Report No. 19920-EC Ecuador Crisis, Poverty and Social Services (InTwoVolumes)Volumell: Annexes June 26, 2000 Human Development Department Latin America and the Caribbean Region Document of the World Bank EXCHANGE RATE (1999 average) Sucres 1 1,786.8 = US$1.00 S1,000 = US$0.09 SCHOOL YEAR Coast: April - January Inland: October - July FISCAL YEAR January 1 - December 31 ACRONYMS AND ABBREVIATIONS BCE Banco Central del Ecuador PUCE Pontificia Universidad Cat6lica del Ecuador BMI Bono Matemo Infantil CAF Corporaci6n Andina de Fomento CEM Centros Educativos Matriz (School Network) CEPAL Comisi6n Econ6mica pam Latinoamerica y el Caribe, (Economic Commission for Latin America and the Caribbean) CEPLAKS Centro de Planificaci6n y Estudios Sociales CONAMU Consejo Nacional de la Mujer COPEFEN Unidad Coordinadora del Fen6meno El Niflo DDSB. Debt and Debt Service Reduction DINIEB Direcci6n Nacional de Educaci6n Intercultural Bilingfle DL Decentralization and Social Participation Law (1997) EB/PRODEC Educaci6n Basica - Proyecto de Educaci6n ECD Early Childhood Development ECV Encuesta Condiciones de Vida (LSMS) FASBASE Proyecto de fortalecinmiento y Ampliaci6n de Servicios Basicos de Salud en Ecuador, (Basic Primary Health Project) FGT Foster - Greer- Thorbecke FISK Fondo de Inversi6n de Emergencia, (Emergency Social Fund) GDP Gross Domestic Product IDB Inter-American Development Bank, (Banco Interamericano de Desarrollo) IESS Instituto Ecuatoriano de Seguridad Social, (Ecuadoran Social Security Institute) INEC Instituto Nacional de Estadistica y Censo, (National lnsfitutefor Statistics and Census) INECEL Instituto Ecuatoriano de Electrificaci6n, (Ecuadoran Electricity Company) INNFA Instituto Nacional de la Juventud y de la Famnilia, (Nafional Institute of the Youth and the Family) JBG Junta de Beneficiencia de Guayaquil Vice-President David De Ferranti Country Director Andrds Solimano Sector Leader: Constance A. Corbett Sector Director Xavier Coi Task Manager: Donald Winkler ISMS Living Standard Measurement Survey (ECV) M & PL Law of the Modernization of the State, Privatization, and Delivery of Public Services by the Private Sectors (1993) MEC Ministerio de Educaci6n y Cultua, (Minisny of Education and Culture) MHEB Municipal Health Board MHCN Municipal Integrated Health Care Networks MODERSA Modernizaci6n Gesti6n Hospitalaria y Reforma Sector Salud MOE MSP Ministerio de Salud Publica, (Ministry ofPublic Health) MSB Ministry of Social Welfare MTI Medium Term Improvement NGO Non-Governmental Organizations, (Organizaciones No-Gubernamentales) OECD Organization for Economic Co-operation and Development OREALC Oficina Regional de la Educaci6n para Amdrica Latina y el Caribe ORI Operaci6n de Rescate Infantil, (Operation Child Rescue) PACMI Programa de Alimentacion Complementario Materno Infantil PAHO/WHO Organizaci6n Panamericana de la Salud, (World Health Organization) PCE Gasto per Capita, (Per Capita Expenditures) 'PDI Programa Desarrollo Infantil ?PP Purchasing Power Parity PRAF Programa de Asistencia Familiar, (FamilyAssistant Program) PROMECEB Proyecto de Mejoramiento de Calidad y Eficiencia de la Educaci6n Bdsica, (Projectfor Improvement ofEfficiency and Quality of Basic Education (1DB-Financed) PRONEPE Programa Nacional de Educaci6n Preescolar Alternativa SEDES Secretaria de Desarroilo Social SEGEPLAN Secretaria General de Planificaci6n, (General Planning Secretariat) SIMUJER Sistema de Indicadores sobre la Situaci6n de las Mujeres y de las Diferencias entre Hombres y Mujeres en el Pais s.InSE Proyecto de Sistema Integrado de Indicadores Sociales del Ecuador SISVAN National System of Nutritional Vigilance SOLCA Sociedad de Lucha contra el CAncer SSFA Seguro Social de la Policia, (Police Social Securty) STD Standard TEMSS Third International Mathematics and Science Study TQM Total Quality Management UNDP Programa de Desarrollo de las Naciones Unidas, (United Nations Developnent Program) USAID United States Agency for Intemational Development VAT Value Added Tax, (Impuesto de ValorAgregado) ANNEXES Table of Contents Annex 1: Statistical Annex Annex 2: Methodology Used for Deriving Estimates of Poverty Annex 3: Protecting the Poor in Ecuador: Priorities and Options for the Bono Solidario Annex 4: An Emerging Education Strategy for Ecuador Annex 5: Health in Ecuador Annex 6: Incidence Analysis of Social Expenditures and Public Subsidies in Ecuador Annex 7: Employment Generation in Ecuador Annex 8: Indigenous Education in Ecuador Annex 1: Statistical Annex Adela Garcia Aracil, M6nica Sarmiento, Renee Hendley 1999 Index of Tables Table Al. Distribution and Incidence of Poverty in 1998, by Geographic Area Table A2. Distribution and Incidence of Poverty in 1998, by Consumption Quintiles (Percentages) Table A3. Household Characteristics by Consumption Quintiles and Poverty Status Table A4. Mean Total Consumption Expenditures per Capita and per Household in Ecuador, 1998 Table A5. Age Distribution of Households in Ecuador, 1998. Percent of Household With Children in Both Age Groups Table A6. Population According to Sex, Disease Symptoms and Age Groups by Consumption Quintiles Table A7. Type of Health Providers Attending to Individuals with Disease Symptoms by Consumption Quintiles Table Ag. Enrollment of Children Less Than 6 Years Old in Daycare and Preschool. Table A9. Education Enrollment Rates Table A10. Distribution of Education Enrollments by School Type (Percentages) Table A1l. Mothly Education Expenditures by Households with Children between 6 and 17 Years Old in Public Schools Table A12. Monthly Education Expenditures by Households with Children between 6 and 17 Years Old in Public Schools Table A13. Monthly Education Expenditures by Households with Children between 6 and 17 Years Old in Private Schools Table A14. Monthly Education Expenditures by Households with Children between 6 and 17 Years Old in Public and Private Schools 'Table Al5. Self reported Number of Days of School Missed Monthly Table A16. Children's Economic and Educational Activity. Age Group 6-l1 (Percentages) Table A17. Children's Economic and Educational Activity. Age Group 12-14 (Percentages) Table Al8. Children's Economic and Educational Activity. Age Group 15-17 (Percentages) Table A19. Average Salary per Hour and Monthly Age Groups Table A20. Average Hours Worked per Week Central Government Expenditure in the Social Sectors Table A21. Summary of total Central Govermnent Expenditures (Real 1995 Sucres) Table A22. Social Sectors as a Percentage of GDP Table A23. Central Govemment Expenditures in Social Sectors and Additional Programs in 1999 (Millions of Real 1995 Sucres) Table A24. Central Government Expenditures in the Social Sectors - Personnel vs. Other Costs (Real 1995 Sucres) Table A25. Central Government Expenditures in the Social Sectors - Recurrent Costs vs. Capital Investment (Real 1995 Sucres) Central Government Expenditures in the Health Sector Table A26. Central Government Expenditures in the Health Sector by Institution (Real 1995 Sucres) Table A27. Central Govemment Expenditures in the Health Sector by Activity (Real 1995 Sucres) Table A28. Central Government Expenditures in the Health Sector by Personnel and Other Costs (Real 1995 Sucres) Table A29. Central Government Expenditures in the Health Sector by Personnel and Other Costs (Percentage haring of total health expenditures) Table A30. Central Government Expenditures in the Health Sector by Recurrent and Capital Costs (Real 1995 Sucres) Table A31. Central Government Expenditures in the Health Sector by Recurrent and Capital Costs (Percentage sharing of total health expenditures) Table A32. Central Government Expenditures in the Health Sector by Province (Real 1995 Sucres) Table A33. Central Government Expenditures in the Health Sector by Province (Percentage sharing of total health expenditures) Central Government Expenditures in Education Table A34. Central Government Expenditures in Education by Institution (Real 1995 Sucres) Table A35. Central Government Expenditures in Education by Personnel and Other Costs (Real 1995 Sucres) Table A36. Non-personnel Expenditures as a Percentage of Total Education Expenditures Table A37. Central Government Expenditures in Education by Personnel and Other Costs as a Percentage of Total Education Expenditures Table A38. Central Government Expenditures in Education by Recurrent and Capital Costs (Real 1995 Sucres) Table A39. Central Government Expenditures in Education by Recurrent and Capital Costs (Percentage sharing) Table A40. Central Government Expenditures in Education by Activity (Real 1995 Sucres) Table A41. Central Government Expenditures in Education as a percent of GDP (Real 1995 Sucres) Table A42. Central Government Expenditures in Education as a percent of Non-Debit Spending (Real 1995 Sucres) Table A43. Per Student Spending (Real 1995 Sucres) Table A44. Per Student Spending (Real terms average 1995 exchange rate) Table A45. Central Government Expenditures in Education by Province (Real 1995 Sucres) Table A46. Central Government Expenditures in Education by Province (Percentage sharing) Central Government Expenditures in the Labor Sector Table A47. Central Government Expenditures in the Labor Sector by Institution (Real 1995 Sucres) Table A48. Central Government Expenditures in the Labor Sector by Activity (Real 1995 Sucres) Table A49. Central Government Expenditures in the Labor Sector by Personnel and Other Costs (Real 1995 Sucres) Table A50. Central Government Expenditures in the Labor Sector by Personnel and Other Costs (Percentage sharing) Table A51. Central Government Expenditures in the Labor Sector by Recurrent and Capital Costs (Real 1995 Sucres) Table A52. Central Government Expenditures in the Labor Sector by Recurrent and Capital Costs (Percentage sharing) Central Government Expenditures in Social Welfare Table A53. Central Government Expenditures Social Welfare by Institution (Real 1995 Sucres) Table A54. Central Government Expenditures in Social Welfare by Personnel and Other Costs (Real 1995 Sucres) Table A55. Central Government Expenditures in Social Welfare by Personnel and Other Costs (Percentage sharing) Table A56. Central Government Expenditures in Social Welfare by Recurrent and Capital Costs (Real 1995 Sucres) Table A57. Central Government Expenditures in Social Welfare by Recurrent and Capital Costs (Percentage sharing) Table A58. Central Government Expenditures in Social Welfare by Province (Real 1995 Sucres) Annex 1: Statistical Annex Table Al. Distribution and Incidence of Poverty in 1998, by Geographic Area Population % National All Poor % All Poor Extreme Poor % Extreme Population Poor Costa 6,109,026 55 2,646,758 52 790,170 42 Rural 2,052,163 18 1,334,604 26 473,236 25 Urban 4,056,863 36 1,312,154 26 316,934 17 Sierra 4,708,827 42 2,275,786 45 1,017,881 55 Rural 2,353,073 21 1,687,092 33 874,092 47 Urban 2,355,754 21 588,694 12 143,789 8 Oriente 376,452 3 169,986 3 57,047 3 Rural 302,023 3 151,706 3 54,106 3 Urban 74,429 1 18,280 0 2,941 0 National 11,194,305 100 5,092,530 100 1,865,098 100 Rural 4,707,259 42 3,173,402 62 1,401,434 75 Urban 6,487,046 58 1,919,128 38 463,664 25 Source: LSMS 98. Statistical Annex, Page 1 Table A2. Distribution and Incidence of Poverty In 1998, by Consumption Quintiles. (Percentages) Qi Q2 Q3 Q' Qs pExtreme Poor Non-poor Rural Urban Costa Sierra Oriente National Poor Population 20.0 20.0 20.0 20.0 20.0 16.8 45.8 54.2 42.1 57.9 54.6 42.1 3.4 100.0 Extreme Poor 96.4 3.6 0.0 0.0 0.0 100.0 100.0 0.0 75.1 24.9 42.4 54.6 3.1 100.0 Poor 43.7 42.9 13.3 0.1 0.0 36.6 100.0 0.0 62.3 37.7 52.0 44.7 3.3 100.0 Non-poor 0.0 0.6 25.7 36.9 36.9 0.0 0.0 100.0 25.1 74.9 56.7 39.9 3.4 100.0 Source: LSMS 98. Table A3. Household Characteristics by Conswnption Quintiles and Poverty Status Qi Q2 Q3 Q4 Qs OExtren Poor Non-poor Rural Urban Costa Sierra Oriente National Poor Family Size 7.40 6.28 5.65 4.84 4.29 7.60 6.70 4.84 6.13 5.34 5.74 5.56 6.00 5.67 Female Headed 16.22 17.89 17.57 18.64 21.32 16.54 17.13 19.56 15.07 21.27 18.23 19.86 14.56 18.81 AverageAge 23.08 24.01 25.57 25.90 29.91 22.71 23.70 27.37 26.06 25.49 25.49 26.31 22.50 25.73 Number of Children Less Than6Years 1.50 1.11 0.83 0.63 0.46 1.61 1.24 0.62 0.98 0.85 0.94 0.84 1.15 0.90 NumberofElderlyPeople 0.34 0.23 0.26 0.17 0.18 0.34 0.28 0.20 0.30 0.19 0.23 0.25 0.14 0.24 Years Schooling Mothers 3.64 5.21 6.43 8.26 10.90 3.57 4.72 8.79 4.67 8.61 7.41 6.71 6.22 7.08 25 YearsOld 6.31 7.10 8.89 10.04 12.43 6.80 6.96 10.74 6.84 10.60 9.37 8.97 8.86 9.19 Household &Spouse 3.37 4.87 6.14 7.96 11.17 3.28 4.38 8.91 4.51 8.95 7.32 7.05 6.34 7.17 Dependance Rate 3.47 2.62 2.28 2.02 1.63 3.53 2.94 1.96 2.84 2.08 2.46 2.27 3.15 2.40 HouseHasBad Wall 32.31 25.29 17.86 12.19 2.28 32.46 27.77 9.71 26.43 11.90 29.07 4.93 2.13 18:01 House Has Mud Floor 34.13 14.58 7.54 3.66 1.50 36.02 22.41 3.72 23.71 3.94 4.80 22.75 1.91 12.26 Percent Lacking Piped Water 75.41 58.56 41.00 31.34 12.13 74.27 64.07 26.46 71.71 23.18 51.02 32.42 62.69 43.59 House WithoutElectricity 22.31 10.27 4.44 1.38 0.61 22.15 15.05 1.07 18.12 0.26 7.18 6.90 28.33 7.77 Overcrowded 0.37 0.51 0.58 0.71 1.09 0.35 0.45 81.95 0.58 0.71 0.63 0.69 0.62 0.66 Rent Home 21.38 27.00 31.06 34.94 33.02 21.61 25.63 32.73 19.99 36.68 26.33 34.23 26.42 29.66 Source: LSMS 98. Statistical Annex, Page 2 Table A4. Mean Total Consumption Expenditures 1 er Capita and per Household Consumption Per Capita Per Quintile Isucres] Hsucresl Q1 108,659 782,715 Q2 202,049 1,265,781 Q3 299,545 1,684,742 Q4 455,605 2,200,072 Q5 1,119,321 4,618,036 National 436,723 2,109,000 Source: LSMS 1998. Table AS. Age Distribution of Househols in Ecuador, 1998 Percent of Household With Children in Both Age Groups Quintile Age Group Q0 2 35 -15 >=65 0-2 _70.03 80.18 18.79 3-S 6 .70 6-15 50.14 60 _ >=65 41.40 43.87 73.44 Q5 0-2 52.68 5.71 3-5 26.31 61.08 8.59 6-15 17.91 24.31 >=65 7.27 12.80 34.53 Note: The numbers in each row state the number of poor households having members of both age groups as a percentage of all households in the age group stated at the left of the row. Source: LSMS 1998. Statistical Annex, Page 3 Table A6. Population According to Sex, Disease Symptoms and Age Groups by Consumption Quintiles Urban Population Q1 Q2 Q3 Q4 Qs Total Both Disease Symptoms 3,408,693 50.91 50.38 52.83 52.55 53.62 52.49 Age 0 to 5 587,727 65.75 71.83 70.80 72.81 77.05 72.04 Age 6 and Higher 2,820,966 46.76 45.70 50.29 49.65 51.60 49.66 Men Disease Symptoms 1,554,118 42.25 50.86 50.18 48.85 48.91 48.92 Age 0 to 5 293,710 48.20 73.99 74.33 68.14 76.42 69.60 Age 6 and Higher 1,260,408 40.68 45.24 46.57 45.82 46.56 45.74 Women Disease Symptoms 1,854,575 60.07 49.89 55.32 56.15 57.82 55.89 Age 0 to 5 294,017 82.76 69.01 67.15 78.26 77.61 74.65 Age 6 and Higher 1,560,558 53.35 46.15 53.73 53.31 56.11 53.34 Rural Both Disease Symptoms 2,558,546 50.03 54.46 57.68 58.23 62.38 54.32 Age 0 to 5 395,687 63.74 69.71 71.00 72.34 73.40 67.66 Age 6 and Higher 2,162,859 47.66 52.28 55.96 56.87 61.52 52.44 Men Disease Symptoms 1,249,234 47.88 52.28 54.68 53.90 62.13 51.80 Age 0 to 5 208,204 66.19 67.19 73.35 66.03 79.23 67.89 Age 6 and Higher 1,041,030 44.54 49.95 52.67 52.75 60.68 49.47 Women Disease Symptoms 1,309,312 52.33 56.84 60.79 62.42 62.64 56.97 Age 0 to 5 187,483 60.85 72.90 69.21 78.30 66.55 67.41 Age 6 and Higher 1,121,829 50.94 54.76 59.50 60.87 62.36 55.53 Source: LSMS 1998. Stadistical Annex, Page 4 Table A7. Type of Health Providers Attending to Individuals with Diseae Symptonms by Consumption Quinties Urban Population Qi Q2 Q3 Q4 Q5 Total Both 1,590,385 100.00 100.00 100.00 100.00 100.00 100.00 Hospital 259,746 14.22 16.76 14.67 17.44 16.60 16.33 Healih Center 120,080 15.55 10.93 7.56 7.66 4.52 7.55 Healih Subcenter 69,076 6.93 6.20 5.22 4.18 2.63 4.34 Clinic 639,533 27.11 30.33 37.09 39.20 49.82 40.21 Phamacy 375,654 32.39 30.67 29.65 24.41 14.44 23.62 Home 69,042 NA 1.89 2.22 3.38 8.34 4.34 Other 57,254 3.80 3.22 3.58 3.73 3.64 3.60 Men 727,473 100.00 100.00 100.00 100.00 100.00 100.00 Hospital 131,849 16.13 19.85 15.23 20.13 17.73 18.12 Health Center 56,577 21.36 13.43 6.68 6.00 3.89 7.78 Heath Subcenter 35,130 6.96 6.81 5.73 4.56 2.77 4.83 Clinic 272,443 18.42 26.47 38.15 39.94 45.39 37.45 Pharmacy 176,382 33.98 29.15 28.88 22.68 17.32 24.25 Home 24,543 NA 1.20 1.61 2.61 7.37 3.37 Oder 30,549 3.16 3.10 3.72 4.09 5.54 4.20 Women 862,912 100.00 100.00 100.00 100.00 100.00 100.00 Hospital 127,897 12.28 13.05 14.17 15.30 15.81 14.S2 Health Center 63,503 9.67 7.94 8.35 9.00 4.97 7.36 Health Subcenter 33,946 6.90 5.47 4.78 3.87 2.53 3.93 Clinic 367,090 35.92 34.95 36.15 38.60 52.92 42.54 Phanmacy 199,272 30.77 32.50 30.33 25.79 12.42 23.09 Home 44,499 NA 2.72 2.77 4.00 9.02 5.16 Other 26,705 4.46 3.38 3.46 3.44 2.32 3.09 |Rural Both 984,836 100.00 100.00 100.00 100.00 100.00 100.00 Hospital 137,597 15.32 12.68 13.92 12.69 16.44 13.97 Health Center 63,622 6.39 7.78 7.66 6.42 0.54 6.46 Health Subcenter 164,786 22.23 16.02 16.43 14.63 9.59 16.73 Clinic 360,968 30.31 35.20 33.39 44.93 48.70 36.65 Phannacy 192,748 16.91 21.95 23.88 13.93 20.08 19.57 Home 30,306 4.07 3.17 2.45 2.69 2.48 3.08 Odher 34,809 4.77 3.19 2.27 4.71 2.17 3.53 Mein 484,605 100.00 100.00 100.00 100.00 100.00 100.00 Hospital 69,529 15.38 12.63 14.60 13.84 16.40 14.35 Health Center 31,174 5.95 8.07 7.84 6.35 0.66 6.43 Health Subcenter 76,945 22.99 14.02 14.98 14.61 6.98 15.88 Clinic 179,147 29.95 37.31 30.43 47.49 50.15 36.97 Pharmacy 99,500 18.64 20.16 27.91 13.91 21.28 20.53 Home 12,833 2.82 3.39 2.51 1.13 3.10 2.65 Odier 15,477 4.28 4.42 1.72 2.66 1.44 3.19 WOUND 500,231 100.00 100.00 100.00 100.00 100.00 100.00 Hospital 68,068 15.27 12.73 13.28 11.65 16.48 13.61 Health Center 32.448 6.86 7.52 7.49 6.49 0.41 6.49 Health Subcenter 87,841 21.42 17.88 17.80 14.64 12.28 17.56 Clinic 181,821 30.70 33.23 36.19 42.62 47.21 36.35 Phannacy 93,248 15.08 23.62 20.06 13.94 18.84 18.64 Home 17,473 5.39 2.96 2.39 4.09 1.84 3.49 Other 19,332 5.29 2.05 2.80 6.55 2.93 3.86 Source: LSMS 1998. Statistical Annex, Page 5 Table A. Enrolment of Chgdren Less Than 6 Years Old In Dayeare and Preschool Qs Q2 Q3 Q4 Qs Extrem Poor Non-poor Ruanl Urban Costs Siem Oriente Nadonal Poor % Enroled / Totl Popultion 0-2 1.96 1.58 0.57 1.33 4.64 1.54 1.41 2.30 1.80 1.92 1.30 2.58 3.38 1.87 3-5 17.37 26.12 29.17 38.41 51.24 16.32 21.93 39.82 22.06 36.35 29.08 31.28 29.99 30.02 % Enroled hn Public lIsditutlona / Total Enrolled 0-2 85.41 94.97 42.72 67.61 35.70 79.29 90.67 48.77 87.34 56.24 63.12 68.75 95.70 68.36 3-5 85.93 75.64 67.10 64.63 24.64 86.86 77.28 49.78 81.25 51.42 57.61 63.55 80.51 61.13 % Enroled In Private Instkutions Totd Enroled 0-2 14.59 5.03 57.28 32.39 64.30 20.71 9.33 50.68 12.66 43.27 36.88 31.25 NA 31.34 3-5 14.07 24.36 32.90 35.37 75.36 13.14 22.72 50.22 18.75 48.58 42.39 36.45 19.49 38.87 Soure: LSMS 98. Sttdstical Annex, Pag 6 Tabk A9. Education Enrollment Rates Qi Q2 QJ Q4 Q5 Exhtmene Poor Non-poor Ruual Urban Costa Sierra Oriente National Primary Gross EnrollmentRate 1.17 1.14 1.06 1.11 1.07 1.17 1.15 1.08 1.13 1.10 1.15 1.08 1.05 1.12 NetEnrollmentRate 0.88 0.94 0.93 0.97 0.99 0.87 0.91 0.97 0.90 0.96 0.93 0.95 0.94 0.94 Repetition Rate first grade 16.98 13.31 4.93 6.84 3.83 18.78 14.18 5.23 15.57 6.50 10.16 9.89 21.02 10.53 second grade 16.12 8.05 7.33 3.38 3.32 18.83 12.76 3.25 11.44 5.81 7.67 9.82 6.39 8.55 %notenrolled6-11 11.99 5.59 6.62 1.69 0.08 12.36 8.95 2.39 9.13 3.24 6.98 4.54 5.35 5.85 Secundary Low Level Gross EnrollmentRate 0.37 0.61 0.96 1.07 1.08 0.34 0.53 1.03 0.57 0.96 0.76 0.77 0.90 0.77 Net Enroliment Rate 0.19 0.31 0.59 0.69 0.80 0.17 0.28 0.70 0.31 0.63 0.44 0.51 0.58 0.48 Repetition Rate firstgrade 6.85 9.15 14.73 12.40 8.38 8.56 9.01 11.75 8.91 11.73 9.64 11.49 13.41 10.65 second grade 5.90 7.07 4.32 13.10 9.65 8.11 7.91 8.37 6.54 9.19 4.80 9.19 28.92 8.21 %not enrolled 12-14 31.88 27.90 14.75 5.81 5.11 31.24 28.10 8.86 29.49 8.73 13.98 24.47 17.86 18.88 Secundary Higher Level Gross EnollmentRate 0.19 0.35 0.53 0.70 0.89 0.17 0.30 0.73 0.30 0.70 0.52 0.53 0.49 0.52 Net Enrollment Rate 0.11 0.20 0.32 0.47 0.62 0.10 0.18 0.49 0.17 0.47 0.32 0.36 0.27 0.34 Repetition Rate firstgrade 0.52 4.82 10.48 11.51 8.75 2.22 3.79 10.57 5.42 9.46 6.50 10.01 23.17 8.44 second grade 23.81 NA 6.97 4.22 5.89 11.84 5.83 5.90 5.53 6.42 6.77 5.59 5.20 6.21 %notenrolled 15-17 63.47 54.50 35.87 19.98 17.42 64.73 55.54 23.28 56.99 24.63 37.74 39.77 47.30 38.91 Source: LSMS 98. Stadstical Annex, Page 7 Table A10; Distribution of Education Enrollments by School Type. (Percentages) Qi Q2 Q3 Q4 Q5 Total Primary Fiscal 89.81 82.96 75.88 65.91 34.61 73.55 Particular 8.34 15.89 20.72 30.08 61.37 23.76 Municipal del Consejo Provincial Fiscomisional, Fiscomilitar, JGB 1.85 1.15 3.40 4.01 4.02 2.69 Total 100.00 100.00 100.00 100.00 100.00 100.00 Secundary Fiscal 83.32 83.71 78.44 72.68 40.68 68.44 Particular 10.50 12.64 16.43 24.12 52.66 26.63 Municipal del Consejo Provincial Fiscomisional, Fiscomilitar, JGB 6.18 3.64 5.13 3.20 6.66 4.92 Total 100.00 100.00 100.00 100.00 100.00 100.00 Source: LSMS 1998. Statistical Annex, Page 8 Table All. Monthly Education Expenditures by Households with Children between 6 and 17 Years Old in Public Schools Qi Q2 Q3 Q4 Qs Total Primary 13,120 20,634 26,781 41,770 74,349 Secundary Low Level 28,983 41,367 49,479 59,405 107,826 Secundary High Level 56,608 57,583 58,195 87,905 159,359 Transport Primary 937 2,132 2,405 4,064 6,927 SecundaryLowLevel 3,510 6,279 8,431 6,912 13,215 Secundary High Level 5,373 7,853 6,880 9,461 12,554 Text and educational material Primary 5,011 7,159 9,324 12,956 17,368 Secundary Low Level 10,818 14,720 19,434 21,812 33,693 Secundary High Level 23,376 20,525 21,680 27,864 61,534 Other (enrollments, uniforms..) Primary 7,171 11,343 15,052 24,751 50,055 Secundary Low Level 14,656 20,368 21,613 30,682 60,918 Secundary High Level 27,858 29,205 29,634 50,579 85,272 Consumption per capita Primary 106,469 199,145 295,651 448,245 815,011 Secundary Low Level 123,722 204,159 301,544 452,088 793,677 Secundary High Level 125,458 209,224 302,588 464,499 838,967 Per student educational expenditure / consumption per Prinmary 12.32 10.36 9.06 9.32 9.12 Secundary Low Level 23.43 20.26 16.41 13.14 13.59 SecundaryHighLevel 45.12 27.52 19.23 18.92 18.99 Total household consumption Primary 845,522 1,437,782 1,850,573 2,589,409 4,365,905 Secundary Low Level 1,018,731 1,528,933 1,995,391 2,606,158 4,474,248 Secundary High Level 1,050,260 1,573,821 1,928,427 2,896,528 4,120,318 Per student educational expenditure / total household Primary 1.55 1.44 1.45 1.61 1.70 Secundary Low Level 2.85 2.71 2.48 2.28 2.41 Secundary High Level 5.39 3.66 3.02 3.03 3.87 Note: Monthly Education Expenditures = Total expenditure in target level / number of children between 6-17 enrolled in target level. Source: LSMS 98. Statistical Annex, Page 9 Tak A12 Moout Educ.tlm Expe_dkw hL HombA wibt CbiIdrm betwwe 6 _md 17 Yom Old in Pubic Seboob Ql Qz Qs Q4 Qs PEiu- PoWt Ne-powr Rl Urba Cotd Siewm Orate Natloesi Totl Priay 13,120 20,634 26,781 41.770 74,349 12,599 17,336 43,1m 17,780 38,485 21,004 37,766 24,246 28,559 Scomdary Low Levwl 28,983 41,367 49,479 59,405 107,826 29,320 37,171 68,293 41,310 62,933 39,866 70,273 53,348 54,355 Se-dary HihLvel 56,608 57,5U3 58,195 37,905 159,359 49,610 56,052 100,743 59,153 95,991 79,869 91,479 86,4S4 85,276 Tra_rt Primay 937 2,132 2,405 4,064 6,927 591 1,495 4,210 1,542 3,649 760 4,901 1,902 2,639 S@codWYLowL.wl 3,510 6,279 8,431 6,912 13,215 2,763 5,091 9,306 5,127 8,975 1,917 13,496 6,473 7,412 SecdayHlghLc-l 5,373 7,853 6,S80 9,461 12,554 4,368 6,266 10,053 7,25S 9,464 2,266 15,335 18,812 8,323 Text and labcaosi materii Primay 5,011 7,159 9,324 12,956 17,361 5,012 6,334 12,660 6,012 11,804 7,089 11,345 S,454 9,027 SecwudayLowLewvl 10,818 14,720 19,434 21,812 33,693 11,909 13,679 24,081 15,539 22,047 14,740 24,129 16,076 19,423 SecundryHi hLevel 23,376 20,525 21,680 27,S64 61,534 19,769 20,911 35,745 19,451 35,140 31,141 29,545 36,K60 30,577 Other (earoulnuat, utoeu...) Primary 7,171 11,343 15,052 24,751 50,055 6,995 9,507 26,908 10,226 23,032 13,155 21,520 13,890 16,893 SewdaryLowLevel 14,656 20,368 21,613 30,632 60,913 14,64S 18,401 34,907 21,094 31,917 23,210 31,948 30,799 27,520 Secudary ffigh Level 27,358 29,205 29,634 50,579 35,272 25,473 28,875 54,946 32,443 51,387 46,462 46,099 30,812 45,377 % Primry trapot 7.14 10.33 8.9" 9.73 9.32 4.69 8.62 9.62 8.67 9.48 3.62 12.9S 7.S5 9.24 texl 33.19 34.69 34.32 31.02 23.36 39.78 36.54 23.92 33.31 30.67 33.75 30.04 34.M7 31.61 other 54.66 54.97 56.20 59.25 67.32 55.52 54.84 61.47 57.52 59.85 62.63 56.98 57.29 59.15 % SLL tfulsport IZII 15.18 17.04 11.63 12.26 9.42 13.70 13.63 12.26 14.26 4.81 19.21 12.13 13.64 text 37.32 35.58 39.28 36.72 31.25 40.62 36.80 35.26 37.29 35.03 36.97 35.33 30.13 35.73 oher 50.57 49.24 43.68 51.65 56.50 49.96 49.51 51.11 50.45 50.71 58.22 45.46 57.73 50.63 % SHL tansot 9.49 13.64 11.S2 10.76 7.8S 8.81 11.18 9.9S 12.27 9.S6 2.84 17.31 21.75 10.35 te3d 41.30 35.65 37.25 31.70 38.61 39.J5 37.31 35.48 32.88 36.61 3S.99 32.30 42.62 35.86 oder 49.21 50.72 50.92 57.54 53.51 51.35 51.51 54.54 54.85 53.53 58.17 50.39 35.63 53.80 Surc: LSMS 199S. Statiai Ann,r Pop 10 Table A13. Monthly Educaton Expenditures by HoueeWs wlkth Children between 6 and 17 Yean Old In Privte Seee Ql Q, QsQs Q Q Extrme r Ne_per P Rural Uribn CostR Slerr Orite N _Iena Poor Total Primazy 17,165 26,181 42,225 65,844 243,784 17,724 25,481 155,425 45,942 137,682 62,164 221,819 52,199 120,507 Scoundry Low Level 27,307 48,348 .56,281 86,773 314,414 26,934 45,832 202,961 99,259 191,407 106,676 259,164 102,826 174,986 SecundsiyHighLevel 54,666 69,027 68,109 128,460 366,221 41,282 66,801 280,309 91,108 279,942 178,441 348,025 84,315 256,317 Tranport Primay 119 1,973 2,946 4,655 25,205 145 1,345 15,202 4,044 13,160 2,494 26,924 3,636 11,454 Seoundary Low Level 1,970 4,490 5,604 5,922 32,594 724 4,692 19,944 9,202 13,703 7,733 28,437 7,539 17,010 Secundary High Level 3,987 5,900 9,704 7,921 37,410 10,520 7,192 27,619 4,564 28,304 10,471 42,554 17,081 25,334 Text and educalenal materbl Primay 6,404 8,682 10,280 14,799 34,878 6,599 7,809 24,755 9,664 22,596 16,081 27,231 17,066 20,175 Secndary Low Level 10,517 13,861 15,497 21,209 38,851 11,751 12,782 29,S43 24,711 27,189 21,365 33,367 21,805 26,747 Secwxtaryligh Level 