752 878 Money market instruments Loans 2,384 746 870 Deposits 8 6 8 Other liabilities 2. Long term 1,227 1,754 1,907 Bonds and lOUs 400 600 Loans 1,227 1,354 1,307 Deposits Other liabilities IV. Other sectors 16,438 20,372 23,216 1. Short term 4,466 4,608 4,155 Money market instruments Loans 438 692 834 Deposits Commercial credits 4,028 3,916 3,321 Other liabilities 2. Long term 11,972 15,764 19,061 Bonds and lOUs 1,194 1,648 1,683 Loans 9,862 12,440 14,950 Deposits Commercial credits 916 1,676 2,428 Other liabilities - V. Direct investment: inter- company loans 3,357 3,707 4,315 4,319 5,297 4,948 4,822 4,240 4,592 1. Short term 1 8 9 401 133 422 157 405 2. Long term 3,356 3,707 4,307 4,310 4,896 4,815 4,400 4,083 4,187 VI. Gross external debt 26.272 29.034 32.591 34.758 37.177 38.538 40.675 43.396 43.764 Source: CentralBank of Chile Table 13. Republic of Chile Fiscal Accounts - (millonsof pesos) 1996 1997 1998 1999 2000 2001 2002 2003 Central Government Revenues Net tax revenue Gross incomefrom copper Provisional provisions Grants Rents Operational income Other income Gastos Personnel Goods and services Interests Subsidies and grants Provisionalprovisions Other expenditures Gross operating result Source: CentralBank of Chile Table 14. Republic of Chile Income Distribution - Distribution of household per capita income Share of deciles and income ratios Chile, 1990-2000 Share of deciles Incomeratios 1 2 3 4 5 6 7 8 9 1 0 1011 90110 95/80 (i) (ii) (iii) (iv) (v) (vi) (vii) (viii) (ix) (x) (xi) (xii) (xiii) 1990 1.2 2.3 3.0 3.8 4.7 5.9 7.6 10.0 15.3 46.1 37.2 10.8 2.7 1994 1.2 2.2 2.9 3.7 4.7 5.9 7.4 10.1 15.3 46.6 37.7 10.9 2.5 1996 1.2 2.2 2.9 3.8 4.7 5.9 7.6 10.3 15.7 45.8 37.3 11.4 2.6 1998 1.2 2.1 2.9 3.7 4.6 5.8 7.5 10.2 15.7 46.3 39.6 11.9 2.6 2000 1.2 2.2 2.9 3.7 4.7 5.8 7.4 9.9 15.2 47.1 40.3 11.2 2.6 Source:Giovagnoli,Pizzolitto and Trias (2005) based on rnicrodata fromtl Giovagnoli,Pizzolitto and Trias (2005) Distribution of household per capita income Inequality indices Chile, 1990-2000 Countw Gini Ihell CV A(.5) At1,) . A(2), t(0), t(2) . . 0) (ii) (iii) (iv)' (v) (vi) (vii) (viii) 1990 0.562 0.676 1.960 0.265 0.435 0.686 0.570 1.920 Source:Giovagnoli, Pizzolitto and Trias (2005) based on microdata from the CASEN. CV=coefficientof variation. A(e) refers to the Atkinson indexwith a CES functionwith parameter e. E(e) refers to the generalizedentropy index with parameter e. E(l)=Theil. Table 15. Republic of Chile Poverty - Official poverty line National Urban Rural Poor and Poor and Poor and indigent lndigent indigent lndigent indigent lndigent 1987 17.4 43.6 16.7 51.5 20.6 Source: MIDEPLAN, EncuestaCASEN (1987,1990,1992,1994, 1996, 1998,2000 and 2003). REFERENCES Aedo, C. (1997). Organizacibn Industrial de la Prestacibn de Sewicios Sociales. Working Paper R-302. Washington, DC: Inter-American Development Bank. Aedo, C. and 0 . Larraiiaga. (1994). Educacidn Privada vs. Pziblica en Chile: Calidady Sesgo de Seleccidn. Mimeo. Santiago, Chile: Graduate Economics Program, ILADESIGeorgetown University. Agostini, Claudio.(2005).Analysis of the Chilean Tax System. Background paper for the DPR. Aguirre et al. (2004).EJiciencia del Sector Publico y Crecimiento.Mimeo. Araneda, Hernhn and Crist6bal Marin. (2002). Meeting the Challenge of the Knowledge Economy: Lifelong Learning in Chile. Washington DC: The World Bank. Arellano, J. P. (2004). Politicas sociales para el crecimiento con equidad Chile 1990-2002. Mimeo. Santiago, Chile. Arenas de Mesa, Alberto. (2000). Cobertura Previsional en Chile: Lecciones y Desafos del Sistema de Pensiones Administrado por El Sector Privado. CEPAL Serie, Financiamiento del Desarrollo No. 105.Santiago, Chle. Arenas de Mesa, Alberto and Mario Marcel. (1999). Fiscal Effects of Social Security Reform in Chile: The Case of the Minimum Pension. Mimeo presented at the APEC SecondRegional Forum on Pension Fund Reforms, Vina del Mar, April 26 -27. Arenas de Mesa, Alberto and Hector Hernindez Shnchez. (2001). "Cobertura del Sistema de Pensiones de Capitalizacibn Individual Chileno: Diagnostic0 y Propuestas". Socialis: Revista Latinoamericana de Politica Social, No. 4, Abril, pp. 93-124 Arenas de Mesa, Alberto and Paula Benavides Salazar. (2003). Proteccidn Social en Chile: Financiamiento, Coberturay DesempeZo, 1990-2000. Santiago,Chile: International Labor Organization. Arrau, Patricio. (2002). Agenda Pro Crecimiento Propuestas de reforma a1Mercado de capital II. Mimeo. Santiago, Chile. Arrau, Patricio. 2004. Bancarizaci6nen Chile: Un Nuevo Pacto Publico-Privado.GERENS S. A. Santiago, Chile. Aschauer D.A. (1989). "Is Public Expenditure Productive?" Journal of Monetary Economics, Vol. 23 (2). Atanasio, 0 . and M. SzCkely. (2001). "Going Beyond Income: Redefining Poverty in Latin America". In Orazio Attanasio and Miguel SzCkely(eds.). A Portrait of the Poor: An Asset Based Approach. Baltimore,MD: Johns Hopkins UniversityPress. Atkinson, A. and J. Stiglitz. (1972). "The Structure of Indirect Taxation and Economic Efficiency". Journal of Public Economics Vol. 1. Atkinson, A. and J. Stiglitz. (1976). "The Design of Tax Structure: Direct versus Indirect Taxation". Journal of Public Economics Vol. 6. Auerbach, Paula, Maria Eugenia Genoni and Carmen Pages. (2005). Why are Social Security Contribution Rates soLow and Unequal in Latin America? Mimeo. Washington D.C. Ballou, Dale and Michael Podgursky. (1998). "Teacher Recruitment and Retention in Private and Public Schools." Journal of Policy Analysis andManagement, 17(3):393-417. Banda, Enric. (2004). University-IndustryKnowledge and Technology Transfer in the European Innovation System. Barr, Abigail and Truman Packard. (2000). Revealed and Concealed Preferences and Self Insurance: Can we Learn from the Self Employed in Chile. Department of Economics DiscussionPaper SeriesNo. 53. Oxford,UK: Oxford University. . (2002). RevealedPreference and SelfInsurance: Can we Learn From the SelfEmployed in Chile? Policy Research Working Paper. Washington, DC: The World Bank. Barrientos, Annando. (1996). "Pension Reform and Pension Coverage in Chile: Lessons for Other Countries". WorldDevelopment, Vol. 15,No. 3, pp. 309-322. . (1997).Pension Reform and the Individual Capitalization Pension Scheme Stampede in Latin America. Working Paper No. 17.University of HertfordshireBusiness School. . (1998a). "Pension Reform, Personal Pensions and Gender Differences in Pension Coverage". WorldDevelopment, Vol. 26, No. 1pp. 125-137. .