Mexico - Primary Education Project

I M(iL)p I(1-i H (")pY Documer The WorlI R.p-;rt No. '"70-ME Type: e orz k XF<-)fll.* t / Xi:5I It 1 ]-/1 71/' iA,A-l- FOR OFFICIAL AAJ 3c,7-Ai Report No. 9770-ME STAFF APPRAISAL REPORT MEXICO PRIMARY EDUCATION PROJECT AUGUST 28, 1991 Country Department II Human Resources Operations Division Latin America and the Caribbean Regional Office IThis document has a restriced dsbution and may be used by redpents onWy in the performance of their ofaicrat duties. Its contents my not ot_erwise be disdosed wihout World ank authorizatIom  CURRENCY EQUIVALENTS Currency Unit = Peso (MEX$) US$1.00 - 3,022 Pesos (July 1991) FISCAL YEAR January 1 - December 31 ACADEMIC YEAR September 1 - June 30 UNMTS OF WEIGHTTS AND MEASURES Metric British/US Equivalent 1 kilogram (kg) = 2.20 pounds (lb) 1 metric ton (m ton) = 2,250 pounds 1 liter (1) - 0.26 gallons (gal) 1 cubic meter (m3) = 1,000 liters ACRONYMS CAM In-Service Teacher Training Centers (Centros de Actualizaci6n del Magisterio) CAPFCE Federal School Construction Agency (Comite Administrador del Programa Federal de Construcci6n de Escuelas) CONAFE National Council for Educational Development (Consejo Nacional de Fomento Educativo) CONACEN National Advisory Council for Teacher Training (Consejo Nacional Consultivo de Educaci6n Normal) CONALEP National System for Vocational and Technical Education (Colegio Nacional de Educaci6n Profesional Tecnica) DGEI General Directorate of Indigenous Education (Direccion General de Educaci6ii Indigena) DGEIR General Directorate for Evaluation (Direcci6n General de Evaluaci6n y de Incorporaci6n y Revalidaci6n) DGEP General Directorate of Primary Education (Direcci6n General de Educaci6n Primaria) DGPPYP General Directorate of Planning, Programnning and Budgeting in SEP (Direcci6n General de Planeaci6n, Programaci6n, y Presupuesto EPU Education Publications Unit (Unidad de Publicaci6n Educativa) GDP Gross Domestic Product INI National Institute for the Indigenous Population (Instituto Nacional Indigenista) NAfIN National Financing Company (Nacional Financiera, S.N.C.) PCU Project Coordinating Unit SCEP SEP's state-level education entities (Servicios Coordinados de Educaci6n Publica) SEP Secretariat of Education (Secretaria de Educaci6n Puiblica) SHCP Secretariat of Fmance (Secretaria de Hacienda y Credito Publico) SNTE National Teachers Union (Sindicato Nacional de Trabajadores Educativos) SOLIDARIDAD National Solidarity Program (Programa Nacional de Solidaridad) SPCU State Project Coordinating Unit SPP Federal Secretariat of Programming and Budgeting (Secretaria de Programaci6n y Presupuesto) UPN National Pedagogical University (Universidad Pedag6gica Nacional) USEC Undersecretariat for Educational Coordination (Subsecretaria de Coordinaci6n Educativa) UTE Educational Television Unit (Unidad de Televisi6n Educativa)  MEXICO FOR OFFICIAL USE ONLY PRIMARY EDUCATION PROJECT Table of Contents Page LOAN AND PROJECI SUMMARY ................................................. iii BASIC DATA .................,,,,,,,,,,,.................. vi DEFINITIONS ..... . vii I. INTRODUCTION. ...,,,, 1 II. THE EDUCATION SECOR .................................... 2 A. Overview of Sectoral Achievement. 2 B. Sector Organization. 2 C. Issues in Primary Education. 3 D. Government Strategy for the Education Sector. 8 E. Bank Role and Experience in the Education Sector. 9 F. Rationale for Bank Involvement .10 In. TE PROJECT .1 A. Project Concept and Objectives .11 B3. Project Areas and Target Population . 1 C. Project Components .12 D. Project Description .13 1. Educational Services Improvement Component .13 2. Institutional Strengthening Component .20 IV. PROJECT COSTS. PROCUREMENT. DISBURSEMENT. AND AU.I. 2.25 A. Project Costs and Financing .25 B. Procurement .27 C. Disbursements .29 D. Accounts and Audits .30 This report is based on the findings of an Appraisal mission that visited MexAco in March 1991. The mission was composed of Mmes/Messrs. Michelle Riboud (mission leader), Sam Carlson (economist), Eduardo Velez (education specialist), Maria Elena Anderson (operations analyst, consultant), M. Bravo (distance education specialist, consultant), Francois Orivel (education specialist, consultant) and A. Zuiiiga (architect, consultant). Darren Dorldn and David Warren provided editorial and preparation assistance. Th d M .1Ih i a rob tW-toel distrihut *n a 'I gi v he a t1 hv ro1w *tntn e liv in thA -rfnrfr -a  -li- V. PROJEC IMPLEMENTATION AND MA G;inx...... ..... 31 A. Project Readiness for Implementation ..................... 31 B. Project Organization and Management ............................ 31 VI. PROJECI1 BENEh-Fl AND RISKS ...................... .34 A. Project Benefits . . . .............. 34 B. Impact on Women .... 34 C Project Risks ............ 