20,768 17,868 18,515 26,379 55,547 13,799 18,481 44,991 24,224 44,550 39,136 45,431 31,823 42,007 Other (enrollents, unfe-) Primay 10,642 15,526 29,000 46,390 183,701 10,980 16,327 115,467 32,231 101,925 43,589 167,664 31,498 88,878 SecundayLowLevel 14,819 29,997 35,180 59,643 242,969 14,459 28,358 153,175 65,347 145,515 77,577 197,360 73,483 131,228 Secundary Higb Lavel 29,912 45,259 39,890 94,159 273,264 16,964 41,129 207,699 62,321 207,088 128,834 260,040 35,410 18,976 % Prii.a Rpo-t 0.70 7.54 6.98 7.07 10.34 0.82 5.28 9.78 8.81 9.56 4.01 12.14 6.96 9.50 text 37.31 33.16 24.34 22.48 14.31 37.23 30.65 15.93 21.04 16.41 25.87 12.28 32.69 16.74 odw 62.00 59.30 68.68 70.45 75.35 61.95 64.08 74.29 70.16 74.03 70.12 75.59 60.34 73.75 % SLL Iraupoet 7.22 9.29 9.96 6.82 10.37 2.69 10.24 9.83 9.27 9.77 7.25 10.97 7.33 9.72 tetl 38.52 28.67 27.54 24.U 12.36 43.63 27.89 14.70 24.90 14.20 20.03 12.87 21.21 15.29 odher 54.27 62.04 62.51 68.73 77.28 53.68 61.87 75.47 65.83 76.02 72.72 76.15 71.46 74.99 % SHL trAt 7.29 8.55 14.25 6.17 10.22 25.48 10.77 9.85 5.01 10.11 5.87 12.23 20.26 9.8S text 37.99 25.89 27.19 20.54 15.17 33.42 27.67 16.05 26.59 15.91 21.93 13.05 37.74 16.39 udier 54.72 65.57 58.57 73.30 74.62 41.09 61.57 74.10 68.40 73.98 72.20 74.72 42.00 73.73 Sowce:LSMS 1998. S P10loM 1nn Page 1I Table A14. MontMy Educatien Eipm 4dku by .uRel sde with Children between 4 and 17 Yeaw Old in Publi un Privte S dhoob Qi Q, Q3 Q4 Qs Extrm POo Non-po Rurgl Urban Coeta Semn Oriente Natlenal POor Total Primay 13,494 21,517 30,091 49,077 176,324 13,070 18,467 85,385 21,122 73,204 33,030 78,557 27,070 52,137 See_aryLowLevel 28,816 42,116 50,824 66,658 215,315 29,081 38,247 112,692 49,199 103,755 57,738 116,981 56,308 84,724 SeoundaryHighLcvel 56,422 59,0S4 60,174 97,384 286,368 49,103 57,468 170,816 64,529 163,528 111,797 174,978 86,300 139,9S0 Trport Primary 862 2,107 2,521 4,243 17,927 550 1,474 8,306 1,839 6,978 1,267 9,782 2,077 4,899 Secundary Low Lcvel 3,356 6,087 7,872 6,649 23,298 2,558 5,041 12,813 5,651 12,066 3,473 17,191 6,537 9,828 SeoundaryllighLevel 5,240 7,596 7,444 9,101 27,815 4,743 6,388 16,908 6,805 16,381 4,923 24,531 18,665 14,104 Text and educational materal Primry 5,140 7,401 9,529 13,516 27,906 5,158 6,539 17,167 6,446 15,581 9,716 14,865 9,324 11,886 Secondary Low Level 10,7S8 14,628 18,656 21,652 36,377 11,893 13,567 25,980 16,762 23,681 16,512 26,940 16,418 21,267 Secundary High Lcvcl 23,127 20,177 21,048 27,517 57,858 19,406 20,591 39,353 20,254 38,595 33,731 34,716 36,432 34,232 Other (enrollments, unifeos-.) Primary 7,492 12,009 18,042 31,318 130,490 7,362 10,454 59,911 12,838 50,644 22,047 53,910 15,669 35,352 Secondary Low Level 14,672 21,401 24,297 38,356 155,641 14,629 19,639 73,899 26,785 68,008 37,753 72,850 33,353 53,629 SecndaryHighLcvel 28,055 31,311 31,682 60,766 200,695 24,954 30,489 114,555 37,469 108,552 73,142 115,731 31,203 91,644 % Primary tprt 6.38 9.79 8.38 8.65 10.17 4.21 7.98 9.73 8.71 9.53 3.84 12.45 7.67 9.40 text 38.09 34.40 31.67 27.54 15.83 39.46 35.41 20.11 30.52 21.28 29.42 18.92 34.44 22.80 other 55.53 55.81 59.96 63.81 74.01 56.33 56.61 70.17 60.78 69.18 66.75 68.63 57.88 67.81 % SLL trant 11.65 14.45 15.49 9.98 10.82 8.80 13.18 11.37 11.49 11.63 6.01 14.70 11.61 11.60 text 37.44 34.73 36.71 32.48 16.89 40.90 35.47 23.05 34.07 22.82 28.60 23.03 29.16 25.10 Other 50.92 50.81 47.81 57.54 72.28 50.31 51.35 65.58 54.44 65.55 65.39 62.28 59.23 63.30 % SHL trapot 9.29 IZ6 12.37 9.35 9.71 9.66 11.12 9.90 10.55 10.02 4.40 14.02 21.63 10.08 text 40.99 34.15 34.98 28.26 20.20 39.52 35.83 23.04 31.39 23.60 30.17 19.84 42.22 24.46 other 49.72 52.99 52.65 62.40 70.08 50.82 53.05 67.06 58.07 66.38 65.42 66.14 36.16 65.47 Soee: LSMS 1998. S talicl Aamu, Pap 12 Table AIL Sef Repored Numbe of Days of Sol MNsse Mon" Q, Q2 Qs Q Qs EQs Poor N _-p. Rol Urba_ CA SMerra Oleab Natlotal Ae 6-11 Ulnderoontolofthe howhold 1.0 1.6 2.1 0.9 1.4 1.0 1.4 1.0 O.S 1.8 2.4 0.7 0.6 1.3 Oubiithocontcolofthehousehold 2.8 1.6 1.5 1.0 0.9 2.9 2.3 1.2 1.8 1.9 2.2 1.4 1.1 1.8 O0I 2.5 1.4 1.2 1.0 1.0 2.4 1.9 1.2 1.9 1.3 1.7 1.5 1.9 1.6 Total 6.3 4.6 4.7 3.0 3.4 6.4 5.6 3.5 4.4 5.0 6.3 3.6 3.5 4.7 Ae 12-14 Undercontrolofthehousehold 1.1 2.1 0.8 1.3 0.8 1.2 1.5 1.2 1.1 2.0 3.0 0.9 0.9 1.5 Ousidethecontrolofthehouohold 8.3 2.8 2.7 1.3 0.9 9.6 5.4 1.5 3.2 5.3 5.0 2.6 1.7 4.2 Other 3.4 0.9 3.5 0.8 0.7 2.4 3.5 0.9 1.6 2.1 3.4 1.1 1.8 2.0 Total 12.8 5.8 7.0 3.3 2.4 13.3 10.3 3.6 5.9 9.4 11.4 4.5 4.5 7.6 AV 18-17 Underoontrolofthehousehold 1.4 1.1 2.1 1.8 0.8 1.5 2.1 1.6 1.7 1.7 2.2 1.1 0.8 1.7 Outsidethecontrolofthehetoueold 2.1 2.4 1.1 0.8 3.3 2.2 1.9 1.4 2.0 1.3 1.5 1.7 1.3 1.5 Other 0.4 0.6 1.3 0.6 1.7 0.4 0.4 1.2 0.8 1.1 1.6 0.8 0.7 1.0 Total 3.9 4.1 4.4 3.3 5.8 4.1 4.4 4.2 4.6 4.1 5.4 3.6 2.8 4.3 NoteE Under control of the houehld nclude: domestic labor, cost of schooling. work Outside the control of the houehold ncluds: ilhn, s,tie weter. Odtr includes: not inbteestd md the lack of u infirstructures and equiment Source: LSMS 98. SMb8cal Annf t Pap 13 Table A16. Chgdren's Economic and Educatinal Activity. Age Group 6-11. (Percentages) Total Work and OyWork Total OnlyStudyNo Work nor Total school 0.1' Working OnlyStud Girls 6.10 OA9 6.60 89.67 3.73 100.00 Boys 8.82 0.75 9.56 83.69 6.74 100.00 ; Total 7.48 0.62 8.10 86.64 5.26 100.00 i Girls 9.69 1.12 10.81 83.26 5.93 100.00 Boys 13.67 1.43 15.11 75.21 9.68 100.00 ii Total 11.73 1.28 13.01 79.13 7.86 100.00 j Girls 3.30 0.00 3.30 94.70 2.01 100.00 Boys 4.84 0.19 5.03 90.64 4.33 100.00 Total 4.07 0.09 4.17 92.66 3.17 100.00 lstQ. Work and OnlyWork Total OnlyStuy No Work nor Total School Worling School s Girls 9.07 1.76 10.83 82.72 6.45 100.00 Boys 12.04 2.08 14.12 72.54 13.34 100.00 E- TotlW 10.63 1.93 12.56 77.38 10.06 100.00 Girls 9.96 2.19 12.15 80.14 7.71 100.00 Boys 14.67 2.36 17.03 68.87 14.10 100.00 0 Total 12.40 2.28 14.68 74.31 11.02 100.00 3 Girls 5.37 0.00 5.37 93.38 1.25 100.00 Boys 2.59 1.08 3.67 85.73 10.60 100.00 Total 3.84 0.59 4.43 89.17 6.40 100.00 2d Q. Work and Ol Work Total Only Stdy No Work nor Total school Ou1' Workin ny td School Girls 8.62 0.09 8.71 86.50 4.79 100.00 Boys 11.16 0.31 11.48 82.51 6.01 100.00 FI Total 9.87 0.20 10.07 84.54 5.39 100.00 '4 Girls 12.89 0.17 13.06 82.36 4.58 100.00 Boys 12.42 0.55 12.97 81.70 5.33 100.00 i Total 12.65 0.36 13.01 82.03 4.96 100.00 ; Girls 3.37 0.00 3.37 91.59 5.04 100.00 Boys 9.45 0.00 9.45 83.62 6.92 100.00 Total 6.28 0.00 6.28 87.78 5.94 100.00 Source: LSMS 1998. Statistical Annex, Pagel4 Table A17. Clhdren's Economic and Educational Activity. Age Group 12-14 (Percentages) Total Work and OnlyWork Total Only Study No Work nor Total school Ol' Working School Girls 20.84 0.41 31.25 59.17 9.57 100.00 Boys 33A8 14.53 48.01 48.65 3.34 100.00 i Tota 27.26 12.51 39.76 53.83 6.41 100.00 i Girls 24.96 16.36 41.32 44.43 14.25 100.00 Boys 44.43 25.20 69.62 27.14 3.24 100.00 0 Total 34.76 20.81 55.57 35.72 8.71 100.00 ; Girls 16.82 4.62 21.45 73.54 5.01 100.00 Boys 23.17 4.49 27.66 68.90 3.44 100.00 ; Total 20.07 4.56 24.63 71.17 4.20 100.00 1stQ. okw nyWr Total OnyStd No Work nor Tb or l rl Worldng School Tol Girls 23.53 20.55 44.08 43.00 12.92 100.00 Boys 35.27 27.42 62.69 34.12 3.20 100.00 | Total 30.06 24.37 54.43 38.06 7.51 100.00 2 Girls 25.86 24.28 50.15 35.73 14.12 100.00 Bey 40.78 32.53 73.32 22.96 3.72 100.00 Ad Total 33.99 28.77 62.76 28.78 8.46 100.00 Girls 13.58 4.60 18.18 74.07 7.75 100.00 ffi Boys 16.55 10.07 26.62 71.96 1.42 100.00 5 Total 15.36 7.89 23.25 72.80 3.95 100.00 2nd Q. Work and Only Work Total Ol,y Study No Work nor Tohl School Working SchoolToa Girls 20.18 13.97 34.16 50.80 15.04 100.00 Boys 39.15 18.10 57.25 34.11 8.64 100.00 I Tota 29.50 16.00 45.51 42.60 11.90 100.00 S Girls 20.83 15.28 36.11 46.11 17.77 100.00 Boys 46.00 26.21 72.21 23.87 3.92 100.00 A Total 32.68 20.43 53.11 35.64 11.25 100.00 Girls 18.93 IIA3 30.36 59.92 9.72 100.00 i Boys 28.52 5.52 34.05 49.98 15.97 100.00 5 Total 23.99 8.32 32.31 54.68 13.02 100.00 Source: LSMS 1998. Sbtstical Annex, Pape 15 Table Alg. Children's Economic and Educational Activity. Age Group 15-17. (Percentages) Total Work and y Total Ol Std No Work nor Total Total School Work hin schlylWorldog OnyStd SchoolToa ,. Girls 18.69 21.76 40.45 42.78 16.78 100.00 I Boys 31.50 32.48 63.98 29.24 6.78 100.00 I. Total 25.47 27.44 52.92 35.61 11.48 100.00 j Girls 16.93 33.85 50.78 25.86 23.35 100.00 Boys 32.49 50.92 83.41 10.69 5.91 100.00 i Total 24.97 42.67 67.65 18.02 14.34 100.00 Girls 20.14 11.71 31.85 56.84 11.30 100.00 B BOys 30.74 18.51 49.25 43.31 7.44 100.00 Total 25.87 15.38 41.25 49.53 9.22 100.00 lst Q. Work and Oy Work Total Only Study No Work nor Total School Worhing School @ Girls 14.34 41.68 56.03 18.13 25.84 100.00 I Boys 20.07 53.55 73.62 19.99 6.39 100.00 P Total 17.41 48.03 65.44 19.13 15.44 100.00 -4 Girls 11.81 44.62 56.43 15.83 27.74 100.00 E Boys 19.37 69.41 88.78 5.61 5.61 100.00 i Total 15.59 57.01 72.59 10.72 16.68 100.00 Girls 27.31 26.68 53.98 29.89 16.13 100.00 Boys 21.93 11.49 . 33.42 58.12 8.46 100.00 ; Total 23.77 16.68 40.45 48.47 11.08 100.00 2nd Q. Work nd Ol Work Total Only Study No Work nor Total Wcorkad OnyWr Working OnySuy school -4 Girls 15.37 26.77 42.14 28.47 29.39 100.00 I Boys 28.42 45.65 74.08 18.31 7.61 100.00 P Total 22.86 37.60 60.46 22.64 16.90 100.00 Girls 14.55 36.83 51.37 17.67 30.95 100.00 b Boys 27.25 54.16 81.41 13.21 5.39 100.00 ii Total 22.23 47.31 69.54 14.97 15.49 100.00 Z Girls 16.38 14.41 30.80 41.73 27.47 100.00 Boys 30.40 31.33 61.73 26.90 11.36 100.00 Total 23.78 23.34 47.13 33.91 18.97 100.00 Source: LSMS 1998. Stascal Annex, Pagel6 Table Al9. Averag Salary per Hour and Monthly by Age Groups Q, QS Qs Q, Qs Extrm Poor Poor Non-poor Rural Urba Coaa Sem Orente National In ueMr Average salary per hour 10-11 1,718.7S 4,608.36 3,601.36 48,053.34 775.19 1,718.78 2,899.55 32,140.28 1,021.25 14,043.52 14,313.19 3,383.62 13,953.49 10,021.67 i AvenraesWlazryperhour 12-14 2,780.84 5,563.10 4,139.29 1,775.95 2,747.01 2,939.21 4,149.90 3,121.05 3,215.66 4,218.83 4,767.68 2,814.09 2,500.75 3,745.72 Avage salary per hour 15-17 6,075.44 5,310.90 6,988.57 2,752.05 4,030.08 6,787.55 5,054.73 5,850.34 5,686.49 5,089.35 3,702.11 7,483.94 3,377.88 5,358.68 * AverageseluyperhourlO-II 1,489.84 NA 1,162.79 NA NA 1,489.84 1,489.84 1,162.79 1,489.84 1,162.79 1,750.00 1,299.85 NA 1,427.85 I, Avewe salaryperhour 12-14 2,516.84 2,204.18 5,958.86 1,560.69 2,561.14 2,515.63 2,390.44 4,776.14 3,474.76 2,456.57 5,007.41 2,401.03 1,834.48 3,076.54 Z Avege saiy per hour 15-17 7,749.76 5,933.78 9,506.70 2,601.36 2,178.61 8,576.62 6,300.11 6,896.22 6,754.05 6,183.68 2,786.75 9,798.53 4,547.87 6,494.55 I Average aryyperhour 1o-11 1,816.60 4,608.36 3,579.83 48.053.34 775.19 1,816.60 3,042.56 34,361.43 884.75 14,354.55 14,842.31 3,770.87 13,953.49 10,829.92 A Average salayper hour 12-14 3,211.35 7,973.58 2,714.17 1,797.49 2,813.69 3,569.52 5,881.83 2,359.83 2,899.11 5,030.16 4,693.46 3,475.83 3,046.45 4,242.19 Avege salry per hour 15-17 2,208.83 4,172.47 4,046.45 2,855.47 6,090.09 2,204.94 3,054.72 4,683.31 3,150.73 4,068.53 4,584.00 2,293.92 1,184.25 3,774.88 In %ar Aveagemonthlysalwy 10-1 1 30,888.47 68,473.74 111,347.50 345,889.30 5,188.21 29,140.51 58,012.52 180,267.80 10,857.68 290.955.40 179,669.50 32,925.76 2,230.02 91,774.71 a Averagcmonthly slaly 12-14 99,285.60 133,828.80 181,466.60 53,582.88 142,618.10 124,743.80 111,327.20 134,731.60 85,574.09 191,779.00 134,471.50 112,729.00 71,313.04 119,261.70 Averagemonthly salary 15-17 510,434.60 733,561.10 718,539.80 262,002.00 403,681.30 561,318.40 578,693.50 518,469.50 503,739.10 615,321.70 340,386.00 831,461.30 203,147.10 552,386.90 u Average monthly selmy 10-11 43,700.45 NA 112,816.90 NA NA 49,599.91 39,977.56 78,786.88 38,227.19 114,428.60 29,655.72 65,417.96 NA 44,951.55 X Average monthly saliy 12-14 175,026.50 179,922.50 383,571.40 99,846.52 321,197.00 203,844.30 171,345.70 382,782.80 169,515.80 400,747.70 255,382.70 198,716.40 154,714.00 212,508.70 Averagemonthlysalary 15-17 624,527.30 1,016,183.00 1,302,209.00 386,188.20 384,693.80 693,583.40 765,566.70 932,147.20 720,594.90 1,015,105.00 308,930.80 1,317,019.00 311,483.80 813,161.70 Aveamgeonlysaly 10-11 28,563.42 69,885.47 111,285.30 353,990.60 5,18.21 25,374.71 59,788.00 183,829.50 8,002.74 295,019.40 191,048.00 30,358.95 2,463.27 95,498.58 A Averagemontly lfy 12-14 37,878.73 108,829.20 109,540.80 50,797.91 106,84.80 63,965.88 72,357.81 83,310.53 35,458.09 144,721.70 101,393.70 56,521.17 41,087.77 76,667.40 Avergemnt lyaaly 15-17 195,613.30 268,620.90 280,289.10 216,275.00 414,732.30 173,738.80 251,449.90 287,688.40 133,607.50 376,552.60 366,530.70 144,196.00 29,804.94 271,379.70 Tn dolar Aveagemohly slary 10-11 5.67 12.57 20.44 63.51 0.95 5.35 10.65 33.10 1.99 53.42 32.99 6.05 0.41 16.85 i Avagemotlyaay 12-14 18.23 24.57 33.32 9.84 26.18 22.90 20.44 24.74 15.71 35.21 24.69 20.70 13.09 21.90 Avergenmonthlyalynyl5-17 93.72 134.68 131.92 48.10 74.12 103.06 106.25 95.19 92.49 112.97 62.50 152.66 37.30 101.42 Averagemondklylaluy 10I1 8.02 NA 20.71 NA NA 9.11 7.34 14.47 7.02 21.01 5.44 12.01 NA 8.25 Ii Averge monly sWlary 12-14 32.13 33.03 70.42 18.33 58.97 37.43 31.46 70.28 31.12 73.58 46.89 36.48 28.41 39.02 Y Avefrgemonthly sabyl5-17 114.66 186.57 239.09 70.90 70.63 127.34 140.56 171.14 132.30 186.37 56.72 241.81 57.19 149.30 Averp mondtly sIwyy 10-11 5.24 12.83 20.43 64.99 0.95 4.66 10.98 33.75 1.47 54.17 35.08 5.57 0.45 17.53 , Average monthly slay 12-14 6.95 19.98 20.11 9.33 19.62 11.74 13.28 15.30 6.51 26.57 18.62 10.38 7.54 14.08 a Averagemonthlvyalay 15-17 35.91 49.32 51.46 39.71 76.15 31.90 46.17 52.82 24.53 69.14 67.30 26.47 5.47 49.83 Source: ISMS 98. Stbsaical Annex, Page 17 Table A20. Average Hours Worked per Week Qi Q2 Q Qs Extren Poor Nonfpoor Rural Urban Coas Skrm Oriente National PoorPO AR worker 10-11 20.0 23.1 27.3 24.8 25.1 19.2 22.9 22.9 20.3 29.0 26.5 21.2 13.4 22.8 12-14 31.1 29.5 27.2 27.7 32.0 31.2 30.2 28.6 28.7 31.9 29.8 30.1 25.9 29.7 15-17 39.3 37.9 38.0 38.0 41.7 38.7 39.1 38.4 36.7 41.4 39.2 38.5 36.3 38.8 Worldu _nd In the school 10-11 17.8 23.1 25.7 24.3 25.1 17.0 22.1 21.6 19.1 28.4 26.5 19.4 12.9 21.9 12-14 26.1 26.1 23.4 26.6 26.5 25.1 26.0 25.3 24.2 28.5 29.2 23.1 21.3 25.8 15-17 36.1 31.7 32.7 34.4 33.8 34.8 34.9 32.6 30.1 36.2 37.3 28.7 22.2 33.6 Werldng and not In swhool 10-11 31.8 20.1 64.5 50.0 NA 31.1 30.8 60.1 32.5 55.2 26.3 44.0 18.7 34.5 12-14 37.3 35.7 37.8 46.1 59.2 39.0 36.6 44.7 36.2 46.8 31.9 40.8 38.4 38.2 15-17 40.5 41.7 45.0 47.5 55.4 40.1 41.4 48.8 40.5 50.2 41.4 45.4 45.2 43.6 Only who bha a salry 10-11 34.4 31.2 50.6 29.9 24.0 34.4 38.2 32.3 28.5 40.5 41.7 29.4 6.0 36.8 12-14 43.5 38.5 39.3 35.0 44.6 43.4 40.8 40.7 38.5 42.8 39.8 42.2 34.6 40.8 15-17 47.8 44.1 45.9 49.0 54.6 45.9 45.7 49.9 44.7 49.5 46.9 48.2 40.6 47.3 D.ea't have a abry 10-11 18.3 20.9 18.4 23.6 25.2 17.1 19.6 21.3 19.7 21.2 21.5 20.0 13.5 20.0 12-14 27.0 23.5 21.2 25.1 20.8 26.3 25.7 22.3 26.0 20.1 23.9 25.3 23.5 24.7 15-17 31.4 29.3 24.9 27.3 29.3 32.0 30.3 27.0 30.6 24.4 29.4 27.3 34.3 28.7 Worldng and school Wkh salary 10-11 30.9 31.2 50.3 29.9 24.0 30.9 37.8 30.7 24.4 40.2 42.3 25.3 6.0 36.0 12-14 37.5 34.2 35.6 33.6 42.9 37.1 36.0 36.4 29.5 40.1 39.0 32.2 20.7 36.2 15-17 46.3 39.7 42.3 46.3 46.6 42.7 42.6 45.4 38.4 46.4 45.6 41.1 26.9 43.8 Wkthst sabry 10-11 16.6 20.9 16.5 22.8 25.2 15.5 18.7 20.1 18.7 20.4 21.1 18.6 12.9 19.1 12-14 23.7 21.6 19.3 24.2 17.5 22.1 22.7 20.9 23.3 19.6 23.7 21.1 21.5 22.2 15-17 23.7 22.2 23.0 26.6 26.3 25.8 23.5 25.2 25.8 23.2 28.3 20.5 20.3 24.6 W eidn Nd not in chool Wkh saary 10-11 42.6 NA 54.0 NA NA 42.6 42.6 54.0 42.6 54.0 28.0 51.5 NA 44.8 12-14 47.2 44.6 44.1 48.7 64.5 47.7 45.8 50.0 45.9 48.7 42.5 44.5 51.5 47.0 15-17 48.4 46.6 49.0 52.9 61.8 47.1 47.6 54.0 47.3 52.9 48.2 51.4 48.0 49.8 Wll_t 10-11 29.0 20.1 72. 50.0 NA 27.7 28.1 62.6 30.4 56.0 26.0 41.1 18.7 31.7 12-14 32.1 27.7 30.1 44.2 50.0 33.4 31.1 35.5 31.6 31.6 24.6 34.8 31.1 31.6 15-17 33.8 34.2 30.9 32.1 40.2 33.9 33.7 34.7 34.2 31.5 31.1 35.1 43.8 33.9 Soume: LSM5 98. Stbtica Anneo, PU 18 Table A21. Summary of Total Central Government Expenditures (Real 1995 Sucre$) 1995 1996 1997 199S 1999* TOTAL CENTRAL GOVERNMENT EXPENDITURES FOR SOCIAL SECTORS (EXCLUDING BONO SOLIDARIO) 2,106,811.3 2,283,103.1 1,990,249.6 2,114,972.2 1,946,935.4 TOTAL CENTRAL GOVERNMENT EXPENDITURES FOR BONO SOLIDARIO - - - - 560,925.3 TOTAL CENTRAL GOVERNMENT EXPENDITuRES FOR OTHER SOCIAL PROGRAMS (CONAMU/CONSEJOS) - - - 11,4S7.3 TOTALCENTRALGOVERNMENTEXPENDrrURES 10,83S,026.8 10,366,213.5 11,908,251.5 11,568,899.6 13,893,644.1 TOTAL CENTRAL GOVERNMENT EXPENDITURES (EXCLUDING DEBT) 6;229524-8 636d3404-3 6,062,110.5 3,027,930.9 6,616,912.9 PAYMENTS ON InERNAL AND EXTERNAL DEBT + AMORTIZATION 4,608,502.0 3,999,809.2 5,846,140.9 4,270,484.3 7,276,731.2 GROSS DOMESTIC PRODUCT 46,005,438.0 60,726,745.0 79,040,013.0 105,894,895.0 139,606,534.8 TOTAL GOVERNMENT EXPENDTURES AS A PERCENT OF GDP 23.56% 22.09% 24.55% 23.75% 34.67% TOTAL CENTRAL GOVERNMENT EXPENDITURES (EXCLUDING DEBT) AS A PERCENT OF GDP 13.54% 13.57% 12.50% 6.22% 16.51% PAYMENTS ON INTERNAL AND EXTERNAL DEBT + AMORTIZATTON (AS REFLECTED IN MOF DATABASE) AS A PERCENT OF GDP 10.02% 8.53% 12.05% 8.77% 18.16% "1999 E.axn dftaopt "Cod'caIo" of iy 1M C"sloadio: Wtdd BY S: 9Subsuua dt Prer y Coiid 9Mty oe FFwe Table A22. Social Sectors as a percentage of GDP SECTOR 195_ 1996 1997 1998 1999 EDUCATION 3.11% 3.18% 2.S5% 3.1S% 3.06% SOCIAL WELFARE 0.37% 0.60% 0.27% 0.19% 0.77/o TRABAJO 0.04% 0.05% 0.05% 0.05% 0.05% SALUD 1.06% 1.03% 0.93% 0.92% 0.97% BONO SOLUDARIO - - - 1.40% OTHER SOCIAL PROGRAMS (CONAMU/CONSEJOS) - - - - 0.03% TOTAL SOCIAL SECTOR SPENDING AS A PERCENT OF GDP 4.58% 4.87% 4.10% 4.34% 6.29% *1.99 Exprai dw%aqwm9zCo-fiCade" Uofufly IM Cslcn : Wwd*do So.: _tasmia de Preasp.i y C_oobila4 Miisly of Fitt Table A23. Central Government ExpendItures In Socal Sectors and AddIional Programs in 1999 (Millons of Real 1995 Sucre) SECTOR 1995 1996 1997 199 1999* EDUCACION 1,429,697 1,494,139 1,384,731 1,550,419 1,228,090 SOCIAL WELFARE 172,346 279,356 130,762 93,292 310,550 TRABAJO 17,907 24,752 23,413 23,970 20,474 SALUD 486,921 484,856 451,343 447,191 387,822 BONO SOLIDARIO (INCLUYE COSTO SERVICIO TRANSACCION) - - - - 560,925 OTBER** - - - - 11,487 TOTAL 2,104871 2,283,103 1,990,250 2,114872 2,519,348 PIB (MILLIONS OF REAL 1995 SUCRE) 46,005,438 46,916,888 48,502,781 41,701,499 40,075,808 */OFPIB(INREAL TERMS 199S) 4.5S% 4.S7% 4.10% 4.34% 6.29% 1999 Eq s da bb a C -t A W cu 19S9 c 1. Let us use a = 2. Then the first case gives P2 = 0.082 and the second gives 0.025. The drawback to using a = 2 is that the measure is hard to interpret. Source: from "Administering Targeted Social Programs in Latin America: From Platitudes to Practice", by Margaret E. Grosh, The World Bank, 1994, Box 3.1, p. 25 29 Annex 4: An Emerging Education Strategy for Ecuador David Harding and Xiaoyan Liang This paper attempts to review basic indicators of attainment, coverage and quality, as well as education expenditures and finance in the education system in Ecuador, in the wake of the macro-economic crisis natural disasters which occurred over the last two years. It identifies the key problems facing the Ecuadorian education system and suggest possible strategies to guide educational policies and direct future World Bank support. The paper uses the most recent data available on the education system in Ecuador including the 1998 Living Condition Survey (Encuesta Condiciones de Vida). It also collates findings and observations from available studies both within and outside the World Bank. I. Educational Attainment Ecuador's educational attainment as measured by years of education has steadily increased over the past few decades. By 1998, the average attainment of 25 year olds reached 9 years with no significant difference between men and women (Figure 1). This level of average attainment places Ecuador relatively high on the list of countries in Latin America. Figure 2 shows that Ecuador's educational attainment over the last 40 years has been between Jamaica and Honduras. In fact, UNDP's education index (composite of adult literacy and educational attainment) categorized Ecuador as one of the medium education development countries. Figure 1 and 2. Educational Attainment 1200 10.00 6~ 6.00Ed] 2 - Ij 400eAO 0- 200 65 45 25 0.00 AgeCohoit 65 60 55 50 45 40 35 30 25 However, there is considerable disparity in educational attainment across income groups and between rural and urban populations. For example, the extreme poor (consumption quintile 1) in rural areas have less than 4 years of education whereas the rich (quintile 5) in urban areas have attained on average almost 12 years of education, three times that of the extreme poor in rural areas (Figure 3). Figure 3. Educational Attainment by Rural and Urban Areas (Age over 15 and have already left School) 14 12 106 4 - 6 ~~~~~~~~~E Rural 4 O~~~~~~~~Urban 0 Qi Q2 Q3 Q4 Q5 Income Quintile Source LSMS 1998 H. Coverage and Quality Coverage Overall coverage at primary level is relatively high at about 95% with only about 10% difference between the poorest and the richest income groups. At the lower secondary level, the overall coverage is about 44%. However, this figure disguises the huge disparity across income groups at this level. The lowest income group has a net enrollment rate of only 19% whereas the highest income group has 80%, four times more than that of the lowest income group (Table 1). This enrollment gap first appears at around age 11 when children enter lower secondary education and the gap widens thereafter (Figure 4). At the upper secondary level, the overall coverage is about 34%. Again, there is big disparity across income groups. Only 11% of the age cohort from the lowest income group are enrolled at the upper secondary level, compared to 62% of the high income group. Furthermore, gross enrollment rates for secondary education have slightly worsened over the last twenty years (Figure 5). Coverage at the preschool level is very low averaging about 18%, the high income group having a net enrollment of 30%, about 3 times higher than that of the low income group. Enrollment at higher education is even less equitable. Only 2% of the age cohort from the poorest quintile have access to that level of education, compared to 33% for the richest income quintile (Table 1). 2 Table 1. Net Enrollment Rates by Level of Education and by Consumption Quintile Consumption Quintile Level of Education QiL12 2 Q3 Q4 Q5 Preschool 0.10 0.13 0.16 0.19 0.30 Prima 0.88 0.94 0.93 0.97 0.99 Lower Secondary 0.19 0.31 0.59 0.69 0.80 Upper Secondary 0.11 0.20 0.32 0.47 0.62 Higher Education 0.02 0.04 0.06 0.16 0.33 Source LSMS 1998 Figure 4. Age-Specific Enrollment by Income Quintile 1.2- 0.8 0.6 -Qi 0.I4 -. __ 0.2 AQ5 0 6 8 10 12 14 16 18 20 Age Source LSMS 1998 Figure 5. Secondary Education Gross Enrollment 1980, 1994 120 ,,,,, ........ __ ...... -- ...... .' ..i' .... 100 Korea 7 80 * Urily ,o 60 *Ma1z, 40 DR 20 20 40 60 so 100 120 19titl Rate ofErenw.t 3 At the primary level, there is no significant difference in coverage between rural and urban groups. However, a significant gap emerges at the lower secondary level at the age of 11 and continues through to the high education level (Figure 6). Figure 6. Age-Specific Enrollment by Urban and Rural Areas Age-Specific Enrollment by Urban and Rural 1.2 - -- .. . ........................ ............ ........... ..... . ........... M0.8 - ' m 0.6 - >> \+_, . + Urban S 0.4 - _ _ _ _Rural 0.2 -U. 6 8 10 12 14 16 18 20 Age Source LSMS 1998 Why is attendance at lower secondary so low especially for the poor? To answer this question, we use two methods. First, we look at the responses given to the question "why are you not enrolled in school"? Among children 11 to 15 years old, main reasons listed for not attending school include cost (34%), not interested (19%), having to work (16%), lack of access to schools or teachers (10%), sickness (4%) and house work (3%). Figure 7. Reasons for not Attending School (Ages 11-15) 40 34.7 3 5 30 20 5 6 1041 3 5 3 0 Z8 u z Source LSMS 1998 Second, we employ multivariate logistic regression method to investigate how factors such as poverty, mother's education, location, and children's work status affect their probability of enrolling in a school. We found that the probability of enrollment at the lower secondary schools among children 11 to 15 is significantly associated with being extreme poor, living in rural areas, mother's education, and whether or not the child is working. 4 With everything else the same, the probability of enrollment for a child from an extreme poor household is only 700/o of a non-extreme-poor child; the enrollment probability of a working child is only 33% of a non-working child. On the contrary, with every additional year of mother's education, the probability of enrollment increases by 1.24 times; living in an urban area also increases the probability of enrolment by 1.74 (Table 2). Table 2. Predicting the Probability of Enrollment in Lower Secondary schools Variables B (S.E.) Exp (B) Mother's education 0.21 (0.02)*** 1.24 Extreme poor -0.34 (0.13)** 0.71 Female-headed households -0.23 (0.17) 0.80 Indigenous 1.23 (0.25)*** 3.43 Urban 0.56 (0.15)*** 1.74 Costa region 0.22 (0.13) 1.25 Sierra region 0.28 (0.21) 1.32 Work -1.12 (0.13)*** 0.33 Constant 0.62 (0.17)*** _ Note: *** P<=0.001 **P<=0.01 * P<=0.05 Source LSMS 1998 In other words, combining the findings from the two methods, we find that children do not attend lower secondary school mostly because of poverty and low maternal education hence lack of demand and stimulation at home. Lack of access in rural areas is also a contributing factor. Due to poverty many poor children have to work. Data show that substantial proportion of children between the ages of 10 to 17 are already participating in the labor force. Figure 8 below shows that children from the lowest income group are much more likely to be working than those from the high income group at all ages, contributing perhaps to why enrollment at secondary level of education is so low especially for the poor. Figure 8. Percentage of Children Participating in the Labor Force % Children Participating in Labor Force 80 . . . . . . 60 40 '-.