(1998b)Pension Reform in LatinAmerica. Ashgate. . (1999a). Retirement, Household Income, and Economic Vulnerability of Older Groups. Working Paper No. 20. University of HertfordshireBusiness School. . (1999b). Economic Risks, the Labor Market and Older Workers in Latin America. Working Paper No. 21. University of HertfordshireBusiness School. Barro, R. (1991). "Economic Growth in a Cross-Section of Countries". Quarterly Journal of Economics Vol. 106(2). Barro, R (1999). "Determinants of Economic Growth: Implications of the Global Evidence for Chile". Cuadernos de Economia 36 (107): 443-78. Bartelsman, E.J. and M. Doms. (2000). "Understanding Productivity: Lessons from Longitudinal Micro Datasets". Journal of EconomicLiterature, Vol. 38 (September). Bartzokas Mani, S. and A. Bartzokas. (2002). Institutional Support for Investment in New Technologies: The Role of Venture Capital Institutions in Developing Countries. DiscussionPaper 04. United Nations University, Institute for New Technologies. Baytelman, A. and M. Duce. (2003). Evaluation de la Reforma Procesal Penal: Estado de una Reforma en Marcha. Santiago,Chile:Universidad Diego Portales. Bellei, C., G. Mufioz, D. Raczynski, L.M. Perez, Unicef and Asesorias para el Desarrollo. (2003). "Escuelas Efectivas en Sectores de Pobreza: ~Quikndijo que no se puede?". In Hevia, R. (ed.).La Educacidn en Chile hoy. Santiago, Chile: Ediciones Universidad Diego Portales. Benavente, Jose Miguel. (2002). The Impact of Public Financing and Research Groups on InnovativeActivities in Chilean Industry. . (2003).Andisis de la Tercera Encuesta Sobre Innovacidn Tecnoldgica. Santiago, Chile: Universidad de Chile. . (2004a). Investigation y Desarrollo, Innovation y Productividad: un Analisis Econometrico a nivel de lafirma. . (2004b). Technology Innovations in Chile:Evidencefrom a National Survey. Benavente, Jose Miguel, de Gregorio and Nufiez. (2004). Rates of Return to Industrial R&D in Chile. Benavente, Jose Miguel., A. Galetovic aid R. Sanhueza. (2005). "La dinamica industrial y la financiacibn de las pynie". El Trimestre Econdmico 72,217-254, 2005. Bergoeing, Raphael, and Felipe Morande. (2002). Labor Taxes, Employment and Growth: Chile 1998-2001. Center for Research on Econonlic Development and Policy Reform. Working Paper No. 131. Stamford University Bergoeing, Raphael., A. Hernando. and A. Repetto. (2003a). Idyosincratic Productivity Shocks and Plant-Level Heterogeneity. Serie Economia No.173. Santiago, Chile: Departamento de Ingenieria Industrial, Universidad de Chile. . (2003b). The Efect of Policy Distortions on Productiviy Dynamics. Mimeo. Centro de Economia Aplicada. Santiago, Chile: Universidad de Cliile. Bergoeing, Raphael, Felipe Morande and Facundo Piguillem. (2003). Labor Market Distortions, Employment and Growth: The Recent Chilean Experience. Working Paper No. 251. Santiago, Chile: Central Bank of Chile. Bergoeing, Rafael and Andrea Repetto. (2003). Copiar no es malo: Competencia, Adopcidn e Innovacidn. Serie En Foco. Expansiva. Bergoeing, Raphael., N. Loayza and A. Repetto. (2004). "Slow Recoveries". Journal of Development Economics Vol. 75(2), 473-506. Bergoeing, Raphael, A. Repetto, and S. Edwards. (2004). "Productivity Dynamics". Journal of Development Economics Vol. 75(2), 329-332. Bernard, A.B. and C.I. Jones. (1996). "Apples and Oranges: Productivity Convergence and Measurement across Industries and Countries."American Economic Review 86:1216-1237. Berstein, Solange, Guillenno Larrain, and Francisco Pino. (2004). Cobertura, densidad y pensiones en Chile: Proyecciones a 30 anos plazo. Mimeo presented at the seminar "Competencia y Cobertura" hosted the by Centro de Estudios Publicos (CEP) and the Superintendencyof the AFP system (SAFP), 11-12November, Santiago, Chile. Bertola, G. (1990). "Job Security, Employment and Wages". European Economic Review,June, 851-886. Beyer, Harald, B. Eyzaguirre and L. Fontaine. (2000). La reforma educacional: una apreciacidn critica. Documento de Trabajo No. 306. Santiago, Chile: Centro de Estudios Pcblicos. Beyer, H. and R. Vergara. (2001). "Productivity and Economic Growth: The Case of Chile". In N. Loayza and R. Soto (eds.). Economic Growth: Sources, Trends and Cycles. Santiago, Chile: Banco Central de Chile. Beyer, Harald and Salvador Valdes. (2004). Propuestas para aumentar la densidad de cotizaciones. Mimeo presented at the seminar "Competencia y Cobertura" hosted the by Centro de Estudios Publicos (CEP) and the Superintendency of the AFP system (SAFP), 11-12November, Santiago,Chile. Bitran, Ricardo & Associates. (2002). The Risk of Poverty from Health Shocks in Chile. Background Paper for Chile: Household Risk Management and Social Protection. Washington, DC: The World Bank. Blanchard, Olivier and Francesco Giavazzi. (2004). "Improving the SGP through a proper accounting of public investment". In: Reformer the Pacte de StabilitB et de Croissance. Paris: Conseil d9Analyse~conomi~ue. Blank, Rebecca. (2003). US. Welfare Reform: What's Relevant for Europe. CESifo Economic StudiesVol. 49. Blomstrom, Magnus, hi Kokko, and Fredrick Sjoholm. (2002). Growth and Innovation Policies for a Knowledge Economy: Experiencesfrom Finland, Sweden, and Singapore. Boadway, R., M. Marchand, and P. Pestieu. (1994). "Towards a Theory of the Direct-Indirect Tax Mix". Journal of Public Economics Vol. 55. Boarnet, M. y W. Bogart. (1996). "Enterprise Zones and Employment: Evidence from New Jersey". Journal of UrbanEconomics. Bosch, M., D. Lederman, and W. Maloney. (2005).Patenting and R&D: A Global View.Mimeo. Boskin, M. and W. Gale.(1987). "New Results on the Effects of Tax Policy on the International Location of Investment". In: M. Feldstein (ed.). The Effects of Taxation on Capital Accumulation. University of Chicago Press. Botero, J., S. Djankov, R. La Porta, F. Lopez-de-Silanes,and A. Shleifer.