35 VII. AGREEMENTS REACHED AND RUECOMs.EN1). . . . ..ONS.. .36 ANNEXES (see also detailed Annex List, page 39) 1 - Comparative Education Data in Selected Countries 2 - Mexico: Educational Statistics 3 - Mexico: Efficiency Indicators 4 - Public Education Expenditures and Financing 3 - Sector Organization 6 - Project Description 7 - Project Costs 8 - Disbursement 9 - Implementation Schedule 10 - Key Monitoring Targets 11 - Key Documents Available in the Project File MAP IBRD No. 23053  - Hi - MEgCO PRIMARY EDUCATION PROJECT LOAN AND PROJECT SIMMARY Borrower: Nacional Financiera, S.N.C. (NAFIN) Guarantor: United Mexican States Beneficiaries: Secretariat of Education (SEP) and SEP's State Level Education Entities (SCEPs) in the states of Oaxaca, Chiapas, Guerrero and HidalgoY Amount: US$250.0 million equivalent Terms: Repayment in 17 years, including 5 years of grace at the standard variable interest rate. Project The project would improve the quality and efficiency of primary education, Objectives: focusing on four states (Oaxaca, Guerrero, Chiapas and Hidalgo) with the highest incidence of poverty and lowest education indicators. These objectives would be achieved through: (i) reducing the high repetition and dropout rates; (ii) raising the level of cognitive achievement of children; and (iii) strengthening management of the primary education system, including program design, implementation, resource allocation and evaluation. Project The project consists of two major components: (a) the Educational Services Description: Improvement Component (72 percent of the total project cost including contingencies) would: (i) provide educational materials, books for school libraries and some bilingual reading books in primary schools, mostly rural; (ii) upgrade skills of underqualified teachers and provide in-service training for primary school teachers and teacher trainers, using traditional methods as well as distance education technologies; and (iii) rehabilitate, replace and construct educational facilities; and (b) the Institutional Strengthening Component (28 percent of total project cost) would: (i) provide work incentives for teachers; (ii) improve the supervision system; (iii) strengthen the organization of the support service delivery system and the management capacity of SEP at the federal and state levels; (iv) support improvements in the information, monitoring and evaluation systems; and (v) conduct selected studies to assess the impact of project activities and help refine future program design. Ihe project implementation would be carried out by SEP at the central level and by SEP delegations (SCEP) at state level; the latter are responsible for the actual delivery, operation and management of educational services. Project The proposed project would support the Government programs for improving both the Benefits: quality and efficiency of primary education, focusing on four states with high incidence of poverty and low education indicators. The level of cognitive achievement of children as well as their probability to complete the primary education cycle are expected to increase. Benefits would thus consist in a higher rate of human capital accumulation contributing 1/ These State Level Education Entities (Servicios Coordinados de Educaci6n Publica) are delegations from SEP to the states.  I iv I to -xonomic development. They would also consist in a more equal distribution of economic opportunities among Mexican children. The project would improve the efficiency of public resource use by reducing repetition and dropout rates and by increasing the amount of investments with high social rates of return. Institutiona! strengthening would contribute to a better aliocation and increased productivity of human resources within the sector; and to an enhanced capacity to evaluate and monitor educational policies and programs. Approximately half of the children who would benefit from the project would be girls. Thus additional benefits are expected, as research has demonstrated that increasing the level of education of women has a significant positive effect on the health status and educati3nal attainment of their children and contnbutes to lower fertility. The impact may be particularly significant for girls from the indigenous population. Project Risks: The main project risks are associated with possible delays in project execution due to: (i) possible teachers' union opposition to Government policy changes; and (ii) limited implementation capacity at the state leveL The risk of teachers' union opposition is difficult to address. However, the improvement of work conditions for teachers and supervisors that the project would bring about and the recent success of the Government (and in particular, of the SCEP directors of the states involved) in handling union and employment issues provides some confidence that this risk can be minimized. The project attempts to reduce the second risk through: (a) the provision of training and technical assistance to the states; and (b) close monitoring from the central level. To increase SEP staff familiarity with Bank procedures, a project launch seminar to cover implementation aspects would also take place soon after loan effectiveness. The risk of slow implementation would also be alleviated, on the one hand, by the high priority given by the Government towards improving economic and social conditions in the poor southern project states, and on the other hand, by the strong commitment expressed in repeated occasions by SEP towards the project objectives as well as the dedication and work already invested in the project preparation.  ESTIMATED PROJECr COSW1 ProLject Comgonents Local Foreign .Total --US$ million--- Educational Servloes Improvement Coiuponent 184.2 34.3 218.5 Educational Materials 49.7 11.1 60.8 Teacher Training 32.8 6.1 38.9 Educational Facilities 101.7 17.1 118.8 Institutional Strengthening Component 80.3 6.2 86.5 Teacher Work Incentives & Supervision 67.4 3.3 70.7 Organization/Management 7.8 1.2 9.0 Information, Monitoring, Evaluation 5.1 1.7 6.8 & Studies TOTAL BASE COST 264.5 40.5 305.0 Physical Contingencies 12.3 2.1 14.4 Price Contingencies 30.8 1.8 32.6 TOTAL PROJECT COST 307.6 44.4 352.0 Financing Plan Local Foreign Total --US$ million----- Federal Government 102.0 - 102.0O' IBRD 205.6 44.4 250.0 TOTAL 307.6 44.4 352.01 Estimated Disbursements Bank Fiscal Year 1992 1993 1994 1995 1996 1997 -US$ million------------------- Annual 15.0k' 2&0 64.0 80.0 52.0 11.0 Cumulative 15.0 43.0 107.0 187.0 239.0 250.0 Economic Rate of Return: Not Applicable 2/ Includes taxes and duties (estimated at US$39.0 million). 