-. --AgelO-ll 20 Age 12-14 20 J-Age 15-17 Ql Q2 Q3 Q4 Q5 Consumption Quintile Source LSMS 1998 5 Strategies have to be found to drastically increase access to lower secondary education particularly for the rural poor while, at the same, addressing the considerable inequities existing at the primary level which contribute to the widening gap at subsequent levels of education. Such strategies for increasing access to rural education may include demand-side financing programs to compensate families for both direct and indirect costs of schooling, programs aiming at improving home education environment and raising awareness of the importance of education, and increasing access to lower- secondary schools in rural areas such as EDUCO in El Salvador, CONAFEI in Mexico, Escuela Nueva. The use of distance education modalities such as Mexico's Telesecundaria (see Boxes 3 -4) may also need to be explored. Internal Efficiency and Quality Repetition and dropout rates are frequently used as measures of internal efficiency of the education system. Repetition and dropout are costly in terms of the resources required to produce a successful graduate of the system. Sometimes, when measures of quality are lacking, they are also used as quality indicators. Figure 9 and 10 present dropout and repetition rates at primary schools. Children in rural areas are twice more likely to dropout than those in urban areas. The amazing disparity appears at the 6th grade when two thirds of the 6h' graders in rural areas drop out of the system. Repetition is also more severe in rural schools than in urban ones, in Costa region than in Sierra and Oriente regions. On average, 15% of children in rural Costa primary schools repeat at least once. Figure 9 and 10. Primary Dropout and Repetition Rates P-" Drpmt RPt. (f p-eit) Pr-wy ReWM- RPe (. percent) 70 - ~~~~~~~~~~~~~20- 50 / -__ _ _ _ 15 10 ~~~~~~~~~~~~~~~~0 Costa Sian 0Oiete Source: SIISE Comparison with other countries in terms of education achievement can not be made because Ecuador has not participated in any of the international standardized testing such as TIMMS and UNESCO education achievement study. However, data from the 1997 National Achievement Test in language and mathematics shows that within Ecuador, there is great disparity in education achievement between private and public, rural and urban, and across regions. Urban private students out-perform urban public and rural students in both language and mathematics for the 3rd, 7*, and 1'& years. Students from rural schools performed worst. Students from Sierra region achieved better scores than those from the Costa region in both language and mathematics for the 3,rd 7th, and 10'h years (Figures 12-15). 6 Figures 12 and 13. Achievement in Language and Communication 1 | _4 _ _14 E Ur__ ban P 8 - - | Uiea hf~~~~~~~~~~ ua - - I0cot il ~ 1' 1'' a ]E62 - - ~ - O.- 0~~" - - 3fd Yew 7b Yew lObYea 3rd Year 7th Year 10th Year Source: Nadional AchievementTest, 1997 Figure 14 and 15. Achievement in Mathematics 10~~~~~~~~~~~~~~~~ O . -: , 'k _ IhEN _____ hr h rd Year 71h Yew Oi MYea 1*lllrkA Source: National Achievement Test, 1997 At the tertiary level, there are also indications of low quality and efficiency. Combination of counting pre-university students as enrollees in higher education and the long duration of study mean that the actual number of persons having received higher education in Ecuador is much lower than the gross enrollment of 18 % (see Wu et al, June 1999). Internal efficiency is low: on averages it takes 13.4 years to produce a high education graduate. Graduation rate in public universities rarely exceeds 10-15 % of a cohort. Further, cost is incredibly high: the unit cost of $22,460 is actually higher than annual tuition fee of an US Ivy League University. As an indicator of the quality of the higher education system in Ecuador, the number of papers published in international journals averaged only 0.02 per researcher per annum. This number is below Brazil (0.04), Argentina (0.12), Mexico (0.05), Chile (0.15) and Colombia (0. 1 1) (Wu, 1999). Teachers A critical factor affecting inefficiency and low quality is in the recruitment and retention of quality teachers particularly in rural areas. There is evidence that there is a high incidence of absenteeism among teachers in rural areas, due to lack of motivation. A good incentive structure may serve to counteract this. Teachers in Ecuador do not fare as well as their colleagues in other countries. Liang (1999) found that even after controlling for hours worked per week and levels of education, teachers in urban Ecuador are paid 27% less than their counterparts in the labor force (Table 3). 7 Table 3. Hourly Wage Premium for Being a Teacher in 12 Samples (Not Taking into Account Teachers' Three-month Vacation) Countries Hourly wage "premium" of being a teacher Bolivia NS Brazil -0.07 (0.03)* Chile NS Colombia 0.20 (0.02)*** CostaRica 0.15 (0.04)*** Ecuador (urban) -0.27 (0.04)*** El Salvador 0.09 (0.03)** Honduras 0.29 (0.04)*** Panama 0.12 (0.03)*** Paraguay NS Uruguay (urban) 0.10 (0.03)*** Venezuela 0.08 (0.04)- p<=O.10 *pp<=0.05 ** p<)0.01 ***p<=0.001 Source: Liang, 1999. Teacher Pay in Latin America In summary, inefficiency and low quality permeates the education system and he impact is especially severe on the poor and the rural. Mechanisms will have to be found for improving the internal efficiency of the entire education system as well as enhancing the quality and equity at all levels. 11., The Indigenous Populations In Ecuador, the indigenous population is estimated at 12 million, distributed among 22 provinces although different data sources show marked discrepancies in the number of indigenous people; this discrepancies are attributed to difficulties relating to the definition of "indigenous" and problems of self-reporting. 25 % of the total population identifies itself as Indigenous or Afro-Ecuadorian and it is calculated that 80% of this population lives in the rural areas. Indigenous population in Ecuador are more likely to be poor than the non- indigenous. While 47% of the indigenous population are classified as extremely poor - in the bottom 20%/o consumption quintile, only 17% of the non-indigenous are considered extremely poor. Also, indigenous people have on average lower literacy rate and lower educational attainment. Figure 16 sows that almost half of the rural indigenous women are illiterate. Additionally, rural indigenous population on average attain less than 4 years of education (Figure 17). 8 Figure 17. Educational Attainment by Ethnicity 8 6 ~~~~~~~Urban 4 ~~~~~~Rumi Indigenous Non-indigenous Source: LSMS 1998 When looking at the enrollment rates in lower secondary schools, we found that holding everything else constant, being indigenous means greater probability of enrolling in a lower secondary school than non-indigenous, more than 3 times greater. However, this does not mean that indigenous children on average have higher enrollment at this level. Using indigenous as the only explanatory variable for enrollment, we found a significantly lower enrollment among indigenous population than non-indigenous. The disparity only disappears when we hold mother's education and poverty constant. This finding indicates that the enrollment disparity between the indigenous and non- indigenous can be explained away by disparities in mother's education and poverty. Ulsually poor rural indigenous families have considerable financial difficulties in paying school fees, materials and for transportation, thus contributing to the low participation in schools. Also, indigenous parents tend to have little or no formal education themselves and the lack of stimulation and support at home is cited as a reason for low participation and poor performance of their children at school (Ivarsdotter, 1999). In fact, with the same level of mother's education and poverty, the indigenous children (at least among those who are sampled) are 3.4 times more likely to enroll in a lower secondary school. Such findings imply that policies aiming at improving the living standards of indigenous population will help boost their children's enrollment. However, since a large portion of the disparity can be attributed to inter-generational effect, it may take a few generations before disparity will completely disappear, though programs aiming at improving home educational environment such as community-based early child educational programs may help alleviate the disparity. 9 There is evidence of labor market discrimination against the indigenous population. The household survey data show that the return to an additional year of education for the indigenous population is only 8%, as compared to 12% for the non-indigenous population. Reasons for this "discrimination" are less clear. It could be that the quality of education received by indigenous and non-indigenous groups differ, favoring the non- indigenous. Or it could be as simple as a ianguage factor since the indigenous group either don't speak Spanish or don't speak it as well as the Box 1. Guatemala: Bilingual Education non-indigenous people. One possible strategy for bilingual education is to develop the two languages to adequate proficiency level during Bilingual Education primary schooling through simultaneous or parallel bilingual education programs. However, there are very few A recent study developing countries except Guatemala, where the simultaneous bilingual education model is used. The commissioned by the IDB has experiment in Guatemala shows promising results not only suggested that one major because of the increase in students' enrolment and priority for the indigenous and completion but also the achievements in both language and Afro-ecuadorean populations is academic subjects. The six-year bilingual school students access to bilingual education out-performed Spanish-only medium students significantly (Encalada et al., IIDB, 1999. in Mathematics and Social Studies, and at the same time Oven eh last yD r,19 they do not fall behind in Spanish language despite spending Over the last ten years, a less time in the language ( Dutcher, 1995). Despite these strategy of intercultural and positive signs, many countries might be reluctant to adopt bilingual education has been this model. One major concern is the cost of this type of implemented. The National bilingual education. Many experimental programs show the Directorate of Bilingual and cost higher than a nonnal programr... .However, the Intercultural Education has effectiveness of bilingual education programs should be made a major effort to develop a explored further. The initial study on the cost effectiveness of *. ~~~~of bilingual education in Guatemala shows very positive system of bilingual and indications in tenns of cost-savings (Patrinos & Velez, intercultural education with an 1996). Availability of competent bilingual teachers, appropriate pedagogy and affordability for development and production of bilingual administration. learning materials, and acceptance and strong support for the program from the minority communities are fundamental Despite many achievements factors for their success. neither the internal and external Source: Win Aung (1999) efficiency of the program is known, nor its impact on improving the student learning in indigenous communities, the relationship between bilingual education and productive work, and the perception of the communities concerning the curriculum, teacher development, and the designation of educational authorities and their role in the development of the communities ). These programs need to be evaluated before large-scale implementation Employing mother tongue instruction in the early years of schooling in other countries in the region, has been shown to increase school effectiveness through higher learning achievement thereby increasing the internal efficiency of the educational system by producing lower dropout and repetition rates. While bilingual education is undoubtedly more expensive than traditional Spanish-only schooling, learning outcomes are often greater; greater learning outcomes are often reflected in lower repetition rates which, in turn, reduce the cost of schooling. Studies such as those of the bilingual 10 education program in Guatemala demonstrate that higher learning outcomes and improved efficiency can more than offset the cost differential involved in bilingual; education. (see Box 1). IV. Expenditure and Finance Central Government Expenditure Public financing for education in Ecuador comes almost exclusively from the central government. Public spending on education as a percentage of GDP has remained relatively steady (3.1%) over the last 5 years and has remained a priority within the social sector (Figure 18). In 1995, the distribution of education spending is 35% for primary, 35% for secondary, 23% for higher education, and another 7% for administration, literacy campaign and infrastructure. This composition of the education spending can be translated into relatively high unit cost at the high education level (US$1300) and low unit cost at the primary level (less than US$200) (Figure 19). The unit costs also have remained rather steady over the last few years. More than 99% of the education budget is allocated for recurrent costs. In both primary and secondary education, more than 99% of the recurrent budget is spent on teachers and other personnel, leaving less than 1% for non-salary expenditures on instructional materials and other learning aids (Figure 20). 11 Figure 18. Social Sector Expenditure as a % of GDP 3.50% - ... ............ . .__.. 3.00% , EWCATION 2.50% -BIENESTAR SOCIAL 2.00% - TRABAO 1.50% -D _________ _________ _______BONO SOLtDARIG 1.00% . ............... .............. 30 .OTH. -allER SOCIAL PROGRA. 0.50% - S ... ;._ ____s , (CONAMU/CONSEIOS) 0.00% 1995 1996 1997 1998 1999* Source: Ministry of Finance Figure 19. Unit Costs by Education Level 16000 1,400.0 1,200.0 1.000- ,n 800.0 - Ttwy 400.0 200.0I k aN Source: SINEC, MOF, World Bank Calculations Figure 20. Central Government Expenditure in Education by recurrent and investment spending - 1998 100 900 * 70 so = | NON-PERSONNEL so 50 _ _* PERSONNEL 30 20 10 0% FMUMW NVEETMM41 12 Household Expenditure Private financing of education is substantial in Ecuador. In 1998, private financing of education at all levels amounted to 3% of GDP, very similar to the level of government financing which is 3.1% of GDP. This level of private financing is substantially higher than for example in OECD countries where on average the government spends 4.7% in education and private financing is 1.2%. This finding may indicate the need to be cautious in devising cost recovery policies to mobilize more resources for education. Distribution of Education Expenditures The distribution of public expenditure on education can be examined in several ways. First, we will look at the distribution across geographical areas to see whether poor areas receive more than its share of population; Second, we will use benefit incidence analysis to examine the share of public expenditure received by the extreme poor, regardless of location. By canton The SEDES organization grouped all the cantons in Ecuador into 5 categories, according to its average poverty status: priority 1, 2, 3, 4, and 5 areas. The priority 1 and 2 cantons are where the percentage of population who are poor exceeds 80%. To examine how public expenditures in primary and secondary education are distributed across the cantons, we use the personnel cost as a proxy for total expenditure, since we learned that more than 99% of primary and secondary expenditure is on personnel. Table 4 shows that there is much room for improvement in targeting the poor priority 1 cantons. So far, the share of public primary and secondary education budget received by these cantons roughly correspond to their population share (15%). A more progressive public expenditure scheme would have devoted a much higher percent of public money (in relation to the population share) to poor areas and a much lower percent to rich areas. Table 4. Share of Primary and Secondary Public Expen iture by Canton Priority 1 Priority 2 Priority 3 Priority 4 Priority 5 Population share 0.15 0.12 0.10 0.26 0.38 Primary expenditure 0.15 0.21 0.11 0.09 0.45 share I__ __ _ I _ _ __ Secondary expenditure 0.16 0.26 0.11 0 0.0938 sh are__ _ _ _ _ _ _ _ _ _ _ __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ By poverty Benefit incidence analysis shows that public spending on primary education is progressive when mapping to consumption quintiles. The bottom 20% of the population enjoys 24% of total public subsidy in primary education whereas the top 20% of the population receives about 11% (Figure 21.A). 13 Figure 21: Distribution of Government and Private Spending on Primary Education (By Consumption Quintiles) 0.5 04 0.3 0.2 _-t A B C F13Z, d kw *l d -a4th *S I Source: LSMS 1998. However, public spending on all levels of education (including secondary and high education expenditures) is not distributed equitably. While the top 20% of the population get 25% of public subsidy in all levels of education, the bottom 20% only gets 12% (Figure 22.A). Figure 22. Distribution of Government and Private Spending on Education (By Consumption Quintiles) 0.50 | 0.40 - 0.30 _ 0.20 1 __- 0.10 _ A B C Dii3t 2d 03d 04h *541 Source: LSMS 1998. Compared to Peru, Argentina, and Dominican Republic where the bottom 20% of the population receive respectively 21%, 26%, and 23% of public subsidies in all levels of education, Ecuador's public financing across all levels of education is much less progressive (Table 5). 14 Table 5. Public Expenditure in Education Received by the Bottom Quintile I Expressed as Percent of Total E nditures) Education Level Ecuador Peru* Argentina** Dominican _ I Republic*** Primary Education 24 28 33 29 All levels 12 21 26 23 *Wu, Kin Bing. 1999. "Sector Study on Education Finance and Rural Education in Peru." The World Bank. Washington, DC. Febnuary. "The World Bark. 1999. "Argentina Poverty Report: Poor People in a Rich Country." The World Bank Washington DC. June. ***The World Bank 1999. "Dominican Republic: Social and Struchiral Policy Review." The World Bark Washington, DC. July. Source: World Development Indicators, 1999 The inequitable distribution of education financing in Ecuador is mainly due to the lower levels of enrollment of the poor at higher levels of education (lower secondary and beyond) where per pupil subsidy is higher. Strategies to reverse this trend must be considered either through increasing enrollment of low income groups and/or through reducing the public subsidy at these levels. Another source of inequity comes from private expenditure on education. The richer spends substantially higher amounts of private money on their children's education than the poor. Benefit incidence analysis of total expenditure in education (both public and private) shows that even at primary level, the inequity is substantial: the bottom 20% receive about 12% whereas the top 20%/o receives more than 40% (Figure 21.C). This is clearly a contributing factor to the disparities in learning achievement between the poor and the rich. V. Benefits of Education About 30% of those who are between 15 to 65 years old and who have already left school are working in the formal wage sector. Returns to education are estimated using the age earnings function method (Table 6). Job experience, years of education, and hours worked per week are employed as predictors for monthly wage not including other forms of compensation (in its natural log form). We find that each additional year of education is associated with 11% of premium in monthly wage - private rate of return to education. When looking at men and women separately, we find that private return is higher for females than males: for every additional year of education, females enjoy almost 15% of monthly wage premium, as compared to less than 11% for males. These figures are consistent with the findings from "Ecuador: Is society getting What It Needs from Public Spending on Education", using the 1994 LSMS data. 15 Table 6. Private Rates of Return to Education ________ ________ All Male Female Constant 9.99 (0.12)*** 10.4 (0.14)*** 9.58 (0.19)*** Years of education 0.114 (0.00)*** 0.107 (0.00)*** 0.148 (0.01)*** Experience 0.04 (0.00)*** 0.04 (0.00)*** 0.04 (0.01)*** Experience squared -0.0005 (0.00)*** -0.0004 (0.00)*** -0.0005 (0.00)*** Hours/wk (log) 0.48 (0.03)*** 0.43 (0.00)*** 0.42 (0.05)*** Source: LSMS 98. When looking at the wage premium associated with different levels of education, we find that from 1994 to 1998, there is a decline in the rate to primary (from 40% to 33%) and secondary education (from 54% to 48%). However there is a significant increase in the return to higher education (from 48% to 73%) (Table 7). These changes may have contributed to the increasing inequity of the Ecuadorian society as the poor are not benefiting as much from the higher education. Table 7. Earnings Premium Associated with Levels of Education Education Level 1994 1998 Primary (vs. no education) 40 33 Secondary (vs. primary) 54 48 Higher (vs. secondary) 48 73 Source LSMS 1998 VI. Institutional Issues The analysis above suggest that there are substantial problems in the Ecuadorian education system related to equity of access at all levels, but particularly in the lower secondary and above. The low income groups, the rural, and the indigenous have significantly lower access, attainment, and achievement. In order to rectify the situation, the Ministry will have to develop a clearly strategic framework in which these issues of equity are prioritized and vulnerable groups are targeted. An examination of the organization of the Ministry reveals that there are weaknesses in this regard; there remain fundamental institutional weakness that need to be addressed before progress can be made in the education sector. Lack of Comprehensive Strategic Framework for Reform The Ministry of Education has been engaged in a process of reforming the education system; proposals having been submitted to Congress but approvals are still pending. Although there is some degree of consensus on the need for major organizational reforms and a number of principles have been articulated on developing a strategic plan, the government is lacking a comprehensive strategic framework for these reforms giving a clearer focus on reducing inequities and internal inefficiencies in the system with specific strategies and implementation plans for addressing these 16 dleficiencies. In line with the strategic framework, the capacity of MEC needs to be strengthened for national level policy making, setting policies for the equitable and efficient resource distribution. Lack of Accountability and Transparency The education system is beset with problems that inhibit a rational allocation and utilization of public expenditures. The examples of budgeting and accounting, and the supply and remuneration of teachers demonstrate the problems of the lack of accountability and transparency in the sector and the need for substantial reforms in this area. Budgeting and Accounting The current budgeting and accounting system, in particular, leads to a lack of accountability and transparency The problem is that no single entity is responsible for the sector as a whole nor is there any entity that reviews the combined effect of the various subsectoral budget proposals or their intrasectoral distribution. Investment proposals are vetted by the Planning Secretariat (SEGEPLAN) and then submitted to the MOF. Too many budgetary units exist with each secondary school have a line item in the national budget. However, MOF is the only sector which has access to all the information necessary to make an analysis of intrasectoral allocations (although such an analysis is not considered to be within the responsibilities of the MOF). As a result, allocations tend to be the result of influence rather than sectoral policy and the complete information on budgetary allocations only emerges after the budget has been sent to Congress for approval. The allocation of funds is made purely on the basis of previously budgeted levels (with allowances for salary increases, estimates of growth and inflation). According to the system of decentralized responsibility, expenditure for primary education is authorized and disbursed at the provincial level of the MEC. However, the effect of this is that, instead of bringing the decision-making closer to the level of the school and its needs, an additional level of bureaucracy is added to the process, resulting in cumbersome and inefficient processes. The need of the schools are not met and no school can make decisions related to key decisions ( such as the selection of teachers) since all are taken by the Provincial Directorate. MEC is never in possession of full information on education expenditure. Data from each secondary school and CEM is unprocessed at provincial or national level and forwarded to MOF. Therefore MEC is unable to compute, for example, a total public wage bill for education. Teacher Remuneration and Supply The remuneration system for teachers is overly complex, non-transparent, and unrelated to teacher performance. The system consists of extra salaries, bonuses, and allowances which, in their complexity makes the system conducive to manipulation and 17 corruption. Teachers are classified into 16 categories which have little relevance to responsibilities, function, or performance, but only to length of service. This results in considerable confusion concerning salaries, with individual teachers not being aware of how their salaries are computed or what their monthly salary should be. Again, teachers are hired by the provincial directorate through a very lengthy and cumbersome process and, since the selection is made far from the schools, teacher skills are often not related to the needs of the schools. The bureaucratic nature of the system is compounded by the fact that there neither a central registry of qualified teachers nor clear selection criteria for recruitment. V. Strategies for Reform The following five main strategies are suggested for addressing the problems identified in the situation analysis of the education sector made in this paper: (i) Developing a clear strategic framework: At the outset, there is an urgent need to clarify reform goals and strategies and to achieve a consensus through a participatory process within the country. These goals and strategies need to take into consideration the considerable inequities which exist in the education system, particularly with regard to the access, attainment and learning achievement of the rural poor and indigenous population. (ii) Institutional reform: re-organizing the Ministry of Education The challenge facing Ecuador is to define the responsibilities of the central and sub-national governments in line with the principles of decentralization which it has adopted. The role of the central ministry would therefore become one of selecting what it can do best and in promoting the development of new decision-making processes by clearly redefining and articulating the roles and responsibilities of local and central governments. The legislative process towards decentralization has to be accompanied by a real process of devolving power closer to the local and school level in order to serve the needs of the school; this could be done by increasing school autonomy, the role of the private sector, and community participation. Related to this, there is the need to streamline the budgeting and accounting mechanisms to improve accountability and transparency. One entity (preferable the MEC) needs to have an oversight of the budget and expenditures in order to ensure that it is in line with sector strategy and to facilitate the desired intersectoral distribution. Finally, a new system of teacher supply, remuneration and incentives need to be designed and implemented in the interest of accountability and transparency. A management information system, including teacher management, needs to be set up to improve the flow of information and to facilitate decision making. 