(2003). The Regulation of Labor. NBER Working Paper No. 9756. Cambridge, Massachusetts:National Bureau of Economic Research. Botero, J., R. La Porta, and F. Lopez-de-Silanes. (2004). "The Regulation of Labor". The Quarterly Journal of Economics, Vol. 119(4), pp. 1339-1382. Bradley, H., J.F. Helliwell and J.R. Livernois. (1981). "Efficient Taxation of Resource Income: The Case of Copper Mining in British Columbiay'.Resources Policy Vol. 7. Brunner, J. (2004). "Politicas de mercado de educaci6n superior: necesidades de informaci6n." In J. Bmnner and P. Meller (eds.). Ofertay demanda deprofesionales y te'cnicosen Chile. El rol de la informacidnptiblica. Santiago,Chile: RIL Editores. Brunner, J., G. Elacqua, A. Tillett, J. Bonnefoy, S. Gonzhlez, P. Pacheco and F. Salazar. (2005). Guiar a1 mercado. Informe sobre la educaci6n superior en Chile. (Processed.) Santiago, Chile. Buscaglia, Edgardo and Maria Dakolias. (1999). Comparative International Study of Court Performance Indicators: A Descriptive and Analytical Account. Legal and Judicial Reform Unit Technical Papers. Washington, DC: The World Bank. Caballero, R. (2002). Coping with Chile's External Vulnerability: A Financial Problem. Cambridge, MA: MassachusettsInstitute of Technology(MIT), Department of Economics. Caballero, R. and A. Galetovic. (2003). Un andisis econdmico de la negativa de Ripley de adherir a la promocidn 'tres'cuotas, sin pie y sin intereses'. Report commissioned by Ripley to be presented before the ComisidnResolutiva, 2003. Caballero, R., K. Cowan, E. Engel, and A. Micco. (2004). Microeconomic Inflexibility and Labor Regulation: International Evidence. NBER Working Paper No. 10744. Cambridge, Massachusetts: National Bureau of Economic Research. Caballero, R., K. Cowan, and J. Kearns. (2004). Fear of Sudden Stops: Lessons from Australia and Chile. Prepared for IDB conference Financial De-dollarization: Policy Options, December 1-2,2003. Caballero, R., E. Engel, and A. Micco. (2004). "Microeconomic Flexibility in Latin America." Economia Chilena, Volume 7 (2): 5-26. 2004. Also: NBER Working Paper No. 10398. Cambridge, Massachusetts:National Bureau of Economic Research. Caballero, R. and M. Hamniour. (1996). "On the Timing and Efficiency of Creative Destruction". QuarterlyJournal of Economics, Vol. 111No. 3. Calvo, G. (1997). "Argentina's Experience alter the Mexican Crisis". In G. Perry (ed.). Currency Boards and External Shocks: How Much Pain, How Much Gain? Latin American And Caribbean Studies.Washington, DC: The World Bank. Calvo, G., and E. Talvi, (2004). Sudden Stop, Financial Factor and Economic Collapse:Lessons from Chile and Argentina. Prepared for the Barcelona Foruni From the Washington Consensus Towards a New Global Governance, September24-26,2004, Barcelona, Spain. Canton, Eric, Debby Lanser, Joelle Noailly, Marieke Rensman, and Jeroen van de Ven. (2004). Science to Industry Knowledge Spillovers; Theoryand International Experiences. Mimeo. Carlton, D. W. (1983). "The Location and Employment Choices of New Firms: An Econometric Model with Discrete and Continuous Endogenous Variables". Review of Economics and Statistics Vol. 65. Caselli,Francesco and SilvanaTenreyro. (2005).Is Poland the next Spain? Mimeo. Clausing, K. (1998). The Impact of Transfer Pricing on IntraJirm Trade. NBER Working Paper No. 6688. Cambridge, Massachusetts:National Bureau of Economic Reseach. Clotfelter, C. T. (1983). "Tax Evasion and Tax Rates: An Analysis of Individual Returns". Review of Economics and Statistics Vol. 65(3). Cohen, W. and D. Levinthal. (1989). "Innovation and learning:The two faces of R&D". Economic Journal, pp. 569-596. Comin, Diego. (2004). R&D: A Small Contribution to Productivity Growth. NBER Working Paper No.10625.Cambridge, Massachusetts:National Bureau of Economic Reseach. Corbo, L. and H. F. Parro. (2005). Institutions, Economic Policies and Growth: Lessons From the Chilean Experience. Central Bank of Chile Working Papers, No. 317,April 2005. Corfo. (2003). La situacidn de la Micro y Pequefia Empresa en Chile. Gobierno de Chile Comite' defoment0 de la Micro y Pequefia Empresa. Mimeo. Couch, J. y J. Barret. (2004). "Alabama's Enterprise Zones: Designed to Aid the Needy?" Public Finance Review. Coughlin, C. C., J. V. Terza and V. Arromdee. (1991). "State Characteristicsand the Location of Foreign Direct Investment within the United States", Review of Economics and Statistics Vol. 73. Cowan Kevin and J. De Gregorio. (2005). International Borrowing, Capital Controls and the Exchange Rate: Lessonsfrom Chile. Central Bank of Chile Working Papers, No. 322, May 2005. Cowan, Kevin, Alejandro Micco and Carmen Pages. (2004). Labor Market Adjustment in Chile. Mimeo. Washington, D.C.:InterarrnericanDevelopnieiit Bank. . (2005). "Labor Market Adjustment in Chile". Economia: The Journal of the Latin American and CaribbeanEconomicsAssociation LACEA,Vol5,pp. 183-218. Cowan, Kevin., A. Micco, A. Mizala, C. PagCs and P. Romagera. (2004). Un Diagnostic0 del Desempleo en Chile. Centro de Micro-Datos. Santiago,Chile: Universidad de Chile. Cowan, Kevin and A. Micco. (2005a). Crecimiento, Empleo y Estabilidad del Ingreso. Documento Mimeograjado. Washington D.C:Banco Interamericanode Desarrollo. . (2005b). The Credit Cycle in Chile: The effect on SME. Mimeo. Washington, DC: Interarnerican DevelopmentBark. Cox, Cristian. (2005). "Las politicas educacionalesde Chile en las ultimas dos dkcadas del siglo XX". In Cox (ed.). Politicas educacionales en el cambio de siglo. Santiago, Chile: EditorialUniversitaria. Crane, S. and F. Nourzad. (1990). "Tax Rates and Tax Evasion: Evidence from California AmnestyData". National TaxJournal Vol. 43(2). Cremer, H., P. Pestieu, and J.C. Rochet. (2001). "Direct Versus Indirect Taxation: The Design of the Tax StructuredRevisited". International Economic Review Vol. 42(3). Crespi, G. (2003). PyME en Chile: nace, crece y.... muere. Analisis de su Desarrollo en 10s ziltimos siete Aios. Fundes International. Cuervo, L. and Mella, C. (1998). Investigacidn Sobre Convergencia Regional. Mimeo. University of Concepcion. De Ferranti, David, Guillenno E. Perry, Indermit Gill, J. Luis Guasch, William E. Maloney, Carolina Sinchez-Pkamo and Norbert Schandy.(2003). Closing the Gap in Education and Technology. World Bank Latin American and Caribbean Studies. Washington DC: The World Bank. De Ferranti, David, Guillenno E. Perry, Francisco H.G. Ferreira and Michael Walton (eds.). (2003).Inequality in Latin America: Breaking WithHistory? Washington, DC: The World Bank. De Gregorio, J. (2004). Economic Growth in Chile: Evidence, Sources and Prospects. Central Bank of Chile Working Paper No. 298. De Neubourg, Chris, Julie Castonguay and Keetie Roelen. (2005). Social Safety Nets and Targeted Social Assistance: Lessonsfrom the European Experience. Washington, DC: The World Bank. Deaton, A. and R. Miller. (1995). International commodity prices, macroeconomic performance andpolitics in Sub-Saharan Africa. Princeton Studies in International Finance, No. 79. DESUP. (2005). Compendia de Educacibn Superior. Santiago, Chile: Ministry of Education. Devarajan, Shantayanan, Vinaya Swaroop and Zou Heng-fu. (1996). "The Composition of Public Expenditure