3/ Includes Special Account deposit of US$12.0 million to cover retroactive financing of US$6.0 million for eligible expenditures incurred after March 22, 1991.  Ivi DAMI DATASr A. Geel Country Data GNP Per Capita (USS) ZO2O 1989 Area (Thousand Square Kilometers) 1,958 1989 Population Estimate (Millions) 81.1 1990 Urban Population (% of Total) 72 1989 Population Projection (Millions) Population for Year 2000 103 1989 B. Social Indicators Crude death rate (per 1,000) 6 1989 Average Annual Rate of Population Growth (%) 2.1 1989 Total Fertility Rate 3.4 1989 Infant Mortality Rate (per 1,000 live births) 40 1989 Life Expectancy at Birth (years) 69 1988 Adult Illiteracy Rate (% of persons 15 and over) 9.7 1988 C. Education Data Enrollment Rates Preschool (Age 4-5) (%) 68 1990 Primary (Net) (Grades 1-6) (%) 98 1990 Lower Secondary (Grades 7-9) (%) 57 1990 Education Expenditure Data Total Public Education Spending (% of GDP) 3.8 1989 Federal Education Spending (% of Total 79.0 1989 Public Education Expenditures) Federal Education Spending (% of Total 6.4 1989 Federal Expenditures) Federal Primary Education Spending (% of 23.5 1988 Total Federal Education Expenditures) Total Primary Enrollment (millions) 14.5 1990 Federal (%) 72 1990 State (%) 22 1990 Private (%) 6 1990 Primary School and Efficiency Data Number of Teachers (thousands) 467 1990 Student/Teacher Ratio 31.1 1990 Incomplete Schools (% of all primary schools) 20.0 1990 Unitary Schools (% of all primary schools) 21.0 1990 Repetition Rate (%) 10.0 1990 Dropout Rate (%) 5.3 1990 Completion Rate (%) 55.3 1990 Sources: World Development Report, 1991; World Tables (Third Edition); Basic Education Sector Report, World Bank, 1990; The State of the World's Children, 1988; INEGI - Resultados Preliminares del Censo de 1990, Primer Informe de Gobierno, 1989; and mission estimates.  I vil-I DEE;INMTONS Crude Death Rate : Number of deaths per 1,000 population in a given year. Infant Mortality Rate The number of deaths of infants under one year of age in a given year per 1,000 live births in that year. Life Expectancy at Birth : The average number of years an infant would live if the current as/sex-specific mortality trends prevailing at the time of birth were to continue. Total Fertility Rate : The average number of children who would be born alive to a woman during her lifetimne if she were to pass through her childbearing years conforming to the age-specific fertflity rates of a given year. Enrollment Ratio . Percentage of children of a given age group enrolled in schools at a particular level of education. Unitary Schools : Multi-grade schools with a single teacher. incomplete Schools . Primary schools with less than six grades. Completion Rate : The ratio between the number of students exiting grade six in a given year T and the number of students entering grade one in year T-6 (close approximation of the percentage of students of a given cohort who completed primary school).  1. INTRQDUCI O 1.01 Between 1950 and 1974, Mexico enjoyed a remarkable period of high growth, low inflation and moderate external debt accumulation. This era of economic stability came to an end when greater Government involvement in the economy Implied growing public expenditures unmatched by revenues. Inflation and external debt became increasingly important sources of public finance. Nonetheless, th, pace of Mexico's economic growth remained rapid until August 1982, when rising world interest rates, falling oil prices and refusal by extemal creditors to roll over Mexico's short term debt put an end to expansionary policies. The country entered a long period of declining income per capita and major structural reforms which have transformed Mexico into one of the most open economies in today's world. Those reforms include trade and financial market liberalization, tax reforms and a significant plan of public enterprise divestiture. They also include a major retrenchment of public expenditures. Since 1989, real per capita income growth has been restored, but at a slower pace than during the pre-1982 boom period. 1.02 During the period of rapid economic growth (1950-74), social welfare and standards of living improved considerably and there is evidence of progress on all social indicators. Illiteracy declined from 40% to 28% of the population over 14 years of age, the rate of infant mortality fell from 96 per thousand to 68, and average life erpectancy increased from 41 to 63 years. From 1974 to the onset of the economic crisis, progress in social indicators continued but at a slower rate. In spite of these achievements, poverty remained widespread and large regional and income disparities persisted. A 1980 study of changes in standard of living showed that, across states, the differences were considerable with leading states enjoying a living standard more than 100% higher than poorer states. With the economic crisis, social problems have become more acute as the poor, who are largely concentrated in rural areas, have particularly suffered from the deterioration in real income and the fiscal retrenchment, which affected social services expenditures more acutely. 