18 (iii) Reducing the inequity in access to and quality of basic education, particularly for the rural poor. According to the analysis made above, the following four interventions are suggested for increasing the access and learning outcomes among the rural poor: Early Childhood Education Investments in early childhood care and education targeted on the rural poor and indigenous, in conjunction with parenting and caregiving Box 2. Colombia's Escuela Nueva: programs, can be a way to A Strategy for Reaching the Rural Poor compensate for the deficits in the Based on its 1978 study detailing major weaknesses in the provision of experience, nutrition and health of rural basic education, the Government of Colombia gave top priority to the children from these rural education and produced a ten-year rural education program which households and prepare them for built heavily on the Escuela Nueva program. Escuela Nueva was created in formal schooling. Approaches 1976 after a decade of experimentation and was supported at various times such as community-based forms by UNICEF, USAID, IDB and the World Bank.. By 1978 more than 500 such as community-based forms schools were involved and this gradually expanded under World Bank of day care, supporting and financing to 17, 948 schools by 1989, serving 800,000 students. Major educating parents will need to be characteristics of Escuela Nueva include: One or two teachers offering all based on a thorough analysis of five years of primary education in one or two multi-grade classrooms; how children are supported and flexible promotion, special instructional materials for individual and group cared for in the local cultural work and teachers' guides, curriculum relevant to rural community, school government to ensure participation of students, parents, and community; context and an identification of study corners and class libraijes; self-monitoring mechanisms for students. the gaps which exist in such Rojas andd Castillo (1988), Psacharopoulos, Rojas, and Velez (1992), and support and care. McEwan (1995) have found that ES schools had significantly improved student outcomes, reduced dropout rates among the rural poor, as well as Bilingual Programs for student and community participation as measured by activities such as adult education, agricultural extension, athletic competitions, health Indigenous Education: campaigns and community celebrations. Social self-esteem and civic behavior higher than in traditional schools. Moreover, EN schools have There is an urgent need to statistically significant effects on Spanish and mathematics achievement in assess the impact of the System of third and Spanish in fifth grade. High benefits relative to cost (parent Nacional Educacion Intercultural volunteerism reduces monetary outlays for building and construction and Bilingue. Since the inception of contributes to gains in student achievement). the bilingual education system in 1988, there has been no evaluation of the achievements and problems to allow for the introduction of corrective action in the future. Such an evaluation needs to be independent of DINEIB and the Ministry of Education to guarantee objectivity and to recognize the opinion of primary stakeholders such as the children, teachers, directors, parents, community leaders, technical experts, and Ministry officials. Strategies of bilingual education may serve to increase learning outcomes among the indigenous groups (see Box 1). 19 Demand-side financing Box 3. El Salvador: EDUCO A new vision of shared responsibility for meeting educational needs has had a profound impact on education in El Salvador in the 1990s, resulting in significantly expanded coverage at the preschool and basic levels. Facing the daunting task of educational sector reforn, the Salvadoran Ministry of Education sought an innovative and effective approach to expanding coverage in rural and marginalized areas. EDUCO ("Education with the Participation of the Community") provided a model for that approach, creating a framework and processes for responsibility to be shared among government authorities, community NGOs, and parent associations. Originally initiated in 1991, EDUCO works to provide preschool and basic education to rural and needy communities through a decentralized system of service provision, encouraging community participation in the educational process. Program activities include teacher training, organization of parents and teachers into community education associations (responsible for the management of educational resources at the local level), and development of didactic materials for parents and teacher guides. The program's administrative component has focused on design and actualization of legal and financial mechanisms which enable educational sector decentralization. The outcomes of EDUCO programs demonstrate both quantitative advances in coverage and qualitative improvements compared with traditional education. Since 1992, EDUCO has provided basic education for 168,672 children, employed 3,871 teachers, and formed 1,700 community associations. The education provided in EDUCO communities is characterized by lower rates of teacher absenteeism, increased amount of instructional time, a higher average mathematics achievement test scores in first and second grades, and greater parent involvement in the education of their children than in the traditional school system. Coordination of efforts at national, regionaL and local levels through EDUCO programs has significantly improved the education system in El Salvador, and today EDUCO continues to expand, spreading the benefits of community involvement in the education to children who might never have had educational opportunity through the traditional system. A high priority need to be assigned to developing demand-side financing mechanisms to attract and retain children in school. Interventions to reduce the direct and indirect costs of education for the rural poor will need to be carefully targeted. Possible interventions could be in subsidizing families for lost earnings (opportunity costs) through the provision of scholarships, and experimenting with flexible school calendar school hours because of seasonal work obligations. Financial incentives could be given to parents to send children and keep them in school through the provision of allowances accumulated as the child transfers from one grade or level to another. Such interventions would be targeted particularly at the very poor and/or indigenous populations who would be most vulnerable in the time of macroeconomic crisis and, as the data above indicates, whose participation and achievement in school is most affected. Sources: Guido Bejar (1997) and Meza (1997). 20 Enhancing community involvement Building on the experiences with the school network programs (CEMs), strategies need to be developed to developing a vision of shared responsibility for meeting the needs of the rural poor. The example of Educo (see Box 3) may serve to illustrate a particular strategy targeted at the rural poor and marginalized areas in which parents, community, NGOs, and government work collaboratively to increase participation and quality in pre- school and basic education. (iv) Increasing access/participation at lower secondary level especially in poor rural areas Building on the experiences of decentralized programs in the rural and peri-urban areas, the government should explore strategies for making access to education more equitable for the rural poor especially to increase access and achievement in lower secondary schools. Again, effective strategies could be in the area of demand side financing such as grants and allowances directed at the most poor. In a period of macro-economic crisis, the opportunity costs of having children in school are most felt in households with children of lower-secondary age and higher. The level of lower-secondary is where the gap in enrollment between the poor and non-poor is particularly noticeable and the disparity widens from the age of 11 (see Figure 5 above) and this would seem to be the critical stage for demand-side interventions of this kind. A further strategy which could be explored is that of combining distance education modalities with conventional on-site education at the levels of lower-secondary and secondary such as been successfully demonstrated with Telesecundaria in Mexico (see Box 4). The use of distance education can be used to overcome the problems of access in remote areas as well as problems related to the supply and quality of teachers in such areas. 21 (v) Reforming higher education: Box 4. Mexico: TELESECUNDARIA: Increasing Access to Secondary Education for the Rural Poor Telesecundaria provides a combination of distance and on-site secondary education for grades seven to nine in rural areas of Mexico where schools or teachers are in short supply and education is of poor quality. Educational television programming is broadcast through EDUSAT, Mexico's educational broadcast system, and transnitted through Solidaridad 1, a government satellite, to 13,785 schools in two sessions daily. A school can be equipped and wired to receive Telesecundaria broadcasts for a total cost of US$2,000. Technical adaptations are made where necessary. For example, 10 percent of Telesecundaria schools use solar power. Telesecundaria schools are initiated at the request of local communities that demonstrate that 15 or more primary school graduates will participate and can identify a location for school facilities. The Mexican Ministry of Education then provides a teacher, a television set, a satellite dish, instructional program, and textbooks. Today, 16 percent of secondary students in the country attend Telesecundaria programs instead of traditional or technical schools. Telesecundaria programs emphasize active learning and collaboration, parental participation, practical application of lessons, and student presentations to their community. The cuniculum is designed to be adaptable to student needs, local context, and available resources. The teacher's guide provides descriptions of possible limitations teachers may face given the availability of learning materials and tools, and suggests alternatives and ways to overcome those limitations. With support from the Wodd Bank, the Mexican Government has initiated both a high school (grades 10- 12) and junior high school program for adults. In future, Telesecundaria schools may access the Internet via Red Escolar, which began in 1997 as a computer technology project aimed at seventh to ninth graders. Telesecundaria has inspired similar educational technology projects in Costa Rica (1997), Panama (1995), and El Salvador (1998). A Telesecundaria program aimed at Latino communities in the United States is planned for the summer of 1998. Jose Calderon Strategies are urgently needed to increase the internal efficiency, quality, relevance and equity of the diverse HE programs in the country. In particular, a system need to be developed for a system of performance-linked subsidies to higher education institutions based on quality, program relevance to the labor market and the efficiency of financial support mechanisms for encouraging the participation of students with limited economic means. In addition to this, the development of a system of information would enable the dissemination of data on standards and increase the accountability and transparency at this level. Given the high degree of inequity which exists among the income groups for access to higher education (see Table 1), the increasing private rates of return for this level of education, and the vastly higher unit costs, cost recovery mechanisms should be explored to release more funds from tertiary education to basic education in order to increasing access, equity and quality. 22 Annex 5: Health in Ecuador Patricio Marquez and Edmundo Murrugarra Health Status of the Ecuadorian Population. Health indicators of the Ecuadorian population rank among the lowest in the Americas. Table 1 shows that infant mortality was 39.4 per one thousand births in Ecuador, higher than that of Chile (13), Costa Rica (11.8), Cuba (7.9), Colombia (24) or Venezuela (22).' The mortality rate of children under 5 years old is 48.3 per thousand, larger than that observed in Chile (14.8), Costa Rica (27.3), Cuba (10.7), Brazil (44.2), or Paraguay (47). The maternal mortality rate in Ecuador is also high (159 per 100,000) compared to Chile (25), Costa Rica (29), Cuba (33), Venezuela (56), Brazil (119), and Colombia (87). Only Bolivia and Peru evidence somewhat worse conditions in mortality rates and incidence of diseases, while Paraguay evidences weaker health infrastructure and services than Ecuador. The health profile is characterized by the coexistence of infectious and communicable diseases and non- cofmmunicable conditions and injuries. The poor suffer higher incidence rates in many types of diseases as a result of higher exposure to risk factors and limited access to effective medical care. Wide Gap in Access to Health Services. Access to health care is limited for about 3 million Ecuadorians, or about 30% of the population. The Ministry of Health (MOH) provides regular access to about 45% of the population, and other public and charitable organizations provide health care to another 5%. Formal health insurance coverage (public or private) is very low, particularly in comparison with other Latin American countries: the medical program of the Ecuadorian Social Security Institute (IESS) covers only 10% of the population, the Peasant Health Insurance Program (Seguro Social Campesino) reaches an additional 8%, and private health insurance companies and the armed forces each cover 2.8% and 0.7%, respectively (see Table 2). Family members are not generally covered through IESS, which is one of the reasons why health insurance coverage is quite low in Ecuador. Another reason is the low rate of IESS coverage, 28%, among the economically active population. The Seguro Social Campesino covers about 16% of the poorest quintile of the population. The IESS covers basically the top two quintiles of the population, and a small fraction in other quintiles. Overall, 79°/0 of the population is not covered by any health insurance in Ecuador. l Health in the Americas (1998). Table 1. Health Indkators for selected Latin Amercan Countries Chile Cuba Costa Ecuador Bolivia Peru Colombia Venezuela Argentina Brazil Paraguay Rica Mortality rates Infant (per 1,000 live births, l.b.) 13.0 7.9 11.8 39.4 59.0 43.0 24.0 22.0 20.9 39.8 36.0 Children 1-4 (per 1,000) 14.8 10.7 27.3 48.3 81.0 68.0 29.6 25.0 24.3 44.2 47.0 Maternal (per 100,000 l..b.) 25 33 29 159 390 265 87 56 44 114 123 Infrastructure (% pop) Access to drinking water 85 92 100 70 61 70 80 79 65 69 39 Access to excretal disposal 98 90 100 57 46 74 66 72 75 67 32 Hospital beds (p/1,000 pop) 3.1 6.0 ... 1.6 1.4 1.8 0.8 2.3 4.6 3.5 1.5 Physicians per 10,000 pop 11.0 53.0 14.1 13.2 5.8 10.3 9.3 24.2 26.8 12.7 4.9 Diseases Tuberculosis incidence 10.3 7.6 16.0 60.4 91.5 112.8 24.5 - 16.4 27.6 18.0 (per 100,000 pop) Malaria pop. in risk areas, % - - 36.0 24.7 40.7 45.6 39.5 4.2 0.7 4.6 29.6 Source: Health in the Americas (1998). Notes: Mortality rates are from 1997, and maternal mortality from 1992-1997. Infiastructure indicators are from 1995 for all countries. Tuberculosis information (BK+) is for 1996, and Malaria for 1997. 2 Table 2. Health Status and Use of Health Care facilities Per capita Expenditures Quintiles Area of residence Overal Poorest n m IV Richest Urban Rural 1.Children under 5 years old Immunization (%/*) Tuberculosis 88.1 96.2 95.3 98.9 98.2 93.2 92.4 92.8 Diphtheria-.Tetanus-Pertussis 85.5 87.8 92.3 92.6 97.6 89.6 89.5 89.5 Poliomyelitis 84.7 85.6 91.5 91.0 96.9 87.9 88.7 88.3 Measles 69.8 71.2 75.4 78.3 83.0 74.9 72.2 73.7 Sickness in last 2 weeks (%) Had Diarrhea? 25.2 27.4 26.0 21.2 23.3 24.0 26.6 25.1 Location of service, if sick Public Hospital or Center 15.1 19.1 21.9 15.0 12.9 17.4 16.7 17.2 Private Clinic orPhysician 6.4 11.1 16.0 24.2 37.7 18.9 9.6 14.4 Pharmacy and other 14.2 16.9 17.3 12.4 08.5 14.8 14.9 14.8 At home 64.3 52.9 44.9 48.3 41.2 48.9 58.9 53.5 Any Respiratory Sickness? 39.5 45.7 43.0 47.6 48.8 45.0 42.3 43.8 Location of service, if sick Public Hospitat etc 21.7 27.5 20.8 14.3 7.0 21.5 18.1 20.2 Private Clinic 5.9 10.9 18.3 24.3 42.6 21.2 9.8 16.3 Pharmacyand other 16.8 12.5 17.2 15.3 5.7 16.7 10.9 14.3 At home 55.6 49.1 43.8 46.2 44.7 40.6 61.3 49.3 2. Individuals older than 5 years old Sick during last month % 44.7 44.8 47.9 45.1 42.4 43.0 47.7 | 45.1 Had malaria?% J 1.3 0.7 0.4 0.1 0.1 j 0.3 0.9 f 0.7 Type of attention, if sick %_ _ Self-medicated 31.7 29.3 30.6 24.9 23.8 26.2 31.2 28.6 Public 14.9 15.5 16.0 18.6 15.5 15.7 16.3 15.9 Private 10.3 17.1 18.8 25.7 31.2 23.0 15.6 19.4 Pharmacy, other 9.4 15.8 16.6 15.9 12.5 17.2 10.4 13.9 Access to services, if sick Transportation tiiiie(iniutes) | 41.6 35.7 41.7 37.3 35.7 1 36.0 41.7 38.4 Waitingtime (minutes) _ 57.8 42.4 36.2 32.0 33.1 | 25.5 60.0 J 39.8 Health Insurance enrollment_ Private 0.1 0.4 1.3 2.7 12.9 4.8 0.3 2.8 IESS 1.0 3.2 6.9 12.5 23.5 12.5 2.6 7.9 Peasants Social Insurance 16.2 12.8 10.8 3.7 1.9 0.6 21.5 10.1 ArmyandPoliceInsurance 0.1 0.1 0.3 1.5 2.5 1.2 0.1 0.7 None 82.3 83.6 81.1 80.4 63.8 82.3 75.7 79.2 3. Reproductive Health All females 14-49 HadPapSmear% 21.7 31.5 42.3 48.9 54.4 47.9 26.9 39.3 Ever heard information on 53.5 71.5 79.6 84.7 90.9 85.6 60.8 75.5 family planning? % Females 15-20 Had Pap Smear? % 0.9 5.9 5.1 7.9 5.6 6.8 2.8 4.9 Ever heard information on I 34.3 58.6 65.4 71.8 79.8 74.7 43.3 60.0 family planning? % I *_*_*_* Source: Enctesta de Niveles de Vida, 1998 3 Health Care Spending. In comparison with other Latin American countries, the proportion of GDP devoted to health care is among the lowest. Total health expenditures in the mid-1990s are estimated at about US$78 per capita, or 5.2 percent of GDP. Recent estimates indicate that direct household expenditures amount to 37% of overall health spending--more than 70% for purchasing pharmaceuticals--while MOH spending accounts for 28% and EESS for 24%. The share of health care spending corresponding to the MOH has significantly decreased since the mid-1980s: as a proportion of GDP, from 1% to about 0.8%, and as a proportion of total government spending (excluding debt Figure 1. Public Expenditure in Health Ecuador 1989-1999 6.5% 8.0% 5 .5%- .0% -7.. 4.5% - i1Y89 1991 1993 1995 1997 1999 Soue: Cuentas NadonAles, Ecuador. service), from 8.2% to 5.5% (see Figure 1). As a result, MOH real per capita spending has declined by approximately 40%. While the general medical care program under IESS spends about US$87 per affiliate, MOH must stretch its resources much further, allowing average per capita spending of about US$20. The Ecuadofian poor suffer a disproportionally high incidence of disease, causing a correspondingly disproportional spending of their income on health care; the urban poor spend an average of 12% of their total household expenditures in health care and the rural poor 170%, as opposed to less than 10% for the non-poor in both areas. MOH expenditures can be grouped in three categories: Operational; Equipment, Machinery and Infrastructure; and Transfers and Loans. Operational expenses represent 85%-90%/ of health spending, while Equipment, Machinery and Infrastructure account for 6.5%, and Transfers and Loans for about 5%. Desegregated data shows that expenditures in medicines and other medical supplies have decreased from 7. 1% of total MOH expenditures in 1993 to 0. 2% in 1996. Since 4 1993, personnel (not administrative) expenditures have remained around 64%. This suggests that expenditures in medicines and other medical supplies have been reduced in order to keep up personnel expenditures, hence reducing the quality of care provided in public facilities and transferring their costs to the patients in form of user charges. Access to Health Services Among the Poor. Access to private and public health services is limited for the poor. Table 2 describes the type of health care that sick individuals are receiving by expenditure level and location of residence. While traveling times are not substantially longer for the poor and rural population, waiting times are much longer, raising important concerns about the quality of care provided. On average, rural individuals take 6 additional minutes to get to a health center, but then must wait an average of one hour compared to a 25 minutes wait for their urban counterparts. Self-care and self-medication are widely common in Ecuador, especially for the poor. Self-medication and health counseling at the pharmacy represent the most chosen health alternatives for the extreme poor in Ecuador (41% overall), and these practices are very common in the treatment of children who suffer from diarrhea or respiratory sickness, representing 80% and 72% of the health care contacts, respectively, for these illnesses. (see Table 2). Important differences are observed in reproductive health practices and family planning information across socioeconomic levels. In the poorest population quintile, only one out of five women in reproductive age have ever had a Pap smear, while half of those in the richest two population quintiles have been given the test. The statistics on "'family planning information ever heard" indicates that only one-half of poorest females have such exposure. These figures are even more dramatic among women between 14 and 20 years old, which represents a latent problem to be addressed. These differences partially reflect the reproductive health care practices in rural areas, and the different access to information as a function of socioeconomic status. Substantial differences in the ability to detect sickness are suggested from the uniform self-reported morbidity indicators across socioeconomic groups. As the poor and extreme poor are generally less educated they may tend to report lower morbidity rates, or to delay treatment due to financial constraints. In this sense, health status of the poor is not only lower but also underestimated, and addressing this problem may help in reducing the costs of future necessary medical care. Equity of Health Expenditures Two types of evidence are provided to evaluate the equity of health expenditures. First, the incidence of public health benefits across population quintiles is examined. Then, a canton-level data set on public health expenditures is analyzed to examine whether the government spends more in poorer cantones or in those with a larger fraction of indigenous rural population. Health Expenditures Across Population Quintiles Evidence a Regressive Pattern Using measures of spending patterns across population quintiles and the 1998 5 LSMS, the aggregate distribution of health spending across population quintiles is shown in Table 3. Each row indicates how each type of spending (private or public) is allocated across different quintiles. The private share is larger than the public share among the poorest and richest population quintiles, suggesting that both groups rely on their own resources to deal with health problems. The problems for the poorest quintile might be access (time, transportation, travel,) and cost (foregone wages and fees), while for the richest quintile the challenge might be making decisions on the basis of perceived quality of health care. Both public and private spending are highly unequal. Approximately 51% of private spending and 38% of public health spending is done by the fifth quintile alone, and while public spending is perhaps somewhat less unequal, the resulting aggregate health spending is very unequal. The poorest 40% of the Ecuadorian population get less than 20% of national health spending, while the wealthiest 20% of the population gets about 45% of these resources. Table 3. Share of National Health Spending Across PCE Quintiles Ecuador 1998 Poorest Q2 Q3 Q4 Richest A. Public Spending 7.6% 11.6% 17.6% 25.1% 38.1% B. Private Spending 9.0% 10.0% 11.0% 19.0% 51.0% C. Aggregate spending 8.4% 10.7% 13.8% 21.6% 45.6% Source: Gragnolati (1999) and ECV, 1998. Limited Targeting of Public Health Expenditures to Poor Cantones The allocation of public resources to finance MOH's operational expenditures does not have any significant relation with the incidence of poverty across cantones. However, there is some evidence that those cantones with higher poverty incidence have also higher per capita health investment expenditures, but these items constitute only 11.5% on public spending in health in 1998 (see Figure 2). Recent Reform Measures. Measures to Increase Effectiveness of Public Spending and Target the Poor. Many of the poor cannot wait for the long-term overhaul of the health system. While most are able to bear the costs of basic health care for a period of time by using informal credit arrangements or sacrificing other goods--often food--the current situation is unsustainable. If the deep cuts in the MOH budget are not redressed, access to basic health services by the poor would be further curtailed. In the short term, in conjunction with a greater and sustained commitment of resources for the MOH, the effectiveness of existing levels of funding can be improved. 