and Economic Growth". Journal of Monetary Economics 16: 313-344. Di Gropello, E. and J. Marshall. (2005). "Teacher Effort and Schooling Outconles in Rural Honduras." in E. Vegas (ed.). Incentives to Improve Teaching: Lessons from Latin America. Washington, DC: The World Bank. Diaz, G., A. Gomez-Lobo and A. Velasco. (2005). "Micros en Santiago: de enemigo public0 a servicio phblico". In A. Galetovic (ed.). Santiago: dbnde estamos y hacia dbnde vamos. Santiago, Chile: Ceiitro de Estudios Publicos. Diaz, R., and P. Meller. (2003). Crecimiento Econbmico Regional En Chile: ~Convergencia? Mimeo. Santiago, Chile: University of Chile. DIPRES (2005). Programas de Empleo con Apoyo Fiscal: Sintesis Ejecutivo. Djankov, S., C. McLiesh, and A. Shleifer. (2004). Private Credit in 129 Countries. NBER Working Paper No. 11078. Cambridge, Massachusetts: National Bureau of Economic Research. Djankov, S., R. La Porta, and F. Lopez-de-Silanes. (2002). "The Regulation of Entry". Quarterly Journal of Economics, Vol. 117, Issue 1. Draghi, Mario, Francesco Giavazzi and Robert C. Merton. (2003). Transparency, Risk Management and International Financial Fragility. NBER Working Paper No. . Easterly, W. and S. Rebelo. (1994). Fiscal Policy and Economic Growth: An Empirical Investigation. NBER Working Paper No. 4499. Cambridge, Massachusetts: National Bureau of Economic Research. Elacqua, G. and R. Fabrega. (2004). El consumidor de la educacibn: el actor olvidado de la libre eleccibn de colegios en Chile.Mimeo. Santiago, Chile :Universidad Adolfo Ibafiez. Erard, B. (1993). "Taxation with Representation: An Analysis of the Role of Tax Practitioners in Tax Compliance", Journal of Public Economics Vol. 52(2). Eyzaguirre, N. (2004). El Chile que Queremos. Presentation Prepared for ENADE 2004, December 2004. Eyzaguirre, N., M. Marcel, J. Rodriguez and M. Tokrnan. (2005). "Hacia la economia del conocimiento: El camino para crecer con equidad en el largo plazo." Estudios Pziblicos No. 97 (Verano). Santiago, Chile: Centro de Estudios Publicos. Feinstein, J.S. (1991). "An Econometric of Tax Evasion and Its Detection". Rand Journal of Economics Vol. 22. Ferrada, C and G. Reinecke. (2004). Creacibn y destruccibn de empleos en Chile: Anhlisis de datos longitudinales del a1 ACHS. Mimeo. Santiago, Chile: Departamento de Economia, Universidad de Chile. Feser, Edward. (2002). The Relevance of Clustersfor Innovation Policy in Latin America and the Caribbean. Mimeo. Fink, Carsten and Keith E. Maskus. (2005). Intellectual Property and Development: Lessons from Recent Economic Research. Oxford UniversityPress and World Bank. Fischer, R. and P. Serra. (2002). "Evaluaci6n de la regulaci6n de las telecomunicaciones en Chile", Perspectivas enpolitics, economiay gestidn Vol. 6, pp.45-77. Forest, A. and S.M. Sheffiin. (2002). "Complexity and Conlpliance: An Empirical Investigation". National TaxJournal Vol. 55(1). Foxley,N. A. (2003). Successes and Failures in Poverty Eradication in Chile. Mimeo. Foxley, Alejandro. (2004). Successes and Failures in Poverty Eradication: Chile. Paper prepared for the Shanghai Conferenceof the World Bank. Fuentes, R., M. Laprain, and K. Schmidt-Hebbel. (2004). Fuentes del Crecimiento y Comportamiento de la Productividad Total de Factores en Chile. Central Bank of Chile Working Paper No. 287. Fuentes, Rodrigo y Ver6nica Mies. 2005. Mirando El Desarrollo Econ6mico de Chile: Una Comparaci6n Iiiternacional. Working Paper No 308. Central Bank of Chile. Santiago, Chile. Fuller, B. and Elmore, R.F. (1996). Who chooses? Who loses? Culture, institutions, and the unequal effects of school choice. New York: Teachers CollegePress. Fundes-Sercotec.(2004). La Situaci6nde la Micro y Pequefia Empresa en Chile. Mimeo. Galasso, E. and M. Ravallion. (2003). Social Protection in a Crisis: Argentina's Plan Jefes y Jefas. Policy Research Working Paper No. 3165.Washington, DC: The World Bank. Galetovic,A. (2005). Competitionpolicy in Chile.Background paper for the DPR, June 2005. Galetovic, A., J. R. Inostroza and C. Mufioz. (2004). "Gas y electricidad: iquC hacer ahora?". Estudios PQblicosVol. 96, 49-106. Galetovic, A. and R. Sanhueza. (1999). Precios de reventa y el mercado de 10s combustibles en la XTIRegidn. Report commissionedby Copecpresented before the Comisidn Resolutiva. . (2005). Concentracidn y precios: la relacidn no es como le han dicho. Class notes, Universidadde 10sAndes. Galindo, A. and A. Micco. (2004a). Bank Credit to Small and Medium Sized Enterprises: The Role of Creditor Protection. Research Department. Washington, D.C: Interamerican Development Bank. . (2004b). Creditor Protection and Financial Cycles. Working Paper, Number 443. Washington, DC: The World Bank. Gallego, F. (2002). "Competencia y resultados educativos: teoria y evidencia para Chile", Cuadernos de Economia 39 (118): 309-352. Gallego, F. and N. Loayza. (2002). The Golden Period for Growth in Chile: Explanations and Forecasts. Central Bank of Chile Working Paper No. 146. Gasmi, F., J.J. Laffont and W. Sharkey. (2002). "The Natural Monopoly Test Reconsidered:An Engineering Process-Based Approach to Empirical Analysis in Telecommunications". International Journal of Industrial OrganizationVol. 20,435-459. Gasparini, L. (2003). Empleo y proteccidn social en Ame'rica Latina. Un andisis sobre la base de encuestas de hogares. OIT. Gasparini, L. (2004). Pro-poor and Progressive Growth in the Southern Cone: Evidence from household surveys. Mimeo. Centro de Estudios Distributivos, Laborales y Sociales, Universidad Nacional de La Plata. Gasparini, L., M. Cicowiez, V. Fazio, G. Pizzolito and H. Winkler. (2004). Meeting the Poverty- Reduction MDG in the Southern Cone. Part of a project on Monitoring the Socio- Economic Conditions in Argentina, Chile, Paraguay and Uruguay. Mimeo. Centre de Estudios Distributivos, Laborales y Sociales (CEDLAS). Universidad Nacional de La Plata. Gasparini, L. and F.M. Bertranou. (2005). "Social Protection and the Labour Market in Latin America: What can be learned from Household Surveys?" International Social Security Review, Vo1.58, No. 2-3. Gazmuri, C. (2004). La justicia del Trabajo en Chile: Realidad y Perspectivas. Cuadernos de Investigaci6n No. 21. Direcci6n del Trabajo. Gellhorn, E. and W. Kovacic. (1994). Antitrust Law and Economics in a Nutshell. Saint Paul: West Group. Gentry, W.M. and H. F. Ladd. (1994). "State Tax Structure and Multiple Policy Objectives". National TaxJournal Vol. 47(4). Gentry, W. and G. Hubbard. (2000). TaxPolicy and Enty