1.03 Mexico's potential for growth is still constrained by relatively low levels of human capital formation: in spite of significant resources spent on education, over half of the population has less than six years of schooling. Major efforts are thus required to reach a higher rate of human capital accumulation needed for sustained economic development. In addition, a full success of the adjustment program hinges on policies that are effective in promoting social mobility and aimed at a more equal distribution of the benefits of economic development. These objectives can only be achieved through sustained human capital investments carried out to a much larger extent and with more efficiency than in the past, in order to reach all segments of the population. 1.04 After having focused its attention for several years on stabilizing and liberalizing the economy, the Mexican Government is now giving the highest priority to the development of human resources in order to help restore sustainable economic growth. It is also making major efforts to enlarge the basis for growth and reach the poorer segments of the population with programs aimed at improving the living conditions and earnings capaz -, of the poor. The proposed project would assist the Government in meeting these goals.  -2 - IL THE EDUCATION SECTOR A Overview of Sectoral Achievements 2.01 Mexico made great strides in increasing both the access to and the quality of public education between 1950 and 1980. During that period, illiteracy rates dropped from 40% to 18%, the number of schools tripled, the number of teachers quintupled and school enrollments increased sevenfold. Progress was evident on all fronts. By 1980, 25% of children aged 4-5 attended preschool, 91% of children aged 6-14 attt ided primary school and the enrollment rate in lower secondary school reached 48%. Since 1982, further progress was achieved but at a slower pace. Net enrollment rates at the primary level now reach 98%, and at the lower secondary level, 57%. The main efforts over the last decade have been concentrated on expanding preschool coverage: in 1989, 68% of children aged 4-5 had access to preschool education. In spite of these impressive quantitative achievements, the level of educational atiainment is still fairly low as the average number of years of schooling varies from 3 or 4 years in the poorest states to around 8 in the richest, with a national average of 6 years. Furthermore, several aspects of educational services have worsened due to low levels of efficiency and sharp decreases in the amount of resources invested in the sector. B. Sector Organization 2.02 The formal education system in Mexico includes basic education (preschool, primary and lower secondary) and post-basic education (upper secondary and university). Of a total enrollment of 21.4 million students at all levels of basic education in 1989/90, 68% were at the primary level (grades 1-6). The supply of basic educational services is almost entirely under the control of the public sector, and within it, federal authorities are predominant: only 6% of students attend private schools, 22%, state schools while the remaining 72% attend federal schools. The distribution is identical at the primary level. 2.03 The main institution in the education sector is the Secretariat for Public Education (SEP). SEP is one of the largest government ministries in Mexico, with some 950,000 employees spread across the Federal District and all 31 states. Seventy-four percent of SEP employees are teachers, 25% are general technical and administrative staff, and less than 1% are senior-level directors and administrators. The Directorate of Planning, Programming and Budgeting (DGPPYP) is at the core of SEP, formulating annual budgets, tracking educational indicators and statistics, and overseeing the activities of SEP's state-level entities, the Coordinated Services of Public Education (SCEPs). 2.04 The SCEPs have been establ;shed by SEP in each of the 31 Mexican states as part of the gradual decentralization process initiated in 1978. They have become key institutions for the actual delivery, operation and management of educational services, with a mandate limited to basic education (preschool, primary and lower secondary educition) and teacher training (both pre-service and in-service). By directly controlling all federal schools, that is, over two thirds of total public schools, they constitute the backbone of the decentralization process and their importance would grow further in the future as decentralization is pursued. (Annex 5 provides a detailed description of the current sharing of responsibilities between the Secretariat at the central leveL and the SCEPs). Another important institution in the sector is the National Teachers Union (SNTE) with nearly one million members, which represents Mexico's teachers and plays a key role in matters relative to teacher assignment, hiring and firing practices.  2.05 Primary education makes up the bulk of Mexico's educational system, reaching 14.5 million students each year through a network of nearly 81,000 schools and almost half a million teachers. Primary education is the only compulsory educational leveL Although it has suffered from the greatest declines in public spending since the onset of the economic crisis in 1982, it has retained the largest share (23.5%) of total expenditures within the sector (see Basic Data Sheet). 