6 Reforming health programs has the greatest potential for directly improving the health and nutrition status of the poor. Given limited resources, the most economically efficient way to reduce the disease burden in Ecuador is by supporting programs that provide the largest reduction in mortality and morbidity per sucre spent. In view of Ecuador's health profile, MOH efforts, both in terms of traditional public health activities and personal health services, need to be targeted at specific maternal and childhood problems, as well as particular adult health programs. Building upon the experience from the World Bank-financed II Social I)evelopment Project-Health and Nutrition (FASBASE Project), the Government has defined a program of essential services formed by those interventions that cost- effectiveness analysis has shown to be the very best investments for health: a set of key public health interventions, such as the expanded program of immunizations, micronutrient supplementation, and health and nutrition education; as well as basic clinical services, such as integrated management of the sick child, pre-natal, child delivery and post-natal care, promotion of reproductive care, including family planning services, treatment of sexually transmitted diseases, and treatment of infections and rninor trauma. This approach to expand health care coverage would help avoid the pitfalls of classical universality, whose promise of "everything for everyone" has proven to be unsustainable even in the richest countries. Public finance of such a program in Ecuador is an effective mechanism to reach the Figure 2. Poverty and Public Expenditures in Health Services Ecuador - 1998 300 U 250 3 200 a 150 . =A 100 U , U 50 0 IL, - 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00 90.00 100.00 Source: SIISE (1999) Poverty Incidenoe SISSE (percent) 7 poor since they are disproportionately affected by the disease burden of the conditions listed above and, because of large family sizes (on average, 3.6 children per woman), they would benefit disproportionately from prenatal and post-natal care, child delivery, and childhood services. In addition, by giving MOH and other public programs expenditure priority for a program of essential services targeted at risk groups, e.g., pregnant women and their unborn, nursing mothers, infants and young children, the foundations for all subsequent human capital formation in Ecuador can be built. To maximize the impact of this approach, efforts need to be concentrated on the poor regions of the country. Organizational Measures. A major change in the government view of the health sector was stated in the new 1998 Constitution. It notes that the State guarantees the promotion and protection of health through food safety, water supply and basic sanitation, and the permanent and uninterrupted access to health services, in accordance with principles of equity, universality, solidarity, quality and efficiency. In addition, public health services will be provided only to those unable to pay for such services and no one would be denied emergency services. It is also stated that the National Health System will be formed by public, autonomous, private and community-based entities and such system will operate in a deconcentrated, decentralized and participatory manner. The centralized functions of MOH cover the regulation, control and delivery of services to the population in the prevention and cure of diseases. Consistent with the institutional framework set forth in the Health Code, the Municipalities Law gives powers to the Municipalities, within their jurisdictions, for regulating and delivering water for human consumption, and for sanitation, and, in coordination with the MOH, for regulating hygiene and other related health matters including the hygienic and healthy operation of businesses dealing with food, and public buildings in general. During the 1990s two complementary laws were enacted which provide an enabling framework for decentralization and implementation of the organizational reforms in the health system: the 1993 Law of the Modernization of the State, Privatization, and Delivery of Public Services by the Private Sector and its Regulations; and the 1997 Decentralization and Social Participation Law. Under these legal instruments, the Central Government transferred its powers and functions under the following regimes: * Decentralization, which consists of the delegation of political, economic, administrative and financial management powers and duties from the Central Government to sub-national governments (provincial or municipal). Decentralization is defined as "a transfer of functions to a functionally or geographically decentralized entity, and such decentralization includes also the power to create new entities to discharge functions which were originally centrally assigned." Executive decrees are required for such decentralization. * Deconcentration, which consists of the delegation of administrative and financial functions from the Central Government to its own dependencies. In this framework, the Ministries are required to delegate, within economic or geographic regions, and through 8 ministerial resolutions, its powers and responsibilities with the only specific requirement that: (a) the geographic scope (boundaries) where the delegate will act be specifically mentioned in the legal framework; and (b) the prior approval by the Ministry of Finance and Public Credit of the budgetary transfers related thereto be granted. In this sense, the Constitution posed a serious challenge for policy makers to actually reach these goals. The main components of the organizational reform supported under the World Bank-financed Modersa Project, can be described as political and administrative decentralization, economic decentralization, and hospital autonomy. Political and Administrative Decentralization. A major step in the decentralization process has been the integration of public and private providers into the MIunicipal Integrated Health Care Networks (MHCN), which involve substantial community participation. These networks have been managed by a Junta de Salud (-Iealth Board) that coordinates services and information among health providers in order to guarantee essential interventions. The responsibilities of the Health Boards are: (i) development of a local health plan (including organization and financial arrangements); (ii) categorization and registration of health care users; (iii) development of new health care financing mechanisms (such as a local health fund); and, (iv) supporting the modernization of management structures and practices (information, quality assurance miechanisms, human resources). The following instruments have been essential for the public and private integration. First, the Health Boards have determined the scope of the basic health package and specified its components according to the epidemiological characteristics of the region. The price of such package is then estimated following Cost Manuals generated in the Modersa project. The Health Board decides the volume of the required service, i.e. number of people covered by such package, and incorporates private providers for the delivery of basic package interventions. This stage uses standard contracts and procedures, which are also provided by the project, which clearly specify type of payment, incentive schemes, and quality control mechanisms. An example of the decentralization can be observed with the creation of the Centralized Health System in the Canton Tena, where substantial community participation was observed (See Box 1). 9 Box 1. Centralized Health System in the Canton Tena On November 7th, 1997 the Junta de Salud was established, having the Major as the President and the Director of the Health Area of Napo as the alternate President. The structure of the Junta includes a Health Assembly with political control, the Board with management power, and the Technical Team with responsibilities on the execution. The decentralization required legal instruments as the Ministry-level agreement with the Municipality of Tena, and an Ordenanza to delegate power to the Junta de Salud. A first Plan was constructed based on the citizen's participation through Canton Level Assemblies, Rural Workshops, Technical Workshops, Topic-specific agreement boards, consulting and opinion polls. This Junta has four components: Healthy Lifestyles and Environments (involving promotion, prevention and community participation), Services Network (organization and enrollment of users, entitlement of services, hospital modernization, and consolidation of the fluvial-system), Support System (management, information and human resources) and the creation of the Fondo Local de Salud. The functioning of the Fondo de Salud included the registration and categorization of the population, the definition of the basic package and its cost, establishing a subsidy and co-payment system, the operational management of the Fondo, and contracting registered providers. An interesting innovation was observed in the subsidy and co-payment system, since there are six potential discount cases based on the socioeconomic evaluation: 0, 25, 50, 80, 90 and 100%. For the Ecuadorian system as a whole only four cases are possible: 0, 25 50 or 100% discount on the tariff. This is evidence that with enough entitled flexibility, local Juntas might be able to establish more precise categorization procedures with appropriate discount schemes. These guarantee both cost-recovery and efficiency gains (in the sense of being able to discriminate on prices). The management of the Fondo involves a Manager, a Medical Supervisor, a financial-administrative head and a secretary. The seed-fund is composed of Modersa 60%, local government 10%, co-payment 20% and central government and other institutions 10% reaching about US$ 2.2 Million. Source: Ministry of Public Health (1998) Economic Decentralization and Hospital Autonomy. Recent decentralization and state modernization laws have opened opportunities for reforming public hospitals. Although a large number of public hospitals require substantial improvements, the MOH, with support from the MODERSA Project, has also set up hospital autonomy demonstration models in various cities: in Quito, Enrique Garces Hospital and P.A. Suarez; in Cuenca, Vicente Corral Moscoso Hospital; in Guayaquil, El Nifno and Guayaquil Hospitals; and in Tena, Jose Maria Velasco Ibarra Hospital, following specified criteria, for eventual replication throughout the country. In the initial phase, the 10 goal is to strengthen hospital management in order to improve technical efficiency and quality of care. Substantial cost differences were found among providers, even within the MOH hospitals (see Box 2). The differences in the cost of ambulatory visits and hospitalization reflect differences in personnel qualification and economies of scale in large provincial Hospitals. Box 2. Health Care Costs in Public Hospitals, Ecuador 1999 (in thousand Sucres) Type of Specialized Hospital Hospitalization Cost Ambulatory Visit (Location) per patient Provincial Hospital (Pichincha) 2,667 79 Provincial Hospital (Pichincha) 1,860 30 Specialized Hospital (Guayas) 2,861 79 Source: Lucio and Lasprilla (1999) Deconcentration Instruments. The deconcentration of management and aidministrative functions from the MOH to the directors of its hospitals leading to a gradual autonomy of such hospitals is done through Ministerial resolutions issued as the respective hospital qualifies to participate under the MODERSA Project. The most imnportant powers transferred are: to carry out the contracting of goods and services as auLthorized by the domestic procurement law; to approve the expenditures made under specific contracts for payments to be made out of the approved budget for that purpose; to appoint and discharge its employees (doctors, nurses and others); and, to plan and monitor implementation of investments (for improvements) of the respective hospital. Scope of Hospital Modernization and Criteria for Selection. This decentralization included the modemization of public hospitals, which included the modernization of their organization and govemance structure, management practices, management information systems, and improving health care quality in the medium-term. H[ospital modernization also was to include: (i) implementing autonomous managerial arrangements including user participation; (ii) strengthening planning and decision making; (iii) developing human resources through selective training and continuing education, performance incentives, and career development; (iv) implementing financial administration systems and management information systems; (v) revamping cost recovery policies and systems and developing new resource allocation mechanisms for hospital care; and (vi) financing civil works, equipment, maintenance, and supplies for refurbishing and upgrading existing facilities. Modernizing institutions with strongly entrenched traditions and weaknesses requires experimentation and careful selection of candidates for change. The MOH selected the initial demonstration hospitals using the Hospital Management Assessment 11 Instrument. This evaluation assessed seven key areas: (i) senior management capabilities; (ii) strategic planning; (iii) information systems; (iv) accounting and financial management; (v) human resources and labor relations; (vi) facilities and equipment; and (vii) relationships with the community and other providers. Other factors for selecting Hospitals for the project were: (i) have a MOH resolution with deconcentration of authority provision issued (i.e., transfer of powers from the MOH to hospital directors); (ii) hospitals classified as provincial or national hospitals; (iii) hospital director and staff willing to accept changes in organization, financial management, and quality improvements; (iv) hospital has strong influence in its region and credibility to serve as a model; (v) hospital director signed an agreement with MOH, when required, with endorsement of respective provincial/municipal authorities, to undertake modernization changes; and, (vi) willingness to be part of a "learning network" organized with other hospitals included in similar change processes. Improvements in Hospital Accounting and Reporting Systems and the Extensive Use of Cost Recovery Mechanisms. Cost recovery in public hospitals is an important policy option that benefits both the health system itself and the public that uses it. With demonstrated improvements in quality, Ecuador's MOH could eventually be able to achieve a significant percentage cost-recovery of recurrent costs. To this end, in the demonstration hospitals, support is provided to develop a revised system to charge and collect fees, including legal safeguards that would allow hospitals to manage autonomously the revenues generated by user charges. The MOH has developed a Fee Schedule that includes both Total Costs and Reference Fees (Tarifas Referenciales) estimates for each intervention. The difference is that the Reference Fees include only operational costs excluding medical personnel expenses. As the blind use of such fees might have a regressive effect, the MOH allows hospitals to proportionately adjust the Fee Schedule according to the regional circumstances (Box 3). Moreover, a Socioeconomic Categorization of Users procedure was designed to apply discounts to the user fees according to family characteristics (Box 4). Box 3. Fee Schedule for Selected Procedures (thousands of Sucres) Procedure Cost Total Reference Fee Cholecystitis (surgical 270 110 Appenditicits (surgical) 700 280 Gastrintestinal endoscopy 260 104 Nornal Birth 650 260 Elective Caesarean Birth 1,000 400 Tonsil extraction 444 200 Prostate biopsy 485 194 Source: Annex 1. "Tarifario Referencial". 12 Box 4. Socioeconomic Categorization for Health Care Users in Ecuador In May of 1999 the Ecuadorian government adopted criteria to classify users of public health facilities. This categorization required information about the residence (urban/rural, transportation, owned dwelling, access to water, sewage, electricity and phone), education (household head, number of children attending school or university, type of school), household demographics (female headed, number of dependents, disability, chronic diseases), and, occupation and income (job location, household income, durable goods, financial assets). Questions on each item assigned "points" which may accumulate to a total ranging from 0 to 210 "points." Public subsidies were applied as follows Category Point Paid by User (%) A Less than 50 Free service B 50-100 25% of the Reference Fee C 101-125 50/o ofthe Reference Fee D 126-180 1 00% of the Reference Fee E More than 180 100% of the Total Cost Individuals holding a Bono Solidar:o are entitled to Category A. Individuals enrolled in a public or private health insurance correspond to Category E. Source: 'Annex 2. Criterios para Categorizaci6n Socioecon6mica.' Acuerdo 1292 - Ministry of Public Health. Promoting Allocative Efficiency in Hospitals. The current budgeting system in the MOH hospitals does not promote allocative or technical efficiency (Lucio and Lasprilla, 1999). Budgets are based on historical spending patterns and are not based upon the output of the hospital. Changing the method of paying hospitals is one possible way to improve both the technical and allocative efficiency of hospitals. To this end, the MODERSA Project is supporting: (i) development of a payment methodology; (ii) implementation of "shadow prices" to acquaint hospitals with reimbursement based upon output; and (iii) gradual implementation of new hospital financing methodology. For example, initially the payment rate would be set equal to total hospital budget divided by total number of discharges. All participating hospitals would receive the same payment rate. The uniform payment rate would be used until it can be refined to include additional factors such as case mix, cost of living, and perhaps hospital-specific factors. The purpose of this initial payment rate would be to get hospitals accustomed to working under a per case administration payment system. It would not be used to actually pay hospitals, but instead, it would be used to simulate what a hospital would receive under a per discharge basis. It would also allow time to refine the payment method. 13 Improving Hospital Governance. There is agreement among key stakeholders that important hospital decisions should be made closer to the population served to improve flexibility and responsiveness to the specific needs of the diverse geographical areas and population groups in the country. In order to improve decision making in hospitals, the following measures have been implemented: (i) development of a decision-making framework to support greater autonomy of selected public hospitals by redefining the relationship between MOH and the Provincial and Municipal Health Boards or Councils, autonomous health entities and individual hospitals, with respect to key hospital governance decisions; (ii) implementation of one or more governance models that would allow regional authorities and local communities to share greater responsibility for the governance and management of demonstration hospitals; (iii) design and implementation of improved management processes supported by management information systems, including Total Quality Management (TQM) initiatives, that would support the strengthening of cost accounting, performance improvement and measurement, and reporting systems in the demonstration hospitals; (iv) design and establishment of hospital management educational programs that train individuals for positions as senior hospitals and health system leaders, managers of clinical and administrative departments, and technical support staff; and, (v) strengthening of the educational program infrastructure in hospital and health system management. Human Resource Management. Overall, the MOH concentrates 23.2% of its labor force in health services, the LESS 11.7%, and other private for-profit institutions 35 %. Out of 16,000 physicians in Ecuador, the MOH and the IESS account for about 6,600 physicians in the system compared to 6,000 in private health provider businesses (OPS, 1999). The main problems observed in the human resource management are the lack of labor incentives to provide a health service of improved quality, the inability of Hospital directors to choose an appropriate labor mix between administrative and medical personnel, and the lack of a career profile in the public sector (Lucio and Lasprilla, 1999). These problems are determining factors in the high tumover of qualified personnel, the percentage share of public and private jobs, and other labor arrangement problems to be addressed. With support of the MODERSA project, the decentralization is encouraging output- based incentives--which still require legal instruments--that delegate power to hospitals for hiring, for management of layoffs, for decision-making on training, and for information systems to organize these decisions. As the project involves other public and private institutions, contracts with providers specify appropriate incentives and accountabilities. The process is intended to move the incentive system from one based on fixed salaries and budgets to a scheme were risk is shared between providers and users and salaries are subject to those risks (partial- and full-capitation). The experience of Colinas del Norte suggest potential avenues for reform in the human resources management that need to be strengthen (see Box 5). 