intoEntrepreneurship. Mimeo. Giavazzi, F. and R. Perotti (2005). Chile's Fiscal Rule. Background paper for the DPR. Gill, Indermit S. and Claudio E. Montenegro. (2002). "Responding to Earnings Differentials in Chile". In: Indermit S. Gill, Claudio E. Montenegro and Dorte Domeland (eds.). Crafting Labor Policy Techniques and Lessons from Latin America. New York: World Bank and Oxford UniversityPress. Gill, Indermit, Truman Packard and Juan Yenno. (2004). Keeping the Promise of Social Security in Latin America. Stanford UniversityPress. Giovagnoli, P., G. Pizzolito and J. Trias. (2004). Chile. Part of a project on Monitoring the Socio-Economic Conditions in Argentina, Chile, Paraguay and Uruguay. Mimeo. Centre de Estudios Distrib~~tivos,Laborales y Sociales (CEDLAS), Universidad Nacional de La Plata. Gallego, and N. Loayza, 2003. Gonzilez, P. (2004). "Una mirada econ6mica de las politicas y de las necesidades de informaci6n en educaci6n superior." In J. Brunner and P. Meller (eds.). Ofertay demanda de profesionales y te'cnicos en Chile. El rol de la informacidn ptiblica. Santiago, Chile: RIL Editores. Gonzalez, P. Mizala, A. y P. Romaguera. (2004). Vouchers, inequality and the Chilean experience. Docurnento de Trabajo CEANo. 150.Santiago,Chile: Universidad de Chile. Gordon, R.H. (1992), "Can Capital Income Taxes Survive in Open Ecoiioniies?" Journal of Finance Vol. 47. Gordon, R.H. and J. Mackie-Mason. (1994). WhyIs There Corporate Taxation in a Small Open Economy? The Role of Transfer Pricing and Income Shifting. NBER Working Paper No. 4690. Gordon, R.H. and S.B. Nielseii. (1997). "Tax Evasion in an Open Economy: Value-Added vs. Inconie Taxation". Journal of Public Economics Vol. 66. Gordon, R.H. (1998). Can high Personal Tax Rates Encourage Entrepreneurial Activity? IMF Staff Papers. Washington,DC: International Monetary Fund. Gruber, Jonathan. (1995). The Incidence of Payroll Taxation: Evidence from Chile. NBER Working Paper No. 5053. Grubert, H. (1998). "Taxes and the Division of Foreign Operating Income among Royalties, Interest, Dividends, and Retained Earnings". Journal of Public Economics Vol. 68(2). Grubert, H. and J. Mutti. (1991). "Taxes, Tariffs, and Transfer Pricing in Multinational CorporateDecision Making". Review of Economics and Statistics Vol. 73(2). Grubert, H., T. Goodspeed and D. Swenson. (1993). "Explaining the Low Taxable Income of Foreign-Controlled Companies in the United States". In: A. Giovannini, G. Hubbard, and J. Slemrod (eds.). Studies in International Taxation.Universityof Chicago Press. Hall, Bronwyn H. (2005). GovernmentPolicyfor Innovation in Latin America. Mimeo. Harmon, O.R. and R. Mallick. (1994). "The Optimal State Tax Portfolio Model: An Extension". National TaxJournal Vol. 47. Heckman, J. (1999). Policies that Foster Human Capital. NBER Working Paper No. 7288, Cambridge,Massachusetts:National Bureau of EconomicResearch. Heckman, J. and Pages-Serra, C. (2000). "The Cost of Job Security Regulation: Evidence from Latin American Labor Markets." Economia, 2000, 1(1), 109-44. Hines, J. R. (1996). "Altered States: Taxes and the Location of Foreign Direct Investment in America". American Economic Review Vol. 86. (1999). "Lessons from Behavioral Responses to International Taxation". National TaxJournal. Hines, J.R. and G. Hubbard. (1990). "Coming Home to America: Dividend Repatriation by U.S. Multinationals". In: A. Razin and J. Slernrod (eds.). Taxation in the Global Economy. Chicago UniversityPress. Hines, J.R. and E.M. Rice. (1994). "Fiscal Paradise: Foreign Tax Havens and American Business". QuarterlyJournal of Economics Vol. 109(1). Holzmann, Robert, Truman Packard and Jose Cuesta. (2000). "Extending Coverage in Multi Pillar Pension Systems: Constraints and Hypotheses, Preliminary Evidence and Future Research Agenda". In: R. Holzmann and J. Stiglitz (eds.). New Ideas About Old Age Security. Washington, DC: The World Bank. Honahan, Patrick. 2004. Alternative Approaches to Taxing the Financial Sector; Which is Best and Where Does Chile Stand? in ed. Luis Antonio Ahumada and J. Rodrigo Fuentes, Banking Market Structure and Monetary Policy. CentralBank of Chile. Santiago,Chile Hopenhayn, H. and R. Rogerson. (1993). "Job Turnover and Policy Evaluations: A General Equilibrium Analysis". Journal of Political Economy Vol. 101:915-38. Hoxby, C. (2000a). Peer Effects in the Classroom: Learningfrom Gender and Race Variation. NBER Working Paper 7867, Cambridge, MA: National Bureau of Economic Research. . (2000b). Would School Choice Change the Teaching Profession? Mimeo. Cambridge, MA: Harvard University. Hsieh, C. and M. Urquiola. (2003). When Schools Compete, How Do They Compete? An Assessment of Chile's Nationwide School VoucherProgram. NBER Working Paper 10008, Cambridge,MA: National Bureau of Economic Research. International Labor Organization. (2004). El Sistema de Pensiones en Chile en el Contexto Mundial y de America Latina, Ponencias del Seminario Internacional. Santiago,Chile. .(2003).Panorama Laboral. Oficina Intarnacionaldel Trabajo. Lima, Peru. IADB. (2003).Economic and Social Progress Report 2004 (IPES 2004): Labor Markets in Latin America. Washington, DC: Interamencan Development Bank. .(2004). GoodJobs Wanted.Washington, DC: InteramericanDevelopmentBank. ICARE. (2004).Indice de Confianza. Noviembre 2004. IMF. (2004). Chile:Financial System Stability Assessment, including Reports on the Observance of Standards and Codes on the Following Topics: Monetary and Financial Policy Transparency, Banking Supervision, and Securities Regulation. Country Report No. 041269.Washington,DC: InternationalMonetary Fund. IMF. 2004. Chile--Financial Sector Assessment Program. International Monetary Fund and The World Bank. Washington, DC: InternationalMonetary Fund. -. (2005). Chile: Selected Issues. CountryReport No. 05/316. Waslington, DC: International Monetary Fund. IMD.(2003). World Competitiveness Yearbook.Lausanne, Switzerland:International Institute for ManagementDevelopment. Jadresic, A., V. Blanlot and G. San Martin. (2001). La nueva regulacidn. Santiago, Chile: Dolmen. Kaiser, Robert G. (2005). "Focus on Schools Helps Finns Build a Showcase Nation." The WashingtonPost, May 24. Kamps, Christophe. (2004). The Dynamic Macroeconomic Effects of Public Capital: Theory and Evidence for OECD Countries. Kiel Studies Vol. 331. Heidelberg and New York: Springer. Karsten, S. (1999). "Neoliberal Education Reform in the Netherlands." Comparative Education 35 (3). Kaufmann, D., A. Kraay and M. Mastruzzi. (2003). Governance Matters III: Governance Indicatorsfor 1996-2002. Policy Research Working Paper No 3106.Washington, DC: The World Bank. . (2004). Governance Matters IV: Governance Indicators for 1996-2004. Policy Research WorkingPaper No 3630. Washington, DC: The World Bank. Kesselman, J. R. (1993). "Evasion Effects of Changing the Tax Mix". The Economic Record Vol. 69(205). Klapper, L., L. Laeven, and R. Rajan. (2004). Business environment andJirm entry: Evidence from international data. Working Paper Number No. 3232. Washington, DC: The World Bank. Klenow, P. and Andres Rodriguez-Claire.