2.06 Public primary education services which correspond to a six year cycle, are provided in three variations: Renra. bilingualLindigenous and rural community courses (CONAFE). General primary education represents 94% of total enrollment. Bilingual/Indigenous education, which accounts for 5% of total enrollment, is specifically designed for the indipwnous population and is managed by a different directorate (DOEI) which recruits teachers among the indigenous communities. Over the last ten years, enrollment in these schools has grown steadily at an annual average rate of 5.5%, an indication of strong demand. The CONAFE community school system, in which specially trained lower secondary school graduates are sent to teach in schools built and maintained by the communities themselves, has b--en designed to overcome th problem of constructing and staffing schools in remote and isolated areas where it is very difficult to attract and retain teachers and where, given the small size of the community, it would be extremely costly to establish regular schools. In exchange for one year of teaching, CONAFE teachers receive a small monthly siipend during the teaching year and a three-year scholarship to pursue higher secondary studies thereafter. Room and board are provided to the teachers by the communities. Since its creation in 1971, the CONAFE model has reached rural students who otherwise would have received no education at all. C. Issues in Primary Education 2.07 As mentioned earlier (see para 2.01), primary education enrolment rates have reached a high level in Mexico. In 1989, the net enrollment rate for the age group 6-12 was estimated to be 98%. Although this implies that some 300,000 children still do not have access to primary education and that efforts to reach them need to be strengthened, coverav,e no longer appears as the principal problem at the primary education level 2.08 The main issues in this subsector are its low internal efficiency ad the low level of cognitive achievement of children. These problems are particularly acute in 10 Mexican states, eight of them located in the southern (and generally poorer) part of the country. 2.09 Low internal efficiency is characterized by high repetition (10.2%) and dropout rates (5.3%) at the national level. These rates are comparable to those in many other Latin-American countries and indicate that Mexico shares with them the same efficiency problem. The consequence of low efficiency is a slow flow of students resulting in a waste of scarce resources. Estimates!', based on 1988 data, suggest that these inefficiency costs amount to nearly 12% of total expenditures on primary education. True efficiency costs may, in fact, be even higher as school principals typically underestimate the number of repeaters. The problem is further aggravated by the wide disparities existing between and within states. For example, repetition rates vary from 6.5% in the Federal District of Mexico to 17% in Oaxaca. Within the state of Oaxaca, and in general within southern states, repetition rates as high as 30% can be observed in many communities (mostly rural). Rates are also significantly higher at the early stage of the primary cycle, where they can reach 40% to 50%. So is abandonment: every year, nearly 900,000 d/ See S. Carlson: 'Transforming the vicious circle - the costs and savings of school inefficiency in Mexico: a view from LATHR," No. 8, April 1991.  -4 - children drop out from school, half a million of them before completing the third grade, that is, before mastering the basic writing and reading skills. Fnally, disparities are also strong among the three variants of educational services: general primary schools are the most efficient and CONAFE schools the least. Indigenous/bilingual schools also indicate relatively inefficient student flows, mainly during the last half of the primary cycle. 2.10 High repetition and dropou. rates also result in a relatively low proportion of students completing the full six years of primary education. Available data indicate that only half of the students entering primary school complete the cycle. Again, wide disparities exist between states. In Chiapas, less than one third of the students ffiish primary school while in the Federal District, the proportion is nearly 80%. This implies that opportunities for gaining access to secondary education also vary greatly between states. Children from southern states which display the worst efficiency indicators are only half as likely as those from northem states to gain access to secondary school (1 out of 3 versus 2 out of 3). 2.11 Student cognitive achievement levels are low. Although no systematic nationwide achievement tests are conducted in Mexico, there are data on periodic testing which suggest that the quality of learning could be improved substantially. For example, a SEP evaluation in 1988, which tested students in Spanish, mathematics, and natural and social sciences, indicates that students' average scores in grades four to six were only slightly above 20 (out of a possible score of 100). 2.12 Low educational quality can of course be partly attrbuted to socio-economic factors. The lowest efficiency and quality indicators are observed in states in which the rural population is predominant, where a large portion of the indian population lives, and where natural geographical barriers to access and communication tend to be important. In general, low educational achievement and high incidence of poverty are found to be highly correlated. 2.13 Low efficiency and quality indicators can also be related and determined by a number of school-related factors. Among these, the most important are: (i) deficiencies in teacher training; (ii) lack of educational materials and inappropriate curriculum; (iii) teacher absenteeism and/or misallocation; (iv) weak supervision of teachers; (v) lack of evaluation; (vi) inadequate physical facilities; and (vii) strong financial constraints. 