14 Box 5. A Public HMO: The Community System for Integral Health in Colinas del Norte The Community System for Integral Health (Sistema Comunitario de Salud Integral, SICSI) in Colinas del Norte builds off of the Municipal Health Council model and seeks to improve the incentive environment wherein health workers and service recipients interact. It targets one neighborhood with about 2000 families north of Quito. The SICSI operates like a small public health maintenance organization (HNO) in which families pay small monthly fees to become members of a clinic. Physicians are paid their usual monthly base salary plus bonuses for a) each new member family and b) each individual attended and service rendered. A physician in a project clinic continues to receive her basic monthly salary independent of her productivity. For each new paying affiliated member family, she receives a monthly bonus of 0.1%. This encourages the physicians to help sell the program in the communities they serve, and quite likely, to treat members better than non-members. In addition, physicians are paid for each service rendered, within certain limits. They receive s/1,250 per consultation up to a limit of s/500,000 (or 400 consultations) per month. Finally, no more than 40% of the additional pay for consultations may relate to morbidity, leaving 60%o for "public health" consultations. Affiliated families pay about US$2.50 per month, for which they receive a comprehensive array of health services including preventive care, pediatric care, and pre- and peri- natal care. The local clinic is the main point of contact but has referral rights to the local hospital. By contrast, non-affiliated families continue to be served in the traditional manner; that is, a system with co-payments for most services. In the first four months of the project, 700 families have affiliated with the clinic, and about 400 of these families pay the monthly fee regularly. Families eligible for the Bono can affiliate without paying the fee (in the same manner that if they are not affiliated they do not have to make most co-payments). In addition to the encouragement of HMO-style efficiency incentives, another promising development is the cooperation on related service provision (such as pre- and peri-natal care) that the project has helped forge between the local clinics in the Area de Salud and the hospital geographically located within it. Source: Gershberg (1999) Measures to address the problem of inadequate health insurance coverage. Universal coverage is a basic goal of all health systems. In Ecuador, there is widespread support for an increase in formal coverage through mandatory health insurance, under public and private schemes. This arrangement is ideal for a number of reasons. First, increasing health insurance coverage would reduce the burden of health spending on many uninsured families who are poor or near-poor. Second, because the uninsured tend to delay treatment until they are quite ill and have no choice but to seek medical help, this results in higher family outlays than if the illness had been prevented or treated at an earlier stage; universal coverage would attempt to alleviate this problem. Third, under current arrangements, there is no protection against catastrophic illness costs, which is the basic purpose of health insurance. Fourth, health insurance enables solidarity and cross- 15 subsidization among beneficiaries. Fifth, the number of people relying on the MOH and other public programs, as the "insurers of last resort" will decrease, allowing for a more equitable targeting of public spending on health. Building upon the current legal framework, three scenarios have been discussed before: * Scenario 1 achieves the greatest expansion of health insurance by extending IESS coverage to the 1.7 million dependents (400,000 adults and 1.3 million children) of the 1.1 million current LESS affiliates, as well as 1.2 million self-employed individuals and 2.4 million of their family members (2.1 million children and 300,000 adults). Within this scenario, health insurance under IESS would cover 57 percent of the Ecuadorian population, up from the 10 percent it currently covers, and together with other existing sources of coverage (e.g., private health insurance, armed forces, Peasant Social Insurance), 67 percent of the population would be covered. * The magnitude of this IESS expansion, even with 10 percent of dependents costs being absorbed by the dependents themselves (either through a monthly or annual premium payment or through user fees), and with the self-employed financing 67 percent of the cost of their coverage, would still require the IESS to increase the compensation base that is assessed from the current 60 percent to 91 percent to raise approximately US$172 million. The resources from the MOH and other public programs would be used to fund services for the uninsured population (but the amount of spending per person would be more than 3.4 times higher in real termns that current spending levels, from US$17 to US$57). To achieve public spending per uninsured person equal to three-fourths the per capita spending level of IESS, MOH's spending per uninsured population would have to rise from US$10 at present to US$48 (assuming that resources from other public programs remain constant in real terms). As a result, the MOH would need to increase its spending by US$61 million in real terms to reach this level of spending per uninsured person. This would require a 38 percent increase in government funding and 10 percent cost recovery. MOH expanded resources could be used to offer a more comprehensive program of essential services than the one discussed above. * Scenario 2 offers a less expansive coverage extension by including dependents of current LESS affiliates and the self-employed (but not their dependents). This would result is 45 percent of the population receiving coverage, with 55 percent remaining uninsured (about 6 million persons). For IESS, this more limited expansion requires more limited funding--the proportion of the compensation base that would need to be assessed to fully fund the portion not paid for by the dependents and the self- employed is 82 percent. Given the sizable population for which the MOH and other public programs retain responsibility, this scenario sets a more modest target for spending per uninsured person of one-third the spending level of IESS (from US$17 to US$3 1). The MOH could achieve this goal with a 26 percent increase in government funding (about US$30 million), along with 5 percent cost recovery. * Scenario 3 offers the most limited coverage expansion, extending IESS coverage only to LESS dependents, and thus achieving 35 percent overall coverage (65 percent of the 16 population would be uninsured). This would require approximately US$85 million from IESS if dependents financed 10 percent of their costs, and could be achieved by increasing the compensation base that is assessed to 76 percent. While the per uninsured person spending target remains the same as that in Scenario 2 (one-third the level of LESS per capita spending), public spending per uninsured person rises to US$27 from US$17 at present. The required MOH fiunding increase from the government to reach this level is higher because of the larger uninsured population. Even with 5 percent cost recovery in MOH facilities, the government funding for MOH would need to be increased by 34 percent over 1994 levels in real terms (about US$39 million). Table 4. Options For Health Insurance Reform CURRENT SCENARIO 1 SCENARIO 2 SCENARIO 3 Comprehensive Moderate Incremental Population to be 10% 57% 35% 25% covered by IESS Current Current affiliates + Current Current Affiliates Dependents; self- affiliates + affiliates + employed dependents; dependents + dependents self-employed Salary base for IESS contributions 60% 91% 84% 76% calculation Approx. IESS per $117 $76 $92 $81 capitaspending(1994 US$) Uninsured population under responsibility of 80% 33% 55% 65% MOH and other public programs Public dollars spent per US$17 US$57 US$31 US$27 uninsured person (94 US$) MOH costs - 10% 5% 5% recovery/community financing needed Approx. additional - US$44 MM US$30 MM US$39 MM government funding required for MOH (1994 US$) Source: Mhrquez et al (1999). 17 The Urgency of Reforming the EESS. Given the major institutional weaknesses found in the social security system, the expansion of health insurance coverage as envisioned above would have to go hand in hand with a major restructuring of IESS, including the separation of its pension and medical schemes, and the identification of alternative mechanisms to manage those schemes (e.g., the administrative and financial aspects of health insurance coverage would be handled by various institutions, who would be separated from those that would actually deliver health care services). Moreover, private health insurance schemes would have the complementary role of providing coverage to those persons who desire and can afford to have access to a broader network of providers or additional services than those offered under IESS. Strengthening Alliances for Health Improvements. Another area of less attention is the health service to those poorer households that are mainly serviced by either self-medication or by local pharmacists. While the reform may reach those already serviced by public and private health providers, there is a need for designing innovative mechanisms to exploit the ubiquitous pharmacy network to reach the poorest households. Within the framework of the health reform, this may mainly consist of having the local pharmacies represented in the Health Boards, as well as establishing a network of trained health consultants to improve the common self-diagnostic and self-medication practices. Another alternative is to exploit the work that the NGO have been doing on health. There are an important number of institutions that might be interested and probably able to act as intermediaries in health service provision. Box 6. Social Organizations and Health Social Organizations in Ecuador include Local NGOs, Popular Organizations, International NGOs, Church Organizations and Private Business organizations. While there is a large variation among these institutions in both the fraction of projects, is clear that the proportion of organization is much larger than the fraction of projects. This evidences either the relatively large cost of projects or the fact that several health-oriented NGOs are executing projects of different nature. For example, there are 317 local NGOs related to health, but only 191 projects being executed. Organizations and Projects related to Health i Organizations Projects NGOs 48.1 21.8 Popular Org. 44.5 7.3 International NGOs 50.0 13.6 Source: SIOS - ALTERNATIVA. Organizaciones Sociales de Desarrollo (1997) 18 References Organizacion Panamericana de la Salud (1999) Ecuador. Perfil del Sistema de Servicios de Salud Programa de Organizacion y Gestion de Sistemas y Servicios de Salud. C:EPAR (1998) Cuentas nacionales de Gastos en Salud Ecuador - 1995. Informe Final. CEPAR-INEC-GESE-PATCO. June, Quito. Gershberg, A. (1999) "Education and Health Decentralization in Ecuador: Practice, Plans and Prospects," October. Gragnolati, M. (1999) Incidence Analysis of Social Expenditures and Public Subsidies in Ecuador. Background paper for the Ecuador Social Expenditure Review. August. INEC (1995) Encuesta de Manufactura y Mineria 1995. Lucio, R. and E. Laprilla (1999) Costosy eficiencia de los Servicios de Salud en el Ec-uador a 1999. Estudio preparado para el Banco InterAmericano de Reconstruccion y Fomento. Quito, July, 1999. Lasprilla, E. et al. (1997) Gastoy Financiamiento del Sector Salud- Ecuador 1996. Organizacion Panamericana de la Salud. Salud Publica: reflexiones y experiencias No. 4. Quito. October. Lasprilla, E. et al. (1998) Equity in Health in LAC. Country studies: Ecuador. November. Unprocessed mimeo. Maceira, D. (1996) Fragmentacion e Incentivos en los Sistema de Atencion de la Salud en America Latina y el Caribe. Banco Interamericano de Desarrollo. Oficina del Economista Jefe. December. Marquez, P. F. Sacoto and M. Molares-Halberg (1999) Organizational Innovations: The Lipchin of Health Sector Reform in Ecuador. Washington, D.C. MSP (1998) Tena: En la Ruta de la Saludy el Bienestar. Propuesta de Creacion del Sistema Descentralizado de Salud del Canton Tena. Mimeo. MSP (1999) Ministerio de Salud Publica. Reglamento al DE 502. Anexo 1. Tarifario Referencial. Anexo 2. Criterios para Categorizacion SocioEconomica. Musgrove, P. (1987) "The economic crisis and its impact on health and health care in Latin America and the Caribbean" in International Journal of Health Services, vol 17, pp.411-41. Sr[SE (1999) Sistema Integrado de Indicadores Sociales del Ecuador. Version 1.0. CD- ROM. 19 Annex 6: Incidence Analysis of Social Expenditures and Public Subsidies in Ecuador Michele Gragnolati August 1999 I. Introduction The objective of this section is to examine the extent to which the poor benefit from public spending. We focus our attention on three major activities which have redistributive impact in the Ecuadorian society. First, we examine government's expenditures in education and health. Second, we look at targeted safety nets programs. Third, we examine the price subsidies on cooking gas and electricity. In the Appendices, we discuss the methodology used and show concentration curves and additional tables. We base all calculations on the third round of the Encuesta Condiciones de Vida (ECV), which was conducted in 19981 and on information gathered by a World Bank mission to Ecuador in July 1998. We examine the incidence of government activities by per capita expenditure quintiles. As a reference, Table 1 contains some consumption characteristics of the Ecuadorian population important for the benefit-incidence analysis. Poorer households tend to live in rural areas and to have bigger families. The distribution of consumption is highly skewed towards the richer households: the population in the richest quintile consumes about 20 times more than the population in the poorest quintile and 2 appropriates 60% of overall resources. Table 1. Distribution and Characteristics of Households by Expenditure Quintile Consumption Expenditure Quintile _________________________ _______ _ _Qi Q2 Q3 Q4 Qs Overal Population Share 20.0 20.0 20.0 20.0 20.0 100 Percentage in Urban Areas 22.5 42.6 62.1 73.8 88.2 57.8 Household Share 15.0 17.2 19.0 22.4 25.9 100 Household Size 5.9 5.2 4.8 4.2 3.5 4.5 Per Capita Expenditure Sbare 3.2 j 6.9 11.1 18.8 60.0 100 Source: ECV 1998 A fourth round of ESV in currently being fielded. 2 See Appendix 2 for the Lorenz Curve of per capita expenditure. Because of the extremely unequal income distribution, it is not difficult for a subsidy to be progressive in the sense that its removal would worsen the overall income distribution in society. Rather, we evaluate the progressivity of public subsidies and expenditures by looking at which part of society obtains the largest share as our benchmark is that the poor should be the major beneficiaries.3 3 See Appendix 1 for further discussion on the methodology and the definition of progressivity. 2 II. Social Expenditures The provision of basic social services is an activity through which the government can have a positive effect on the welfare of its people. Cost recovery of social services in Ecuador is very low since public school tuition and health fees are either minimal or absent. Social expenditures are hence financed by general taxation. 1. Education Primary and secondary school in Ecuador last for six years each; university is four to six years depending on the field. In addition, a one year pre-primary school (kindergarten) program is also available. Public primary schools are universally available. Secondary schools are available in cities and towns, but often not in the more remote rural areas. IJniversities are concentrated in the major cities. Table 2 presents the unit subsidy that each child enrolled in public school receives fi-om the government every year, by level of education.4 Because of the much fewer number of enrollments in universities, the unit subsidy that a university student receives is about five times higher than the subsidy received by a child at the beginning of the education career. Table 2. Unit Subsidy by Level (US$) Unit Subsidy Pre-primary Primary Secondary University Overall 1998 117.9 117.9 274.1 552.3 159.4 Source: SINEC The public education system is defined to include fiscaL fiscomisional and municipal schools. Figure 1 reveals that overall public expenditure on education in Ecuador is regressive: the poorest 20% of the population receives only 11% of the government total spending on public education, while the richest 20% receive 26% of government spending. This finding is driven by the combination of higher enrollment rates in secondary schools and universities among the richer Ecuadorians and high unit subsidies per student in each of these levels. Public spending on primary schooling is progressive due to the fact that school attendance is close to universal but many more richer families send their children to private schools.5 The more unequitable distribution is associated with university expenditures of which the poorest 40% of the population receive only 6%. 4The statistics of the Ministry of Education treat expenditures in pre-primary and primary schooling as one category. 5 See Appendix 2 for education net enrollment rates and concentration curves. 3 Figure 1. Share of Government Spending on Education 0.50 0.40 030 0.20 0.10 Pivprnary Pnlary Secondary Univenity AM 13 st 1312m *3rd *4th E5th Source: SINEC, ECV Three findings emerge from the analysis of ECV data, which need to be taken into account to evaluate the role of public education as a socially equalizing force in Ecuador. * The education process lasts longer for richer children . For example, the expected number of years spent in secondary school is six years and two years for children in the richest quintile and children in the poorest quintile respectively. The corresponding figures for university education are 3.7 and 0.1 years.6 * At all education levels, enrollment rates in private schools are higher among children of richer households. If use of private schools translates into higher quality of education and higher returns to education later in life, income inequality between the rich and poor is likely to increase. * Rich households spend more on the education of their children than poor households. Households incur out-of-pocket expenditures to gain access to and improve the benefits of subsidized education. Private expenditures are indeed inequitable in Ecuador. The lowest quintile receives only about 8% of total household investments in primary education, and the upper quintile as much as 55% 7 The end result is that rich children in Ecuador receive a longer and better education, as compared to poor children. 6 See Table 6 in Appendix 2 for the expected number of years of enrollment in each education level. 7 See Appendix 2 for the concentration curves of private spending on education. 4 2. Health Although it has made notable progress in health in recent years, Ecuador still lags behind other Latin American countries. Between 1960 and 1995 life expectancy rose from 55 to 68 years, but in 1995 the infant mortality rate was 36 per thousand as compared to 26 per thousand in Colombia and 12 per thousand in Chile. The basic Ecuadorian health care system consists of the public system and private providers. In the public sector, the Ministry of Health (MSP) and the Social Security Institute (IESS) are the main actors. Health care at all public institutions is nominally free of charge, though users often have to pay for supplies, laboratory work, etc., especially in MSP facilities. Use of health services is low: according to the 1998 ECV only 10% of the population visited any health facility in the month prior to the interview. Because of the historically low and declining quality of the public sector, many people who visit health facilities rely on the private sector (45% of the visits are to private clinics). Even among the poorest groups, a considerable proportion attends private clinics (27%).s Table 3. Unit Subsidy by Type of Public Health Care Facility (sucres/) F I Hospital Unit Subsidy l 1 Hospital | Centro de Salud | Subcentro de Salud 1999 268,348 11,814 9,511 Source: MSP The unit cost of a visit to a public hospital is more than 20 times higher than the unit cost of a visit to a smaller public health facility (centro de salud or subcentro de salua). To calculate the incidence of health expenditures, we use information from the ECV about the last visit to a public health facility. Within public health care facilities, the rich tend to use hospitals, while the poor are more likely to visit health sub-centers and, to a smaller extent, health centers. We find that public expenditure on public health care facilities is regressive with only 8% of total spending arriving to the poorest 20% of the population. 8 See Table 7 in Appendix 2. 5 Figure 2. Share of Government Spending on Health 0.40 035 0.30 ols 0.10 0.05 0.00 1 Hoispital Centro de salud Subhntrm o All dispensalio 1Elu 2st1 E3nd E4th 05th Source: MSP, ECV The ECV also contains information on participation to different kinds of health insurance. For each expenditure quintile Table 4 shows the participation rates to four different health insurance schemes: (1) Private insurance; (2) IESS Seguro general, (3) IESS Seguro campesino, and (4) Seguros ISSFA (Social Security Institute of the Armed Forces) or ISSPOL (Social Security Institute of the Police). More than 80% of the people in lowest quintile do not have health insurance. Table 4. Participation Rates to Different Health Insurance Programs ________ Consumption Expenditure Quintile Qi Q2 | _Q3_ | Q4 | Qs Overall Private Insurance 0.1 0.4 1.2 2.8 12.9 3.5 IESS Seguro general 1.2 3.4 7.5 13.3 26.5 10.4 IESS Seguro campesino 16.3 12.8 10.1 4.5 1.9 9.1 Seguro ISSFA or ISSPOL 0.3 0.1 0.5 1.6 2.9 1.1 None 82.1 83.4 81.2 78.7 60.9 77.2 Source: ECV 6 III. Safety Nets Programs As part of its poverty alleviation strategy, the Government of Ecuador has developed a number of social programs that seek to expand access to or improve the quality of basic services. Most of the more recent social programs have explicitly stated objectives of reaching the poor and targeting elements were built into their design. 3. Bono Solidario The bono solidario was introduced in September 1998 to mitigate the negative iimpact on the poor originating from the elimination of public subsidies on gas, fuel and electricity. The bono consists of a cash transfer to poor mothers with children and to pieople aged 65 and over. In April 1999 the transfers were increased to 150 and 75 sucres for mothers and adults respectively, and were extended to the disabled population. The bono solidario is the main social cash transfer program in place in Ecuador. Figure 3. Share of Government Spending on Bono Solidario 50 45- 35- 30- 25- 20I 15-_I _ 105 5- o- 1 2 3 4 5 U 3QuintilcsI Source: ECV Overall, the bono solidario appears well targeted, at least with respect to other social programs. The poorest 40°Oo of the population receives 54% of the total disbursements (to both mothers and adults). The component to adults is especially pirogressive: the poorest quintile receives 38.5% of the disbursement (results not shown). However, there is considerable leakage to people in the richer quintiles. In particular, the bono seems to be reaching urban areas disproportionately and its coverage in rural areas is much inferior than it should be to reach all the most vulnerable groups. 7 4. Colaci6n Escolar The School Breakfast Program has been operated by the Ministry of Education since January 1990 with support of the World Food Program. The objectives of the program are to provide nutrition supplements to children in primary schools considered at risk for their socioeconomic conditions as a stimulus to school participation and academic achievement. The Ministry of Education has recently launched a program of almuerzo escolar with similar objectives. Figure 4. Share of Government Spending on Colacion Escolar 0.50 0.40 0.30 0.10 0.00 ~~~Olst 02nd E3rI E4th 05th Source: ECV The colaci6n escolar is well targeted and leakage to the richer quintiles is very small. 5. Alimentos Gratuitos The Government of Health (MSP), the Ministry of Social Wellbeing (MBS) and the National Institute of Children and the Family (INNFA) offer free food to children below school age. 8 Figure 5. Share of Government Spending on Alimentos Gratuitos by Source and Income Quintile 0.70 0.60- 0.50 0.40 0.30 0.20- 0.10 0.0 __ INNFA MSP MBS 1131t f2r, E3rd E4t E5th Source: SINEC, ECV The progressivity of programs of alimentos gratuitos varies across institutions. Programs from the Ministry of Health (MSP) are the best targeted: the poorest quintile benefits of 61% of overall spending, while the richest quintile receives only 1%. 9 IV. Price Subsidies on Cooking Gas and Electricity Traditionally, the government has given price subsidies to the consumers of cooking gas and electricity by fixing the consumer price at a lower fair than the production cost. The cooking gas subsidy has been provided by setting the price of canisters at roughly one-fifth the combined production and distribution cost. The electricity subsidy has been provided as discounts in the unit household consumption charge, with higher percentage discounts applying to households with lower monthly usage. These subsidies were not only costly but also inappropriately targeted. The benefits from the price subsidies have been higher in urban areas, although the largest poverty incidence is in rural areas, and have been enjoyed by all consumers, including the richest households. Moreover, neither subsidy reached Ecuador's poorest, i.e. those who did not consume any cooking gas or electricity. In 1998, the Government increased considerably the price of cooking gas and electricity in order to eliminate the subsidies.9 However, the new price was fixed in nominal sucres and the subsidy reappeared after the price of oil, the exchange rate and the inflation rate increased. Price subsidies for cooking gas and electricity are highly regressive. Apart from the share of subsidy distributed for industrial use (7.0% and 25.0 % of the overall subsidy for cooking gas and electricity respectively), subsidies for residential use benefit disproportionately richer households, which consume both gas and electricity in larger quantities. Only 27% of the cooking gas subsidy and 24% of the electricity subsidy were estimated to flow to the poorest four deciles of the population. Figure 6. Share of Subsidies on Cooking Gas and Electricity (Residential Use) 0.40 0.30 0.20 0.10 0.00 Cookin Gas Mmidty 101kst ln 3rd 4th Ef5th Source: ECV 1998 9 These subsidies are currently estimated to be 1.1% (electricity) and 0.7% (cooking gas) of GDP (IMF 1999). 10 V. Summary Table Table 5 summarizes the results of the incidence analysis by showing the share of social expenditures and public subsidies received by each expenditure quintile. Table 5. Share Received by Each Expenditure Quintile _______________________________ .Consumption Expenditure Quintile [ECV981 = ____________________ __ _Qi Q2 Q3 Q4 QS SOCLAL SECTORS (11) Education Pre-primary 0.15 0.22 0.25 0.22 0.15 Primary 0.24 0.24 0.21 0.20 0.10 econdary 0.08 0.16 0.25 0.29 0.22 University 0.01 0.05 0.12 0.30 0.52 All 0.11 0.16 0.21 0.27 0.26 (2) Health Hospitals 0.07 0.13 0.18 0.24 0.38 Centros de salud 0.11 0.18 0.23 0.27 0.21 Subcentros o dispensarios 0.18 0.23 0.22 0.21 0.16 All 0.08 0.13 0.18 0.24 0.37 SAFETY NETS _ (3) Bono Solidario* _ _ _ Mothers 0.25 0.28 0.22 0.15 0.11 Eldery 0.39 0.19 0.19 0.13 0.10 All 0.28 0.27 0.21 0.14 0.11 (4) Colacion Escolar 0.41 0.25 0.16 0.12 0.06 (5) Alimentos Gratuitos _ _ INNFA 0.22 0.43 0.13 0.11 0.11 MSP 0.61 0.12 0.24 0.02 0.01 MBS 0.18 0.23 0.22 0.21 0.16 PRICE SUBSIDIES (Residential Use) _____ (6) Cookinggas 0.12 0.15 0.19 0.23 0.31 (7) Electricity 0.10 0.14 0.17 0.21 0.37 Source: ECV 1998 Data are for ECV99 second trimester income Quintiles 11 VI. References Demery L. (1997). Benefit Incidence Analysis. Poverty Reduction and Economic Management Network, The World Bank (unprocessed mimeo). Parandekar S.D. (1998). Analysis of Private Expenditures on Education in Peru. (unprocessed mimeo). International Monetary Fund (1999). Ecuador: Protecci6n Frente a la Crisis Econ6mica". (unprocessed mimeo). Lucio R. and E. Lasprilla (1999). Costos y Eficiencia de los Servicios de Salud en el Ecuador a 1999. (unprocessed mimeo). World Bank (1996). Ecuador Poverty Report. Washington D.C.. World Bank (1997). Is Society Getting What it Needsfrom Public Spending on Education? Issues in Efficiency and Equity. Washington D.C.. World Bank (1999). Poverty and Social Developments in Peru, 1994-199 7. Washington D.C.. 12 Appendix 1 Methodology Estimating the benefit incidence of public spending involves three steps (Demery 1997): Step 1: Estimating unit subsidies The unit cost of a service is defined as total government spending on a particular service divided by the number of users of that service. Information comes from government statistics on different social services. Whenever the value of the unit subsidy is not available, we estimate the share of each type of service produced (distribution of volume produced) received by each income quintile instead of the share of actual expenditures (distribution of value produced). The two coincide if the unit subsidy is fixed across the income distribution. Step 2: Identifying users Information on who uses the service is obtained from the 1998 and 1999 ECV. Step 3: Aggregating users into groups. To describe how the benefits from public spending are distributed across the population, we aggregate individuals into expenditure quintiles (for the services provided to individuals) and we aggregate households into expenditure quintiles (for the services provided to households). The ECV, which is modeled after the Living Standards Measurement Surveys, provides the information needed to calculate household per capita expenditure. Concentration curves Concentration curves show the cumulative proportion of social expenditure or subsidy that is received by the cumulative proportion of the population ranked by per capita expenditure: The 450 diagonal is the line of "perfect equality", that is the line that would be obtained if each quintile received the same amount of social expenditure or public subsidy. By comparing the concentration curves with the 450 diagonal, we can judge the targeting to poorer groups. If the curve lies above the diagonal, it means that the poorest quintile gains more than 20% of the total subsidy (and the richest quintile, less than 20%). Distributions below the diagonal signify poor targeting. We evaluate the 'progressivity' or 'regressivity' of social expenditures and subsidies by estimating the share that is received by the poorest rather than by comparing the distribution of expenditure and subsidies with the distribution of income. 