(2004).Externalities and Growth.Mimeo. Kloprogge, J. (2003). Educational Policies and Equal Opportunities in the Netherlands. Recent Developments in Policiesfor Children at Risk. Mimeo. Lagos, Ricardo. (2005). Discourse on the State of the Nation. May 21St, 2005 (see http://www.~obiernodechile.cl/2lmayo2004/indice discursos.asp). Landau, D. (1983). "Government Expenditure and Economic Growth: A Cross-Country Studyyy. Southern Economic Journal of Monetary Economics 16:141-164. Larrain, G. (2004). Sistemaprevisional chileno: mcis informacidnpara un mejor ancilisis. Power point presentation.Available in www.safp.cl. Lederman, D. and W. Maloney. (2003). R & D and Development. Mimeo. Lee, Y. and R. Gordon. (2003). Tax Structure and Economic Growth.Mimeo. Levine, R. and D. Renelt. (1992). "A Sensitivity Analysis of Cross-Country Growth Regressions". American Economic Review Vol. 82(4). Lindert, Peter H. (2005). Growing Public: Is the Welfare State Mortal or Exportable? DEC Lectures. Washington D.C.: The World Bank. Link, Albert and John Scott. (2004). The Role of Public Research Institutions in a National Innovation System:An Economic Perspective. Mimeo. Litchfield, J., 2002. Poverty and Income Inequality in Chile, 1987-2002. Background paper for Household Risk Management and SocialProtection. Loayza, N., P. Fajnzylber, and C. Calder6n. (2004).Economic Growth in Latin America and the Caribbean: Stylized Facts, Explanations, and Forecasts. Central Bank of Chile Working Paper No. 265. Santiago, Chile. Long, J.E. and J.D. Gwartney. (1987). "Income Tax Avoidance: Evidence from Individual Tax Returns". National Tax Journal Vol. 50(4). Luger, M. and S. Shetty. (1985). "Determinants of Foreign Plant Start-Ups in the United States: Lessons for Policymakers in the Southeast". Vanderbilt Journal of Transnational Law, Spring. Maloney, William. (2002). "Missed Opportunities: Innovation and Resource Based Growth in Latin America". Economia, 3(1), Fall 2002. . (2005).Patterns of Innovation. Mimeo. Maloney, William and Jairo Nufiez. (2001).Measuring the Impact of Minimum Wages:Evidence from Latin America. Working Paper. Washington,DC: The World Bank. . (2004). "Measuring the Impact of Minimum Wages, Evidence from Latin America." In: James Heckrnan and Carmen Pagks (eds.). Law and Labor Markets. Chicago:University of Chicago Press. Maloney, W. F. and A. Rodriguez-Clare. (2005). IdentzJjling Innovation Shortfalls in LDCs. Mimeo. Mani, Sunil and Anthony Bartzokas. (2002). Institutional Support for Investment in New Technologies: the Role of Venture Capital Institutions in Developing Countries. INTECH Discussion Paper Series.United Nations University. Marcel, M. (2004). Oportunidades y desafos para la politica social en Chile. Presentacibn Seminario "Enfientando 10s desafios del desarrollo en Chile: Los sectores sociales" MinisterioHacienda-BancoMundial. Santiago,Chile. Marimon, Ramon. (2002).Innovation in Spain: Past and Future Challenges. Mimeo. Marinovic, Ivan and Salvador Valdes. (2004). Economias de Escala en la Industria AFP. Mimeo. Santiago,Chile. Maskus, Keith E. (2000). Intellectual Property Rights in the Global Economy. Washington, DC: Institute for InternationalEconomics. McEwan, P. and Carnoy, M. (1999). The impact of competition on public school quality: longitudinal evidence from Chile's voucher system. Mimeo. Palo Alto, CA: Stanford University. . (2000). "The effectiveness and efficiency of private schools in Chile's voucher system". Educational Evaluation and Policy Analysis 33:213-239. Meller, P. (1999). Pobreza y Distribucidn del Ingreso en Chile (De'cada del 90). Serie de Economia No. 69. Santiago, Chile: Departamento de Ingenieria Industrial, Universidad de Chile. Meller, P. and A. Tokman. (1996). "Apertura Comercial y Diferencial Slaralial en Chile". In P. Meller (ed.).El Modelo Exportador Chileno. Crecimientoy Equalidad. Meltzer, A. and Richard Scott. (1981). "A Rational Theory of the Size of Government". Journal of Political Economy 89 (5): 914-27. Messere, K.C. (1993). Tax Policy in OECD Countries: Choices and Conflicts. The Netherlands: BFD Publications. Micco, A. (2005). Microeconomic Flexibility and Firms ' Behavior in Chile. Background paper for the DPR. Micco, A. and C. Pagks (2004). Employment Protection and Gross Job Flows: A DzfSevences in Differences Approach. Working Paper No. 508. Washington, DC: Interamerican DevelopnlentBank. MIDEPLAN. (2005). Informe de Seguimiento y Monitoreo, Sistema de Proteccion Social Chile Solidario. Santiago,Chile: SecretariaEjecutiva Chile Solidario. Ministerio de Trabajo y Prevesion Social. (2004). Analisis y Principales Resultados, Primera Encuesta de Proteccion Social (Historia Laboral y Seguridad Social 2002). Santiago, Chile: Ministerio de Trabajo y Prevision Social. Miodosky, M. and E. Siimott. (2005). The Impact of Chile's Free Trade Agreements on the Import-Competing Agricultural Sub sectors. Mimeo. Washington, DC: The World Bank. Mizala, A. and P. Romaguera. (2001). "Factores explicativos de 10s resultados escolares en la educacion secundariaen Chile." TrimestreEcondmico 272 (October-December):515-549. Mizala A. and P. Romaguera. (2005a). "Teachers' Salary Structure and Incentives in Chile." In E. Vegas (ed.). Incentives to Improve Teaching: Lessonsfrom Latin America. Directions in Development Series. Washington, DC: The World Bank. . (2005b). Calidad de la educacidn chilena: una asignatura pendiente. Mimeo. Santiago, Chile: Ui~iversidadde Chile. Mizala, A., P. Romaguera and P. Henriquez. (1999). Female Labor Supply in Chile. Mimeo. Santiago,Chile: Universidad de Chile. Montenegro, Claudio E. (2001). Wage Distribution in Chile: Does Gender Matter? A Quantile Regression Approach. Policy Research Report on Gender and Development. Working Paper Series No. 20. Washington, DC: The World Bank. Available on line at: www.worldbank.org. . (2002). Unemployment, Job Security, and