2.14 Deficiencies in Teacher Training. Many teachers seem to lack the theoretical skills needed to effectively reach children at the basic education level. First, only 30% on average, and less than 10% in the specific case of indigenous education, have the qualifications currently required by the Federal and State Governments to be hired, and in-service training programs are too short, sporadic and of low quality and effectiveness. Second, the lack of teacher training materials makes it difficult for teachers who do receive training to apply what is learned in the lectures. Third, teachers in rural communities often cannot attend training programs because they are offered mainly in urban areas, involving transportation, lodging and opportunity costs few can afford. Typically, these teachers are those most in need of training, because of inexperience, lack  6 - limited number of schools. It will stress problem-solving and practical applications. It will reverse the current practice of grouping several distinct subjects into general areas of study; instead, subjects will be taught separately in units that are pedagogically articulated to provide increased continuity between the three levels of basic education. TIhis nev curriculum has been developed in consultation with a large number of pedagogical experts and teachers working in a variety of areas and conditions. Once properly tested, it will be gradually implemented in primary schools starting, in 1992/93, with the lowest grades. Textbook conte-nt will be revised concurrently. 2.19 Teacher Misallocation and Absenteeism. Many teachers refuse to work in isolated rural communities for obvious reasons: work and living conditions are much barder, access is more difficult and costly, and rural teachers often spend 1-2 school days each week in transit, commuting from their homes to remote schools at their own cost. Furthermore, there is no incentive in terms of additional training opportunities, hardship pay or faster promotion to encourage teachers to take on more challenging assignments. As a result, teacher absenteeism is high, and some rural schools remain closed or students lose a good portion of class days each year (56 days or 30%o of the school year in southem states). Another consequence is the high turnover rate among rural teachers (for example, 17% have to be replaced every year in the southern states), resulting in the assignment of entry level teachers, who are the least experienced, to the most difficult grades and geographical areasY In other words, the children with the fewest educational opportunities receive the lowest quality instruction. These difficulties are further compounded by the hiring and teacher assignment practices which give great decision-making power to the SNTE; the union strongly supports teacher transfer to urban areas, resulting in overstaffing of many urban schools, while rural schools remain poorly staffed. Union problems are also tied to the problem of low teacher salaries, which have deteriorated in real terms by more than 33% since 1983. 2.20 Weak Supervision. Weak teacher supervision refers to both technicalpedagogical supervision and administrative oversighL It results in poor quality because: (a) teachers receive little or no feedback on their teaching techniques and effectiveness; (b) there is no follow-up on school improvement efforts and students achievements; and (c) there is no control over teacher attendance or effective textbook distnbution. The problem is especially acute in rural schools, which typically lack a principal, and have the least experienced teachers (i.e., those most in need of supervision). In those schools, and in particular in the southem states, supervision is deficient, mainly because of lack of financial resources. The number of schools allocated to each supervisor is often too high to allow for frequent visits-a problem that is compounded by the fact that many schools are dispersed and isolated, and access is thus difficult; transportation time to remote schools, in many cases, exceeds five hours and can reach twelve to twenty hours. Moreover, adequate transportation means do not always exist; and when they do, their costs (along with the costs of food and lodging) have to be supported by the supervisors. Additionally, training of supervisors is inadequate, and support materials to distnbute to teachers are extremely limited. Other causes of weak supervision are essentially bureaucratic: the appointment of supervisors is largely controlled by SNTE, which favors seniority and loyalty to the union as selection criteria, and there is no reward for carryi3g out proper supervision workV f/ For a description of teacher appointment and reassignment policies and practices, see Annex 5, Attachment 1. 7/ The supervisor appointment system is fully descnbed in SEP (DGPPYP), "Supervision in Primary Education (Supervisi6n y asesorfa en educaci6n primaria)," March 1991 (see project files).  of supervision, and the added difficulty of handling multi-grade teaching'! with children from poor communities. 