13 Appendix 2 Concentration Curves and Additional Results Per Capita Expenditure Lorenz Curve of Per Capita Expenditure .4 . o.6 0 . .2 .4 .6 .8 Cumulative Proportion of Population 14 Education Table 6. Net Enrollment Rates and Expected Number of Years Enrolled by Level of Education (Pre-primary, Primary, Secondary and University) Consumption Expenditure Quintile 1st 2nd 3rd 4th 5th PUB PRI AU PUB PRIV All PUB PRIV An PUB PRIV AU PUB PRIV AU PRE-PRIMARY 5 16.3 2.3 18.6 33.3 6.5 39.9 39.0 11.0 50.0 37.0 21.7 58.7 26.3 46.3 72.6 Expected Years in 0.2 0.0 0.2 0.4 0.1 0.5 0. 5 0.2 0.7 0.5 0.3 0.8 0.3 0.6 0.9 Prepunary Scbool PRDMARY 5 9.8 0.6 10.4 9.8 3.9 13.7 S.4 6.7 15.1 13.0 7.6 20.6 6.3 5.2 11.5 6 61.3 3.6 65.0 60.3 18.3 78.6 68.4 18.4 86.8 65.0 30.1 95.1 38.8 60.0 98.8 7 77.4 5.5 82.9 86.4 8.0 94.4 74.4 17.9 92.3 64.9 33.3 9S.2 35.9 63.0 98.9 8 85.0 7.5 92.5 78.6 15.1 93.7 78.1 20.5 98.7 76.0 23.2 99.2 41.3 58.7 100.0 9 88.4 6.4 94.8 82.0 14.7 96.7 72.5 20.8 93.3 72.6 24.8 97.4 46.6 53.4 100.0 10 85.6 4.8 90.4 89.3 9.2 98.5 83.8 13.5 97.3 66.7 32.5 99.1 43.5 56.5 100.0 11 83.3 4.3 87.7 84.6 7.4 91.9 79.8 15.5 95.3 72.7 19.1 91.8 40.6 56.4 97.0 12 55.4 6.3 61.7 60.4 7.4 67.8 39.5 9.2 48.7 39.1 7.8 46.9 18.9 16.2 35.1 13 39.6 1.9 41.5 27.8 3.0 30.8 15.4 1.5 16.9 13.3 0.0 13.3 3.5 1.2 4.7 Expected Years in 6.3 0.4 6.7 6.0 0.7 6.7 5.3 1.3 6.6 5.0 1.8 6.8 2.8 3.8 6.6 Prinmay School SECONDARY 12 10.9 1.7 12.6 20.9 3.0 23.9 39.5 5.9 45.4 40.6 10.2 50.8 32.4 31.5 64.0 13 18.8 1.3 20.1 29.3 5.3 34.6 58.5 6.2 64.6 62.8 16.S 79.6 43.5 47.1 90.6 14 28.7 0.7 29.4 37.3 4.2 41.5 52.8 13.8 66.7 61.1 23.0 84.1 50.0 41.8 91.8 15 24.2 3.1 27.3 40.6 6.0 46.6 59.7 12.6 72.3 58.7 17.4 76.1 38.5 43.3 81.7 16 17.5 2.5 20.0 39.7 4.3 44.0 53.9 9.6 63.5 59.2 18.4 77.6 39.8 43.7 83.5 17 17.2 1.7 19.0 28.0 6.8 34.8 45.3 10.9 56.2 57.3 12.9 70.2 37.6 42.4 80.0 18 7.3 2.1 9.4 27.3 2.5 29.8 31.5 7.1 38.6 28.3 8.8 37.1 24.7 17.7 42.4 19 9.3 4.0 13.3 14.0 1.0 15.0 12.0 5.0 17.0 16.7 6.1 22.8 14.0 4.3 18.3 Expected Years in 1.7 0.2 1.9 2.8 0.4 3.2 4.1 0.9 5.0 4.4 1.3 5.7 3.1 2.9 6.0 Secondary School UNIVERSITY 18 3.1 0.0 3.1 0.8 0.0 0.8 1.6 0.0 1.6 8.8 2.7 11.5 8.8 14.2 23.0 19 0.0 0.0 0.0 2.0 0.0 2.0 8.0 2.0 10.0 15.8 5.3 21.1 11.8 20.4 32.3 20 0.0 0.0 0.0 6.2 1.0 7.2 4.3 4.3 8.6 10.4 3.1 13.5 13.5 22.5 36.0 21 3.3 0.0 3.3 4.4 1.1 5.5 6.2 0.0 6.2 19.6 4.9 24.5 21.0 32.1 53.1 22 0.0 0.0 0.0 3.0 0.0 3.0 8.4 2.8 11.2 14.0 3.7 17.7 14.6 12.5 27.1 23 0.0 0.0 0.0 1.2 2.5 3.7 2.2 0.0 2.2 16.7 4.1 20.8 20.3 9.7 30.0 Expeced Years in 0.1 0.0 0.1 0.3 0.1 0.4 0.7 0.2 0.9 1.5 0.4 1.9 1.9 1.8 3.7 University Source: ECV 1998 15 Concentration Curves of Public Spending by Education Level 1mI l Il ;Pmy A ; .8 - SeidaoSdy 0 N i y ...A///./ ° .6 .4 e .2 n v l I I F O .2 .4 .6 .8 Cumulative Proportion of Population Concentration Curves: Private Expenditure on Education 1-=Pre.prwyO 0 Pimaiy A Sdey 0 .8 .2 4a6i8 .6 U 1:6 ~ ~ ~ ~ ~ ~ ~~~1 .2 0 I T4 ~~~~ .2 ~. 4 6. Cumulative Proportion of Population 16 Health Table 7. Health Visits (Percentage) in the Month Prior to the Interview by Expenditure Quintile. Consumption Expenditure Quintile 1st 2nd 3rd 4th 5th Overall NO 95.1 92.6 90.4 86.4 82.0 89.3 YES 4.9 7.4 9.6 13.6 18.0 10.3 Private Clinic 26.7 27.7 35.1 47.3 60.9 45.2 Pbarnmacy,Home & other 14.0 16.1 11.2 14.4 18.7 12.6 Public 59.3 56.2 53.7 38.3 20.4 42.2 Hospital 32.6 39.5 45.5 51.0 67.2 48.1 Centro-de-Salud 20.6 22.5 24.1 26.2 17.5 22.4 Sub-Centro de Salud 56.8 38.0 28.3 22.8 15.3 29.5 Source: ECV 1998 Concentration Curves by Type of Public Health Facility :I I I I _ la 8 HospiI 0 ~~~~~m1 ~ ~ ~ ~ ~ 1 :.6 .4 .2 0 0 .2.4 .6 .81 Cumulative Proportion of Population 1 7 Colaci6n Escolar Concentration Curve LII 1X I.- *.2 0 .2 .4 .6. Cumulative Proportion of Population Alimentos Gratuitos * EWNFA A MPS ° MBS A. j.1 04 o .2 .4 .6 .8 Cumulative Proportion of Population 1~~~~~~~~~~~~~~ Annex 7. Employment Generation in Ecuador Mari Minowa November 11, 1999 Table of Contents L Crteria for Selecting Safety Net Programs ..............................2 II Alternatives for Income Support and Employment Generation .3 A. Job Skills Training .3 B. Micro-Enterprise Development .3 C. Public Employment Programs .4 II Existing Programs of Employment Generation in Ecuador .6 A. Common Issues of Existing Programs .6 3. Review of Existing Programs .10 1. FISE ...................... 10 2. PRONADER ...................... 11 3. COPEFEN/CORPECUADOR ...................... 12 4. Ministry of Education ................... 13 5. Ministry of Health ................... 14 6. MIDUVI ................... 15 7. Municipal Governments ................... 15 Appendix 1. Characteristics of Public Works Programs in Selected Countries Appendix 2. Impact Evaluation of Public Employment Schemes Appendix 3. Fondo de Solidaridad Appendix 4. COPEFEN: Summary of Funds Allocation and Transfer Employment Generation in Ecuador L Criteria for Selecting Safety Net Programs The Government has various alternative tools of intervention for providing income support and generating employment. They include, among others, programs such as food stamps, cash transfers and vouchers, public employment, micro-enterprise and self- employment assistance, and job training targeted to the poor. Governments need to assess the merits and demerits of each alternative program before deciding to implement specific interventions. The principal criteria for selecting among alternative measures would include the following: 1) Administrative feasibility. This depends on the detailed design of the program, the level of resources available for administration, the existing administrative capacity of the implementing entity, and the degree of imperfection that can be tolerated; 2) Political feasibility. This depends on how the program is promoted to the public, how coalitions of supporters or detractors are built, and the relative power of beneficiaries, suppliers, and administrators; 3) Corresponding effects on the poverty strategy. How will a safety net program affect, for example, the participants' labor supply, participation in other programs, and receipt of private inter-household transfers, and how will those changes affect markets and government finances? What will be the net effect on poverty reduction? 4) Potential for reaching the poor. Will the program reach significant number of the poor (coverage and exclusion error)? How much leakage of benefits will there be to the non-poor (inclusion error)? 5) Tailoring the solution to the problem. The program choice should address the real problem. Where the poor have suffered a loss of real wages rather than a loss of jobs, for example, transfers to the working poor may be more relevant than creating jobs. In the current case of Ecuador, however, both the real income loss and significantly increased unemployment need to be addressed. It is also important to note the heterogeneity of the impact for different groups. In times of economic crises, an additional important factor to consider is the speed with which the program can be implemented to quickly deliver the much-needed protection to the affected people. This objective of quick response can be achieved easier if the framework of a safety net program has already been set up, and can be flexibly expanded to respond to greater demands whenever necessary. The primary issue in designing and implementing a safety net program is how to deliver, efficiently and effectively, program benefits to the targeted groups. Thus, designing a cost-effective targeting mechanism is also a crucial factor for the success of a program. 2 In the following section, we will review the experiences of different types of income support/employment generation programs in other countries, with the objective of applying the lessons learned from those experiences to the current conditions in Ecuador. HI. Alternatives for Income Support and Employment Generation' A. Job Skills Training Publicly funded job skills training may improve trainees' skills and ernployability. However, it can help the long-term unemployed and disadvantaged workers find employment only when the economy is improving and demand for such workers is expanding. Impact evaluation studies have found that the cost-effectiveness of these programs is generally disappointing. The real rate of return of these training programs is rarely positive, and they are no more successful than job search assistance programs in terms of post-program placement and wages, while they cost anywhere between 2-4 times as much. The evaluation studies also have found that short-term training is not an adequate tool for updating and upgrading skills of workers displaced from declining industries. Tr-aining for youth generally does not have a positive impact and improves neither employment prospects nor post-training earnings. Taking costs into account, the real rates of return of these programs both in the short- and the long terms have been found to be usually negative.2 Evaluations do show that small-scale, tightly targeted on-the-job training programs aimed at women and other disadvantaged groups among the long-term unemployed often offer the best returns among different types of training programs. However, when there is little prospect of finding employment after completing training, such programs are too costly and ineffective as tools of providing income support for the poor. B. Micro-Enterprise Development Micro-enterprise development programs are usually taken up by only a small fraction of the unemployed and are associated with high deadweight and displacement effects. Evaluation of these programs in OECD countries found that the failure rate of these businesses is quite high in most cases, though businesses that are assisted through mentoring and business counseling are more likely to succeed. As in the case of training for the long-term unemployed, assistance targeted to particular groups, such as women and older individuals, seems to have a greater likelihood of success.3 Nonetheless, as a tool for employment generation and/or income support in times of economic crisis, micro-enterprise development assistance or micro-credit programs are V very few countries in Latin America have legaDly or administratively enacted an unemployment insurance system. While unemployment insurance could be considered as part of a safety net, its role under the current conditions in Ecuador is limited to providing modest income support for those new poor laid-off from formal sector employment There still remains a large segment of the poor who do not have employment contracts in the regulated sector. 2 See Dar and Tzannatos (1998) for more details of evaluation results from OECD countries. 3 Also see Dar and Tzannatos (1998). 3 not the most cost-effective option. It is true, on one hand, that these programs achieve effective income transfer to the households that own businesses, and are likely to help expand both employment and incomes associated with the business. However, on the other hand, the productivity of resources will be lowest in a crisis period, since the firm's income and profitability depends on the overall level of economic activity. C. Public Employment Programs Public employment programs serve as a safety net, providing income transfer and/or stabilization benefits to the poor, while at the same time using the beneficiaries' labor to build infrastructure for development. The program effectiveness depends on the benefits, costs, and the way resources are raised to finance the program. The cost-benefit calculation must take into account the wage income of the participants, the costs of participation, and the foregone income from alternative sources. It must also take into account the impact of the program on the labor market and the structure of wages. One important design feature of the program is to keep the wage level sufficiently low so that the poor are self-selected into the program. Stabilization benefits depend on the timing of the program, and may be high even when the transfer amount is small. Unlike job training programs or micro-enterprise development support, public employment programs mainly provide current benefits (temporary income) and are poor instruments as a permanent escape route from unemployment. Public employment programs are also expensive. As shown in Table 1, it costs 1.8 to 9.1 times the GDP per capita in these countries to create one temporary job through a public employment program. Despite its relatively high cost, however, a public employment program can be an attractive alternative as a short-term safety net intervention for the following reasons: (a) it simplifies targeting so long as the wage rate is appropriately set to ensure self-targeting to the most needy; (b) it is possible to create long-run benefits through assets created and maintained by the participants; and (c) as work has social merit, it prevents welfare dependency and disincentive effects. Table 1. Annual Cost of Job Creation through Public Employment Program 1. Cost/job (US$) 1401 ' 2120 2580 ' 786 ' 2700 5445 2122 2 .Cost/job (PPP) 7212 9759 14302 3620 9388 12100 10610 3. Per capita GDP (US$) 790 600 380 230 800 600 390 4. Ratio (1/3) 1.77 3.53 6.79 3.42 3.38 9.08 5.44 Source: Dar and Tzannatos (1998). In designing a public employment program, careful consideration needs be given to various key features of the program. The features of a good public employment program would include the following:4 4 The list is taken from Ravallion, Martin, "Appraising Workfare Schemes," (??) 1998. 4 * The wage rate should be set at a level which is no higher than the prevailing market wage for unskilled manual labor in the setting in which the scheme is introduced. * Restrictions on eligibility should be avoided. The fact that one wants work at this wage rate should ideally be the only requirement for eligibility. * If rationing is required because demand for work exceeds the budget available at the wage set, then the program should be targeted to poor areas, as indicated by a credible poverty map. However, flexibility should be allowed in future budget allocations across areas, to reflect differences in demand for the scheme. * The labor intensity (the share of the wage bill in total cost) should be higher than normal for similar projects in the same setting. How much higher will depend on the relative importance attached to immediate income gains versus (income and other) gains to the poor from the assets created. This will vary from setting to setting. * The projects should be targeted to poor areas, and should try to ensure that the assets created are of maximum value to poor people in those areas. Any exceptions - in which the assets largely benefit the non-poor - should require co-financing from the beneficiaries, and this money should go back into the budget of the scheme. * The performance of the program in reducing poverty should be monitored using careful evaluations. The wage level is the most important variable in program design. It is critical in determining how well the program will be targeted to the poor. Maintaining the program wage at or below the level of the ruling market wage for unskilled labor in the local labor market is the best way to ensure this. In doing so, some effort should be made to determine the appropriate subsistence wage and the ruling market wage, and to find out hoow they relate to the legal minimum wage. Again, the wage rate should not be higher thani the market wage for unskilled manual labor in agriculture or the informal sector during a normal year in the setting in which the program is introduced. For example, in the "Trabajar" program in Argentina, the wage rate was set at 2/3 of the average wage for the poorest decile in the Greater Buenos Aires area. The maximum wage allowed is the national minimum wage. Low wage rates ensure that: (a) there is self-targeting of the poor; (b) incentives are preserved for taking up regular work when available; and (c) there is a wide coverage of the poor, given the available budget. A rigorous ex-post evaluation of the program found that the distribution of the benefit is decidedly pro-poor, even after taking into account the foregone income of program 5 participants. Another example of a public employment program is the "Employment Guarantee Scheme" (EGS) in the Maharashtra state in India, which is often considered the most successful direct government effort at reducing absolute poverty in rural areas. If the 5 Jalan, Jyotsuna and Martin Ravallion, "Income Gains from Workfare: Estimates for Argentina's Trabajar Program Using Matching Methods," World Bank, 1998. 5 "Trabajar" program is an example of workfare in a medium income country where the emphasis is more on the value added of assets created through the program, the Maharashtra EGS is a program which focused more on the labor intensity of sub-projects and immediate income gains for the poor. Since the mid-1970s, the EGS has offered employment opportunities for unskilled workers who maintain rural infrastructure through small-scale irrigation and soil conservation projects, reforestation, and rural road construction. The program has been found to be well targeted and to provide sizable net transfers to the poor. However, a study has also found that a dramatic increase in the program wage in mid-1988 has made many of the poor worse off, as the budget constraint resulted in rationing of the benefit.6 (Appendix 1) presents the design features of selected workfare programs from Latin America and other regions. (Appendix 2) summarizes results of the impact evaluation studies for public employment programs in OECD countries. m. Existing Programs of Employment Generation in Ecuador A. Common Issues of Existing Programs Currently, Ecuador has several public programs that could contribute to generating temporary employment. The Emergency Social Investment Fund (Fondo de Inversion Social Emergente, FISE) finances community social infrastructure projects. Special entities, COPEFEN and CORPECUADOR, have been created for reconstruction and rehabilitation of infrastructure affected by El Nifio. There are also infrastructure components within the work of the Ministries of Education, Health, and Social Welfare, which could contribute to employment generation. Another obvious source of public infrastructure investment is the Ministry of Public Works, though their work tends to be large scale and more capital-intensive rather than human-intensive. The Ministry of Urban Development and Housing also invests a substantial amount in public housing construction. Finally, work carried out by municipal governments could also be used for broad-based employment creation. While these programs have potential for employment generation for the poor, the Government has not yet developed a well-defined strategy to organize these programs under the explicit objective of public employment generation. If the institutional infrastructure of the existing programs were to be used for expanding public employment in the short-run, a helpful next step may be to identify a designated super-entity within the Government which has the mandate and power to coordinate these various ministries and other autonomous public entities, and to develop a scheme for generating temporary employment for the poor. In doing so, design features and institutional capacity of each program need to be carefully assessed to determine necessary modifications. For each of these programs, Table 2 summarizes the targeting mechanism, program coverage, share of labor cost, estimated budget for FY99 and FY2000, and estimated unit 6 Ravallion, Martin, Gaurav Datt, and Shubham Chaudhuri, "Higher Wages for Relief Work Can Make Many of the Poor Worse Off: Recent Evidence from Mabarashtra's Employment Guarantee Scheme," World Bank, 1991. 6 cost of the employment created. We will review each program in more detail in the following section. Here, we point to some common issues, identified while reviewing these programs, that need to be addressed if these programs are to function as part of the safety net for the unemployed and the poor. They include the following: * Lack of Clear Targeting. Some programs, notably FISE and the infrastructure components of the projects financed by international lending institutions in the education and health sectors, have explicit selection/targeting criteria based on poverty indices. However, most of the existing infrastructure programs do not have a well-established, poverty-based targeting mechanism. During 1999, much of the infrastructure investment has focused on the reconstruction and rehabilitation of buildings, roads, and bridges damaged by El Nino. The extent of damage, rather than any poverty indicator, has dictated the selection of projects. If these programs were to be used as instruments for a social safety net and income support for the poor, it would be necessary to develop a suitable mechanism of p]roject selection and beneficiary targeting. * Budgetary Constraints Leading to Limited Coverage. Delays in the availability of funds have negatively affected program implementation. Particularly for the reconstruction and rehabilitation of infrastructure damaged by El Nifio, there has been criticism for its slow progress, although part of the explanation may be the initial planning work required. The same budgetary constraints limit the scope of employment generation in a short-run. * Legal impediments for Efficient Program Implementation, Especially the Public Sector Contracting Law. Since the Public Sector Contracting Law (Ley de Contrataci6n Puiblica) applies to the contracting of services in these infrastructure projects, administrative inefficiency in the compliance of the law tends to cause significant delay in contract processing. For these infrastructure projects to serve as effective public employment generation tools, a more flexible and streamlined contracting process is needed. This may include raising of the threshold costs of contracts to allow direct contracting or direct invitation; waiving of ex-ante approval of the National Comptrollers Office for smaller contracts; or decentralization of ex-ante approval authority to local governments. Certainly, the need for speed of implementation has to be balanced with the need to ensure transparent, fair execution of the program. However, the current situation, where the entire process of 7 Table 2. Assessment of the Existing Infrastructure Programs for Employment Generation Program Targeting Mechanism Size/Coverage Share of Labor Cost Activity/Budget for Activity/Budget for Unit Cost of One FY1999 FY2000 Workerday FISE Based on poverty index (As of June 30, 1999) Labor cost share: 300%. Initial total budget: 428 Based on the investment US$20 and prior benefits 1,039 contracts signed for billion sucres. Of which plan of FY99, USS40 received, at the parroquial USS41.6 million. 47 billion sucres for million worth of signed census zone level. Pre- Of which, 585 projects operational costs, and 381 contracts will be allocation of funds at the completed a feasibility billion sucres for outstanding. 