Minimum Wages in Chile: 1960 to 2001. Background Paper for Chile: Household Risk Management and Social Protection. Washington, DC: The World Bank. Montenegro, C., and C. Pagks (2003). Who Benefits from Labor Market Regulations? Chile 1960-1998.Policy Research Working Paper No. 3143.Washington, DC: The World Bank. Montenegro, Claudio E. and C. Pages. (Forthcoming). "Who Benefits from Labor Market Regulations? Chile 1960-1998." In James Heckman and Carmen Pages (eds.). Law and Employment: Lessons from the Latin America and the Caribbean. Cambridge, Mass. and Chicago, IL.: National Bureau of EconomicResearch and University of Chicago. Montero, J.P. and H. Rudnick. (2001). "Precios elkctricos flexibles". Cuadernos de Economia Vol. 38,91-109,2001. Morandk, F., R. Soto and P. Pincheira. (1997). "Achilles, the Tortoise, and Regional Growth in Chile". In F. MorandC and R. Vegara (eds.). Ancilisis Empirico del Crecimiento en Chile. CEP -ILADESIGeorgetownUniversity. Mullin. (2004). Refections on UniversityPoliciesfor their R&D Linkages with Firms -A Survey on Chile. Ontario:Mullin Consulting Ltd. Mullin Associates. (2004). UniversityPolicies in Latin Americafor the Promotion of RcW Links with Firms: A Five CountryAnalysis. Mimeo. NSF. (2002). Science and Engineering Indicators 2002. Washington, DC: National Science Foundation. OECD. (1996). Employment Outlook 1996. Paris: Organization for Economic Cooperation and Development. . (2000). Literacy in the Information Age - Final Report of the International Adult Literacy Survey. Paris: Organization for Economic Cooperation and Development. . (2003a). Chile-Country Survey. Paris: Organization for Economic Cooperation and Development. . (2003b). Literacy Skills for the World of Tomorrow Further Resultsfrom PISA 2000. - Paris: Organization for Economic Cooperation and Development. . (2004a). Chile, Review of National Policies for Education. Paris: Organization for Economic Cooperation and Development. . (2004b). Education at a Glance. Centre for Education Research and Innovation. Paris: Organization for Economic Cooperation and Development* . (2004~).Employment Outlook 2004. Paris: Organization for Economic Cooperation and Development. . (2004d). Science, Technology and Industry Outlook. Paris: Organization for Economic Cooperation and Development. . (2004e). Taxing Wages. Paris: Organization for Economic Cooperation and Development. . (2005). Pensions at a Glance: Policies Across OECD Countries. Paris: Organization for Economic Cooperation and Development. Oyarzun, C. and I. Araya. (2001). "Long-Run Dynamics of Regional Growth in Chile". Estudios de Economia, 28(1), 69-78. Packard, Truman. (2002a). Pooling, Saving and Prevention: Mitigating Old Age Poverty in Chile.Policy Research Working Paper. Washington, DC: The World Bank. . (2002b). Do Workers in Chile Choose Informal Employment? A Dynamic Analysis of Sector Choice. Background Paper for Chile: Household Risk Management and Social Protection. Washington, DC: The World Bank. . (2001). Is There a Positive Incentive Effect from Privatizing Social Security? Evidence from Pension Reforms in Latin America. Policy Research Working Paper. Washington, DC: The World Bank. Pagks, Carmen and Claudio E. Montenegro. (1999). Job Security and the Age Composition of Employment: Evidence from Chile. Working Paper No. 398. Washington, DC: Interamerican Development Bank. Available on line at: www.iadb.org. Papke, L.E. (1991a). "Interstate Business Tax Differentials and New Firm Location. Evidence from Panel Data". Journal of Public Economics Vol. 45. . (1991b). "What Do we Know about Enterprise Zones?" Tax Policy and the Economy. Paredes, Ricardo and Augusto Iglesias. (2004). Analisis de propuestas para aumentar la cobertura de trabajadores independientes en el sistema AFP. Mimeo presented at the seminar "Competencia y Coberturayyhosted the by Centro de Estudios Publicos (CEP) and the Superintendencyof the AFP system (SAFP), 11-12November, Santiago,Chile. Parente, S.L. and E. Prescott. (1994). "Barriers to Technology Adoption and Development". Journal of Political Economy, Vo1.102: 298-321. Patrinos, H. (2003). Private Education Provision and Public Finance: The Netherlands as a Possible Model. Mimeo. Washington, DC: The World Bank. Peters, A. and P. Fisher. (2002). State Enterprise Zone Programs: Have they Worked? Upjohn Institute for Employment Research. Pizzolito, G., P. Giovagnoli and J. Trias. (2004). Poverty and Inequality in Chile: Methodological Issues and a Literature Review. Part of a project on Monitoring the Socio- Economic Conditions in Argentina, Chile, Paraguay and Uruguay. Mimeo. Centre de Estudios Distributivos, Laborales y Sociales (CEDLAS), Universidad Nacional de La Plata. PRIESO. (2000). Encuesta de Prevision de Riesgos Sociales. Social Risk Management Survey. Santiago,Chile. Raczynski, Dagrnar, Claudia Serranoand Manuela Valle. (2002).Eventos de Quiebre de Ingreso y Mecanismos de Proteccidn Social: Estudio en Hogares de Ingreso Medio y Bajo. Background Paper for Chile: Household Risk Management and Social Protection. Washington,DC: The World Bank. Ravallion, Martin. (2000).Are the Poor Protectedfrom Budget Cuts: Theory and Evidencefiom Argentina. Washington, DC: The World Bank. Ray, E. J. (1977). "Foreign Direct Investment in Manufacturing". Journal of Political Economy Vol. 85. Rego, S.O. (2003). "Tax-Avoidance Activities of U.S. Multinational Corporations". ContemporaryAccounting Research Vol. 20(4). Reineck, Gerhard. (2005.) "Income Protection Through Direct Employment Programmes:Recent Concepts and Examples from Latin Americayy.International Social Security Review Vol. 58(2-3). Reinecke, R. and Ferrada, C. (2005). Creacidn y destruccidn de empleo en Chile: Andisis de datos longitudinales de la ACHS. Mineo. Santiago, Chile: Asociaci6n Chilena de Seguridad,Oficina Intemacional del Trabajo. RICYT. (2005). La Red Iberoamericana de Indicadores de Cienciay Tecnologia. Buenos Aires: www.ricvt.edu.ar Rodriguez, Jorge. (1988). "School Achievement and DecentralizationPolicy: The Chilean Case." Revista de Anblisis Econdmico, Vol. 3, No. 1,pp. 75-88. Rodriquez, Jorge and Marcelo Tokrnan. (2000). "Resultados y rendimiento del gasto en el sector public0 de salud en Chile 1990-1999". Serie Financiamiento del Desarrollo 106.CEPAL, United Nations. Rodriguez-Clare, Andres. (2004).Microeconomic Interventionsafier the Washington Consensus. Mimeo. Rofman, Rafael and Eliana Carranza. (2005). La Cobertwa de la Seguridad Social en America