2.15 Finally, a major problem also exists in the overall organization of teacher training. There does not seem to be a good coordination of the more than 100 training programs or courses, which include both preservice and in-service programs, run by the "Escuelas Normales," the National Pedagogical University and other institutions. The 1984 requirement that all new teachers obtain their "licenciatura" (three years beyond high school, ie., 15 years of schooling) has created much confusion. Curricula and instructor skills in the Escuelas Normales" have not yet been adapted to the new requirements. The disorganization has been compounded further by declining teacher trainee enrollments (10% average annual declines since 1981) which have not been accompanied by decreased expenditures or reduced teacher trainers. The results are increasingly high unit costs for teacher training, an excessively high trainee-trainer ratio (9:1), and general inefficiency in resource allocation. 2.16 Lack of Educational Materials and Inappropriate Curriculum. Mexican students benefit from an impressive, large-scale national textbook production system (over 70 million books are produced each year), which was initiated in 1959 and (theoretically) provides a full set of free textbooks to all primary students. However, serious problems remain. At present there is no formal, transparent system for textbook distnbution at the state-level. In many areas, particularly rural regions, students never receive their textbooks, or if they do they receive only some of them and/or receive them one to seven months after the school year begins. The problem is closely linked to weak supervision and to problems of access to rural and remote areas, in particular to the lack of transportation means and/or absence of financing for a distnbution system. It also seems to have been exacerbated by conflicts between SEP and the National Teacher's Union (SNTE) which has been partly controlling the textbook distnrbution. Obviously, the students are the real victins of these shortcomings. As a consequence, because textbooks are the primary, and often the only didactic materials students and teachers have available, faulty distribution means reduced educational quality as well as wasted time and money spent on production. 2.17 Besides the textbooks there are few educational materials available to students and teachers. Materials such as library books, maps, models, posters, school supplies, audio-visual aides and basic equipment for the natural sciences have proved to be useful to enrich, supplement, and apply the lessons provided in the textbook; they provide a variety of learning activities which are mutually reinforcing, an important element of a quality educational program. Furthermore, these materials typically belong to the school (and not the student, as is the case with textbooks), and so can benefit students who have 1-,st their textbooks and/or have never received them. The lack of school library books is especially serious, as few rural Mexican schoolchildren have access to any reading material besides their textbooks. This retards children's progress in developing reading and writing skills. The absence of additional edlucatioaal materials is particularly serious among indigenous children who often do not speak Spanish, usually have less qualified teachers and show the worst education indicators. The completion rate of these children for primary schooling is just 25% (compared to 56% nationwide) and their repetition rates is 24%, almost two and half times the national average (Annex 3, figure 1). 2.18 There is general consensus that the present primary school curriculum is inappropriate. A new curriculum has already been prepared and is being currently tested in a ./ Multi-grade teaching is still prevalent in Mexican primary schools. For example, unitary schools (multi-grade schools with only one teacher) comprise 21% of primary schools.  2.21 Lack of Evaluation. SEP and its General Directorate for Evaluation (DGEIR) have limited resources or capacity to monitor and evaluate academic achievement and standards of teaching across Mexico's educational system. Recently, the DOEIR lost much of its staff due to salary cuts, and many of those who remain lack sufficient training and skills. Furthermore, what evaluation research is conducted is typically limited in terms of sample coverage, periodicity, scope, depth of analysis and accessibility. Compounding the problem, there is practicalty no assessment capacity at the state level. The result is an educational system in which the most productive investments and reforms are difficult to determine due to a lack of understanding of the impact of previous investments. 2.22 Inadeguate Phvsical Facilities. In 1989 the number of classrooms needing to be built or repaired in Mexico was close to 57,000. The current classroom shortage (relative to current enrollments) is estimated at nearly 3,700 at the preschool level, almost 29,000 at the primary level, and approximately 6,000 at the secondary level. While multi-shift operations could mitigate this shortage to a certain extent, such programs have proved difficult to implement in rural areas, where the shortage is most pronounced. Despite these deficits, real primary education infrastructure expenditures have declined at an average annual rate of 183%, and in 1987 they were just 17% of their 1979 level. In addition, many classrooms lack sufficient desks, chairs and other furniture and equipment, and schools often have no funds for basic classroom operation and maintenance. In rural areas it is common to find classrooms with crumbling walls, weak or nonexistent lighting, or sometimes just a roof set over four columns. Community participation via Parents Associations is important for school maintenance, but in many areas these groups are weak or nonexistent. 