25% ofthe canton level, based on study, and 165 projects investment. Expect to total budget of USS40 poverty index. under implementation. complete 1,000 projects million will be transferred (worth USS40 million). to CORPECUADOR. Average project cost: USS40,000. Maximum p_roject cost: USS250,000. COPEFEN No explicit targeting Total investment costs of 30-35% depending on the (As of July 21, 1999) None. US$18 mechanism exists. Funds US$167.7 million for size ofthe project. US$85 million has been (WB loan will close in the (COPEFEN-Municipal were allocated to 1998 and 1999. (WB, transferred to end of 1999.) Governments) municipal governments, IDB, CAF loans and implementing entities. ministries, and other counterpart funds.) Remaining US$76 million public entities, by a need to be transferred by directorate, consisting of Average project size: the end of 1999. the Vice president, USS15,000 for those President' delegate, a implemented by municipal delegate each from MOP, governments (42% of the MOF, and SEDES. total.) CORPECUADOR No explicit targeting Average annual Yet unknown Unless the Currently, has US$30 Will receive 25% of the N/A mechanism exists. "Plan investment of US$200 master plan includes million (of which US$14 "Fondo de Solidaridad" Maestro" to be completed million is planned. smaller (tertiary or million transferred from allocation (Estimated in August 1999. community) road COPEFEN). US$105 million). Division of labor: projects, the expected Additional source funds COPEFEN-rehabilitation, labor share is small. expected, but amount CORPECUADOR- unknown. reconstruction below altitude 1,000m, MOP- reconstruction above altitude 1,000m. PRONADER Project areas selected Total costs for the rural 25-30% US$10.4 million. None. US$9. based on: (a) road component: US$24.5 (The WB financed project concentration of small million over 8 years will close in June 2000.) farmers; (b) high poverty (1991-1999). level; and (c) potential for viable productive Of the total of about 70 activities. contracts signed for 1999, only 7 are large contracts (US$250,000_or greater). ___ 8 | Ministry of Health Reconstruction of basic The nrwect area cove 25-30%. USS6.2 miflionofbo. None, or rea-.inder ofthe US$20. health service about 2 million inhabitants WB financing and the supplemental loan infrastructure. Rapid (i.e. about 20% ofthe Average wage for an counterpart funds. approved in 1999. initial diagnosis in areas population.) unskilled worker: 30,000 In addition, USS20 selected for the FASBASE sucres/day (US$2.7/day). million supplemental loan project, based on poverty Average project duration indicators. is about one montk. MInistry of Educatlon Reconstruction of schools US$22.5 million for 40%. US$22.5 million for 1998 None. US$10. damaged by El Nino. school reconstruction and 1999. Of which Possible supplemental The COPEFEN/DINACE during 1998-1999. For each contract, about USS18 million has been loan. study identified 1,700 35 laborers are hired. committed and US$12.5 damaged schools. The average project size is Average project duration million has been disbursed US$22,500. is 3 months. as of July 20. School infrastructure as All contracts have been 25%/. (ITis labor cost US$10 million (?) N/A N/A part of the regular activity. less than 200 million divides into: 30%/. for Funding per demand, after sucres (approx. professionals, 50% for satisfying technical US$18,000) allowing skilled labor, and 20% for evaluation ofthe proposal. direct invitation. unskilled labor.) Ministry of Public No explicit targeting 16 contracts have been 30% for rehabilitation of N/A N/A N/A Works mechanism exits. signed for US$34 million primary roads. During 1999, 75% of the during Phase , and 8 total investment is for the additional contracts signed "reconstruction/rehabilitat during Phase II (since ion" of infrastructure November 1998). damaged by El Nino. Ministry of Urban N/A 93,000 worker*year for N/A N/A N/A USS25. Development and 170,000 housing units Housing reconstruction, 42,000 worker*year for 100,000 housing units improvement Municipal Governments N/A Of total estimated revenue 30-35°/o. Roughly US$500 million N/A US$18. of about US$500 million in total. (COPEFEN-Municipal per year, at least US$162 Governments.) million, but most likely greater amount, is spent on infrastructure investment. Sources: Various interviews with, and data provided by, the program executing agencies. 9 hiring workers for a small social infrastructure project takes 6 months on average, needs to be improved to allow faster responses. For those projects funded by multilateral financing agencies, the law allows that the terms agreed with the lending agencies supercede those in the national law. This should help the contracting of workers in emergency public employment projects funded by external financing. * Inadequate Institutional Capacity. Staff working on these programs could benefit from training in both their administrative and technical skills. Setting up clear and suitable project selection criteria and targeting mechanisms, for example, would require the technical assistance of an expert. Staff then will have to be trained to efficiently execute these mechanisms, including the management of an appropriate information management system for program monitoring. B. Review of Existing Programs In the short-run, during the remainder of 1999, effort should be made to maximize the temporary public employment generated by the existing social programs. Considering the current status of different programs, several among them can be identified as potential source of employment. In the following, we review these programs to assess whether they are suitable for protection and expansion, and their expected impact on employment generation. 1. FISE FISE is one of the few programs of temporary employment creation which has explicit selection criteria based on poverty indices. More specifically, funds are initially allocated at the canton level based on the poverty index calculated from the 1995 household survey (LSMS) and the 1990 Population Census. Furthermore, each parroquia (or in the case of large urban parroquias, census zones consisting of 2,000 inhabitants) is classified into three categories of urgency based on the "unsatisfied basic needs" index. While the basic principles of the selection criteria need not change, the poverty indices need to be updated with the results of the 1998 household survey. The Implementation Completion Report of the first phase of the FISE project maintains that employment generation for both skilled and unskilled workers in construction is the clearest quantifiable impact of the project. However, the Government has not monitored any quantitative measure of employment generated through the program. According to a rough calculation based on the figures given by the program office (i.e. US$40 million annual program budget, 30% labor cost share, unskilled worker wage rate of US$30/week,) a total of 400,000 worker*weeks of employment are being generated through the program in 1999 (assuming that all the labor costs are for unskilled labor).7 This implies that it costs the Government US$100 to provide one 7 According to the Central Bank data, there were about 280,000 unemployed workers in Cuenca, Guayaquil, and Quito in June 1999. Considering that the share of econonically active population of these three largest cities in Ecuador is about 50% of the national total, we could very roughly estimate that there were 560,000 unemployed in Ecuador as of June. This implies, therefore, that if all unskilled workers employed by the FISE projects worked 10 week of employment for an unskilled laborer. Or alternatively, the cost of creating one workler*day of temporary employment is US$20 (assuming a five-day-week). During the previous years' operation, FISE has experienced some problems due to allegations of corruption. Since then, a new set of operational procedures and project selection criteria has been established, and the program's transparency has been improved. One implementation problem that still remains is the long time (an average of about 6 months) that the process of project authorization requires due to various legal requirements and administrative inefficiencies. If this problem can be alleviated, and the administrative capacity increased with more emphasis on decentralizing program execution, then FISE will be a good instrument for expanding public employment for the poor. For the remainder of 1999, the Government's Emergency Social Program will maintain the budget of about 108 million sucres for FISE, funded by the allocation from Fondo de Solidaridad and external financing.8 If economic and fiscal conditions allow an increase in funding for the FISE, this program can be an important source of temporary public employment with the institutional infrastructure already in place. To maximize the program's capacity in providing employment for the poor, the Government may consider: (a) increasing the budget allocation; (b) promoting and favoring projects with higher labor intensity; and (c) simplifying the procedure of contract pre- authorization by Contraloria and Procuradoria to shorten the project authorization process. 2. PRONADER The National Rural Development Program (PRONADER) is a Bank-funded project, which has been under implementation by the Ministry of Social Welfare since 1991. The project is due to close in June 2000. Among other things, the program involves construction of irrigation schemes, flood control and drainage schemes, construction and maintenance of rural roads, and construction of small dams. For the fiscal year 1999, the program has an allocated budget of US$10.4 million for these infrastructure components, which roughly corresponds to the undisbursed amount of the Bank loan (US$12.6 million as of June 30). The particular projects to be funded by the program during 1999 have already been selected. The program office reports that between the three components of the program (i.e. reconstruction and rehabilitation of irrigation system, rural roads, flood control and drainage systems damaged by El Nino; Community Development Fund; and forestation components,) a total of 1,134,328 laborer*days of employment are expected to be created. It is also reported that the program added a new condition this year that contractors employ local labor rather than workers from outside the community. Out of about 70 contracts to be carried out during 1999, only seven are large contracts greater than US$250,000, with the remainder mostly small only for one week, then about 71% of the unemployed can be employed for one week through the FISE program during 1999. Or alternatively, about 8,300 unemployed workers (1.5%) can work for 48 weeks. s For a summary description of Fondo de Solidaridad, see Appendix 3. 11 contracts lasting 60 to 90 days. From these figures, the unit cost of generating one worker*day of employment can be calculated to be about US$9. Since the project is completing this year, and the follow-up project prepared to be financed by FAO (PROLOCAL) does not have any infrastructure component, the expected impact of PRONADER will be limited to the 1,134,328 worker*days of temporary employment to be generated during 1999. The new program, PROLOCAL, will focus on human capital development and investment in agricultural micro enterprises. It is planned that "mesas de cofinanciamiento" be formed as a venue for community development planning jointly between PROLOCAL and FISE. 3. COPEFEN/CORPECUADOR COPEFEN (Unidad Coordinadora del Fen6meno El Nino) was created in October 1997 with the objective of coordinating the rehabilitation of social infrastructure and roads damaged by El Nino. COPEFEN coordinates the work of the Ministry of Public Works, FISE and municipal governments. The first phase of the program (October-December 1997) corresponded to the preparation and design of the program before serious damage took place; the second phase (January-October, 1998) to responses to emergencies during the rain and flooding; and the third phase (November 1998 to present) to the rehabilitation of damaged infrastructure. The program has a total budget of US$167.7 million, of which US$146.7 million are to be spent directly on infrastructure projects. The funding comes from the World Bank, and IDB loans, and Government counterpart funds. The funds have already been allocated and "committed" to municipal governments (42%), Ministries of Public Works (43%), Ministries of Agriculture, Social Welfare, and Urban Development (3%), and other public entities (FISE, UNDP, CORPECUADOR, etc.) (12%). As of July 9, a total of US$84.2 million (57%) had been disbursed to the recipient entities.9 CORPECUADOR was established in August, 1998 the final political decision of the Alarcon administration, one day before his departure from office. It is headed by the Vice- President of the Republic, and headquartered in Guayaquil. Its funding is expected to come from the export and sales tax of bananas and other tax revenues. However, no funding existed for CORPECUADOR during 1998, and an agreement was signed in December, 1998 for COPEFEN to provide US$14 million to fund the activities of CORPECUADOR during 1999. They agreed on the division of labor, with COPEFEN responsible for "rehabilitation" and CORPECUADOR responsible for "reconstruction" of affected infrastructure. With the US$14 million, CORPECUADOR was to finance reconstruction of 28 bridges (about US$10 million) housing (US$1 million,) and development of the "master plan" (US$3 million). As of late July, six contracts for bridge reconstruction had been signed, and the remaining twenty-two contracts were being finalized. CORPECUADOR has been criticized for the slow progress in its work, which is mainly due to the time spent on developing the "master plan." 9 For a more detailed distribution of the funds, see Appendix 4. 12 In contrast, about 75% of the funds allocated by COPEFEN to municipal governments (total US$61.7 million) have been already disbursed with the work completed, and the remaining 25% of the work is expected to be completed by the end of 1999 or earlier. The average size of the wvork executed by the municipal governments under COPEFEN is US$15,000, and the share of labor cost is reported to be 33% on average. If we use the average wage of US$30/week for laborers, each project generates 165 worker*weeks of temporary employment. The total allocated cost of US$61.7 million would generate 678,700 worker*weeks of employment. This implies that it costs US$91 for COPEFEN-municipal governments to create one week of employment for an unskilled laborer, or US$18 for one worker*day of employment. As for the work carried out by CORPECUADOR, no data is yet available. However, we cannot expect a large impact on employment generation for the poor and the unskilled workers. The civil work under CORPECUADOR will focus on major reconstruction work with high capital intensity, and will not be suitable for the heavy use of unskilled labor. 4. Ministry of Education During 1999, the Ministry of Education plans to reconstruct and rehabilitate school infrastructure damaged by El Nino. The Basic Education Project (EB/PRODEC) financed by the World Bank has US$22.5 million budgeted for the 1998-99 period for infrastructure recornstruction in target areas.10 A joint study by COPEFEN and DINACE identified a total of 1,700 primary schools damaged by El Nino in coastal areas, of which the EB/PRODEC will finance the reconstruction/rehabilitation of 700 schools in rural areas.1' According to the program office, the share of labor cost in the total project cost is about 40%. Thus, US$9 million will be spent on the labor cost. The average size of a project is reported to be US$22,500 with about 35 laborers hired for each project. The EB/PRODEC recommends to the contractor that they hire local unskilled labor, and this recommendation apparently has been followed in most cases. In addition to some engineering work, the projects involve many labor-intensive tasks, such as water tank and septic tank installation, roof and ceiling replacement, and painting of walls. It takes oll average 3 months (or 13 weeks) from the signing of the contract to the delivery of the work. Thus, the total budget of US$22.5 million would generate 455,000 worker*weeks of temporary employment for unskilled labor, translated into the unit cost of US$10 for generating one worker*day of employment. In year 2000, the Ministry plans to construct and install 2,300 pre-fabricated classrooms, in collaboration with the municipal governments. The municipal governments will be responsible for building slab foundation, on which the Ministry will mount the classroom. Each pre-fabricated classroom is expected to take 10 workers one month to construct; thus the project '° Of ithe total budgeted amount, US$18 million has been "committed," and US$12.5 has been actually spent as of late July. " Rehabilitation of other 1,000 schools in urban areas is financed by an IDB-funded project. In addition, FISE will finance the reconstruction of 500 high schools (colegios) damnaged by El Nino. 13 will generate 23,000 worker*months, or 99,700 worker*weeks of temporary employment. The cost of this work is to be funded by part of the fees charged to bounced checks.'2 For further reconstruction/rehabilitation work to continue during 2000, the Government will need to obtain additional external financing. Negotiation for a supplemental loan is currently under way with the World Bank. 5. Ministry of Health The Ministry of Health has been funding the rehabilitation of health infrastructure damaged by El Nino through the Bank-financed FASBASE project. The infrastructure component of the project has been decentralized to sub-national units with financial and administrative management capacity, such as parroquias or the headquarters of "areas de salud " While initially contracts were given to qualified contractors from outside the community, now most contracts are small and managed locally, with the condition that local labor be hired. The selection of infrastructure projects under FASBASE has been based on demand, with a quick initial assessment of damages. The FASBASE office reports that it has taken only 8 to 15 days, on average, from the identification of demand and submission of estimated project costs to the signing of a contract. This program has taken full advantage of the fact that programs financed by international lending agencies are exempt from the condition imposed by the Public Sector Contracting Law (Ley de Contrataci6n Puiblica). Thus, the time spent on preparing and obtainin3 approval for contracts has been apparently much shorter than with other programs such as FISE. To improve the targeting of the program to the poor and the most vulnerable groups, a new set of project selection criteria has been developed recently. Based on poverty indices, 300 of the poorest parroqulas have been selected from 176 poor, or extremely poor, cantons. Additional criteria include: indigenous population; status of basic health service; availability of qualified human resources; and areas in the frontier. The World Bank financing of FASBASE will be completed in 1999. However, given the current crisis and increased need for additional funding, the Bank has agreed to provide a Supplemental Loan of US$20.2 million. Of this amount, US$11.02 million will be spent on infrastructure. Using the reported 30% as the share of labor cost in the total project cost, and US$30/week as the average wage for unskilled workers, this additional funding would create 92,000 worker*weeks, or 460,000 worker*day of temporary employment. The unit cost of generating one worker*day of employment through the FASBASE infrastructure component is calculated to be US$20. 12 This unique scheme was initiated in 1974 by the military. Initially, the collected money went to INNFA. Now, the Superintendencia de Bancos collects about US$6 to 7 million of bounced check fees every year, which is used for investment in school infrastructure and Ministry of Social Welfare programs. 13 Since FISE is also externally financed, it, too, can be exempted from the law. However, the implementing unit of FISE seems to have been unable to benefit from this exemption. 14 6. MIIDUVI According to the 1LO estimation, construction and/or reconstruction of 170,000 public housing units would generate about 93,000 worker*years of employment, and additional improvement of 100,000 housing units would generate 42,000 worker*years of employment.'4 Since the total cost of these investments is not available, we use the figures reported by COPEFEN for its work with MIDUVI in constructing/rehabilitating housing, to calculate the unit cost of temporary employment generation. It is reported that US$4 million is budgeted for the reconstruction of 1,000 housing units, at an average cost of US$4,000 per unit, of which US$2,000 is for the housing infrastructure itself, and the other US$2,000 is for related basic service infrastructure, such as water and sanitation. Thus, we can assume that US$680 million and US$200 million are spent respectively for the above-mentioned reconstruction and improvement. The unit cost of generating temporary public employment through this program is calculated to be US$6,500 per worker*year, or US$126 per worker*week, or US$25 per woriker*day. Although raising the labor intensity of these public housing projects could be considered, careful consideration needs to be made on the trade-off between employment generation objectives and the quality of housing produced. 7. Municipal Governments The Central Bank data show that duringl998 the total tax revenue for all municipal governments was US$78.6 million, of the equivalent of 0.4% of the GDP. In addition, muinicipal governments, received transfer of US$162.0 million from the central government, according to what is set forth in the "Ley Especial de Distribuci6n del 15%".1 With the additional transfers of US$274 million from other sources, the total municipal government revenue was US$514.6 million. 16 At the same time, the Central Bank data also show that the share of municipal government expenditure-- including the transfers from the central government-- was only 12.1% of the total public expenditure in 1998, amounting to about US$549 million. Although the revenue and expenditure figures do not correspond exactly, they give us the magnitude of money spent by municipal governments. Unfortunately, no data are available on the actual share of expenditure spent on infrastructure projects. However, it should be noted that the "Ley Especial de Distribuci6n del 159%' constrains that the transferred money be used only for investment, and not to cover current 14 OIT, OficiMa de Area y Equipo Tecnico para los Paises Andinos, "Propuestas de Politicas de Empleo para Ecuador a Corto y Mediano Plazo," julio 1999. '5 "La LeyEspecial de Distribuci6n del IS5' issued in March 1997 guarantees transfer of at least 15% of the central government budget to municipal governments. 16 Figures are taken from Wiesner, Eduardo, "La Descentralizaci6n, el Ajuste Fiscal y el Desarrollo Municipal en el Ecuador," Banco Interamericano de Desarrollo, BogotA, mayo 1999. 15 expenditure. Thus, we could assume that at least US$162 million (about one third of the total expenditure) was directed to infrastructure investment during 1998. This is a significant amount relative to the budget size of other programs reviewed here. Since municipal governments rely heavily on transfers, the current economic and fiscal crisis, and the resulting arrears of the transfer payments, have caused a serious problem. The expected transfer under the "Ley de 15%" in 1999 is US$268 million, which amounts to 11% of the total central Government budget, or 76% of the revenue of the US$353 million expected from financial transaction taxes during 1999. In theory, the fiscal impact of decentralization should be neutral to the public sector as a whole. To the extent that institutional and political capacity exists, municipal governments can take over from the national government execution of investment projects, and the process of transfer need not be hindered by the fiscal crisis. In reality, however, many municipal governments still lack the institutional capacity to effectively carry out many of the tasks currently performed by the national government. To help municipal governments play increasingly important roles in executing investment projects (including temporary employment generation through infrastructure projects,) we need to: (a) strengthen technical assistance for improving their administrative, financial, and other technical capacities; and (b) set up an incentive system that encourages and rewards efficient use of transferred funds. The development of municipal government capacity is a long-term goal. For the short- term, existing programs, such as FISE, COPEFEN, etc., should strengthen their collaboration with selected municipal governments in promoting temporary public employment generation. 16 Appendix 1 Characteristics of Public Works Programs in Selected Countries fiacd wit' WBhep mncia gvrnets:ovrmet with ?,?t'?5; .:?,:??'..,-''' .''-.f:"'"and expanded in 1997.) ,, mgiA0W&. Largely goverrmnent 60% WB, 30% state and Mostly the national ??i.;.}-?>.,.->.-finan ced, wit Bhelp. municipa govenuments, govermnent witl 10% beneficiaries usually contribution of state and in labor and materials. municipal governments. u.mb (May 1997-October 1998) (up to September 1998) Benefl?, iaries¢ t.'. . 300,000 workers 1.2 million workers %ofPopu. . -. 18,000 projects approved (60-70% of workers in drought-affected areas IT < 5>5, -; . jj C ? ;?.; ??> . . . ? ...........wanting work at the program wage rate.) iomeand/or Other:-. None, except the selection Rural areas. Only one Elb ityCi i of sub-projects based on participant per household, ?; , ;??.?-w????.? . g=>ji . - .... . . poin ts system. and no other family income source. (About US$ 100 .million/month.) of E m o en .C r.eate _ _ _ __.-_ _ _ _ _ _ _ _ _ _ _ t -. Wage C(st to-... 0.20 Estimated 0.75 Program W age/Remarks 2/3 of the average wage About equal to the average for the poorest decile in wage for casual wage GBA. Max. wage = labor in norml year, and .?.i' '''national minimum wage. close to the minimum wage. (Well above the shadow wage during the ____ ____ ____ ____ ___ drought) Tat;,,&.3Gr , ',,,.', Poor, unemployed workers Rural poor. Rural unemployed, ______________________ dro.e?t-affected workers. Targeting Me ?hanism. Self targeting through low Geographic targeting at Geographical targeting wage rates. municipality and rural based on rainfall data. Sub-prjects are selected community levels. (Limited to rura areas.) through points system Community-ba ed Jobs are rationed based on which places an emphasis selection by the municipal income level and losses .........?-.? ,.-,5,, on geographically council. due to drought targeting poor areas. Project-based selection by incentive structure and cost-sharing matrix. Thgex.p .r., .... ., 50% o of participants from 930/oof disbursed funds No data available. Bur .? .... 5 the poorest decile of reached targeted observation suggests that families. 80% from the beneficiaries (1995,) absolute rainfall data does poorest quintile. improvement from 44% not reflect the importance before program of relative rainfall. ,,_. ..... _ .,,, ?:i, ,,, reformulation. 17 Appendix 1 (Continued) Yea# ler~ented 1995 1987-1990 1987 m *Finaxiduig External Exteral (85%) through Exteral and national social fund government M3bErotBEnifieinries 25,000/month 8-9 million person days 40-45 million person days . (from 9/1995 to 8/1996) /year /year 60,000/month =r~t4 Xost (wage |kou- (total US$51.4 million) 8.0 0.5 ' gg WU RatioofW4gW ggsmarket wage for unskilled j "STaretGi~ou Rural and urban por. _______ Strenuous physical labor Highly self-targeted by the and far project site. very low wages. Selection by the implementing