Latina. Mimeo. SAFP. (1999). El Sistema Chileno de Pensiones. Santiago, Chile: Superintendencia de Administradoras de Fondos de Pensiones. Sanguinetti, Pablo. (2004). Innovation and R&D Expenditures in Argentina: Evidence from a Firm Level Suwey. Mimeo. Sapelli, C (2003). Chile: Effects of Old Age on Health Sewices utilization and Consequences on System Design. Santiago, Chile: Pontifica Universidad Catolica de Chile, Instituto de Economia. Sapelli, C. and B. Vial. (2002). "The performance of private and public schools in the Chilean voucher system" Cuadernos de Economia 39 (118):423-454. Sarria-Allende, V. and R. Fisman. (2004). Regulation of Entry and the Distortion of Industrial Organization.NBER Working Paper No. 10895. Sawada, Y. and A. Ragatz. (Forthcoming.) "Decentralization of Education, Teacher Behavior and Outcomes: The Case of El Salvador's EDUCO Program." In E. Vegas (ed.). Incentives to Improve Teaching:Lessons from Latin America. Washington, DC: The World Bank. Slernrod, J. (1995). m a t Do Cross-Country Studies Teach about Government Involvement, Prosperity, and Economic Growth? Brookings Papers on Economic Activity No. 2. Washington, DC: The Brookings Institution. Spengler, J. (1950). "Vertical Integration and Antitrust Policy7'.Journal of Political Economy Vol. 58,347-352. Soto, R. and A. Torche. (2004). Spatial Inequality After Reforms in Chile: Where Do We Stand? Mimeo. Sutton, J. (1991).Sunk Costs and Market Structure. Cambridge, MA: MIT Press. Swenson, D. (1994). "The Impact of U.S. Tax Reform on Foreign Direct Investment in the United States". Journal of Public Economics Vol. 54(2). Syverson, C. (2004). Market Structure and Productivity: A Concrete Example, Journal of Political Economy Vol. 112, 1181-1222. Szekely, M. and M. Hilgert. (1999). The 1990s in Latin America: Another Decade of Persistent Inequality. IDB-OCE Working Paper No. 410. Washington, DC: Interamerican Development Bank. Tan, Hong. (2005). Policiesfor increasing SME absorptive capacity. Mimeo. Tanzi, V. and H.H. Zee. (2000). "Tax Policy for Emerging Markets: Developing Countries". National Tax Journal Vol. 53(2). Tarzijan, Jorge. (2004). Evaluation de la Autorizacion de Planes de Precio por Permanencia a las AFP. Mimeo. Chile, Santiago: Catholic University. Thorn, Kristian, Lauritz Holm-Nielsen, Daniel Humrnel, and Maarja Soo. (2005). Science & Technology in Latin America: Policy options and lessons learned. Mimeo. Thorn, Kristian., L. Holm-Nielsen,and J. Jeppesen. (2004).Approaches to results-basedfunding in tertiary education: identzbing finance reform options for Chile. Policy Research Working Paper No. 3436. Washington, DC: The World Bank. Tokman, A. (2004). Education y Crecimiento en Chile. Mimeo. Santiago, Chile: Banco Central de Chile. .(2005). "Diagnbsticos y propuestaspara la educacibn chilena". Economia Chilena Vol. 8 (1). Tokman, M., and A. Zahler. (2004). Innovation para un Crecimiento sostenido: Siete Lecciones para Chile. Mimeo. Tokman, V. (ed.). (1992). Beyond Regulation: The Informal Economy in Latin America. World Employment Program, ILO. Torche, F. (2005). "Unequal But Fluid: Social Mobility in Chile in Comparative Perspective". American Sociological Review Vol. 70, pp.422450. Torche, Aristides and Gert Wagner. (1997). Previsidn Social: Valoracidn Individual de un Beneficio Mandatado. Cuadernos de Economia No. 103.PUCC. Traverso,A. (2004). "Las nuevas reformas que la reforma engendra." En Foco 26, Expansiva. Valdks, Salvador (2002). Social Security Coverage In Chile, 1990-2001. Background Paper for Regional Study on Social Security Reform. Office of the Chief Economist, Latin America and Caribbean Regional Office.,Washington, DC: The World Bank. . (2005). "Para aumentar la competenciaentre las AFP". Estudios Pziblicos 98, 87-142. Valdks, S. and I. Marinovic. (2004). Contabilidad regulatoria: las AFP chilenas, 1993- 2003.Working Paper. Santiago,Chile: Instit~~tode Economia, Universidad Catblica. Vegas, E. and I. Umansky. (2005). Improving Teaching and Learning llrough Effective Incentives: m a t Can We Learn from Education Reforms in Latin America? Washington, DC: The World Bank. Venables, Anthony J. (2004). Technology,Innovation and Trade. Mimeo. Venturelli, A. (2003). Una estimacidn del efecto distributivo de las bandas de precio, 1984- 2000. Documento de Trabajo No. 162. Santiago, Chile: Centro de Economia Aplicada, Universidad de Chile. Vodopivec, M. (2004).Income Supportfor the Unemployed. Washington,DC: The World Bank. Vroman, Wayne. (2002). Unemployment Protection in Chile. Background Paper for Chile: Household Risk Management and SocialProtection. Washington, DC: The World Bank. Wasylenko, M. (1997). "Taxation and Economic Development: The State of theEconomic Literature". New England Economic Review. Weiss, L. (1989). Concentration and Price. Cambridge, MA: MIT Press. Wisecarver, D. (1985). "Regulacibn and derregulacibn en Chile 1973-1983". Estudios Pziblicos Vol. 22, 115-167. Whish, R. (2001). CompetitionLaw. Fourth edition.London:Butterworths. Woodward, D. (1992). "Locational Determinants of Japanese Manufacturiilg Start-ups in the United States". Southern Economic Journal Vol. 58(3). White, F.C. (1983). "Trade-Off in Growth and Stability in State Taxes". National Tax Journal Vol. 36. Whitehouse, E.R. (forthcoming).Guide to World Pension Systems. Washington, DC: The World Bank. World Bank. (1997). Chile:Poverty and Income Distribution in a High-Growth Economy: 1987- 1985. Report No. 16377-CH. Washington, DC: The World Bank. . (2002). Chile: Millennium Science Initiative - Implementation Completion Report. Washington, DC: The World Bank. . (2003). WorldDevelopment Indicators. Washington DC: The World Bank. . (2004a). Chile. A Strategy to Promote Innovative Small and Medium Enterprises. Report No. 291140 CL. . (2004b). Doing Business in 2004 - Understanding Regulation. Washington, DC: The World Bank. . (2004~).Household Risk Management and Social Protection in Chile. Washington, DC: The World Bank. . (2005a). Doing Business 2005 -- Removing Obstacles-toGrowth. Washington DC: The World Bank, International Finance Corporation and Oxford University Press, 2005. . (2005b). Investment Online Climate Survey. http://iresearch.worldbank.org/ics/isp/index.isp World Bank Institute. (2005). Knowledge Assessment Methodology. http://info.worldbank.org/etools/kan?2005/ Young, K.H. (1988). "The Effect of Taxes and Rates of Return 011Foreign Direct Investment in the United States". National Tax Journal Vol. 41(1).