2.23 Strone Financial Constraints. Between 1970-1982, federal spending for the education sector grew in real terms at an average annual rate of 14.1%, double the growth rate of GNP. Following the onset of the economic crisis in 1982, however, federal spending for education decreased in real terms at an average annual rate of 2.1%, four times faster than the decline in GNP and almost three times faster than the decline in overal public spending. By 1988 the federal education budget represented just 6.3% of overall federal public spending, down from 17.8% in 1976, and as a percentage of GDP total education spending declined from 5.3% in 1980/81 to 3.5% in 1988/89 (a 32% decrease). Given that this decline occurred while total enrollments, teaching staff and the number of schools were all increasing, this resulted in drastic reductions in real per pupil expenditures. In fact, these cuts have been more severe than those in other Latin American countries (Costa Rica, Chile and Venezuela) enacting fiscal austerity measures. Lastly, the balance of spending between recurrent and investment expenditures has changed. Almost 95% of federal spending now goes to recurrent costs (personnel and operating costs), up from 81% in 1970. 2.24 Public educational spending bI state governments has also declined in real terms, at an average annual rate of 11.6% between 1982-1987. It is clear that the decentralization program which began in the early 1980s has not yet brought about an increase in the responsibility and participation of the states in education financing. Although private spending is shown to have increased, it remains a marginal source of financing in the sector (see Annex 4, Table 1). 2.25 By educational level, primary education expenditures have suffered the most of all. Between 1979 and 1988 federal spending for primary education decreased in real terms at an average annual rate of 6.3%, while enrollments increased by 3.9%. In constant 1980 pesos, actual spending per pupil decreased by 46.5% between 1981 and 1988. Lower secondary education has also suffered; between 1979 and 1988 real expenditures declined at an average annual rate of  -8- 2.1%, while enrollments increased at an average annual rate of 5%. This translated into a decline in real spending per pupil of 42.7%. Meanwhile, real expenditures for preschool education, special education, upper secondary and post-graduate education all increased. Viewed as a whole, basic education (preschool, primary and lower secondary) expenditures decreased from 50% of total education spending in 1982 to 45.8%, while spending for upper secondary and university education increased from 29% to 38.5% of total spending. The fact that pupils at the basic education level make up approximately 85% of all Mexican students (21.4 million in 1989) indicates both the severity and breadth of the cutbacks in education expenditures. 2.26 Compared to many other Latin American countries (e.g. Chile, Costa Rica, Venezuela), Mexico has traditionally maintained low levels of education spending in relation to GDP, particularly in terms of per pupil spending on primary education. For example, Mexico's recurrent expenditures per primary student as a percentage of GNP per capita have averaged 4.5% between 1965 and 1985 while, in comparison, those in Chile, Costa Rica and Brazil have averaged 8.8%, 12.4% and 8.1%, respectively. Furthermore, there has been a long-term and increasing bias (as indicated by trends in spending) toward higher education, coupled with a recent emphasis on expanding preschool opportunities. Between 1983 and 1988, real per student spending for higher education increased an average of more than 15% per year, and averaged 15 times per pupil spending for basic education in 1988, while basic education spending declined at an average rate of 3% per year. Moreover, the fiscal austerity measures in Mexico since 1983 have placed priority on debt service payments and Mexico's productive sectors (industry and agriculture), such that the share of federal budget allocations to the education sector decreased. In summary, an education budget that was relatively small prior to the 1982 economic crisis has shrunk, and within the sector, preschool higher and post-graduate education have received preferential treatment over the primary level, although research has repeatedly shown that investment in primary education has a higher rate of return. D. Government Strategy for the Education Sector 2.27 The Education Modernization Program 1989-94 (PME) was announced by President Salinas de Gortari in October 1989. The PME expands on the National Development Plan 1989- 94, which reflects the Government's strong commitment to building up Mexico's human capitaL The PME has two overriding objectives: (i) to improve the quality of the educational system, with special emphasis on basic education; and (ii) to raise the level of schooling throughout the population. Additional objectives are: (iii) to strengthen community participation in all aspects of the sector; and (iv) to gradually pursue the decentralization of the educational system and adjust the distribution of educational functions to better serve the diverse. interests of society and needs for modernization. 2.28 More specifically, preschool and primary education coverage and quality are to improve through provision of educational materials, special support for community and bilingual schools, school construction and rehabilitation and new distance education programs. Incomplete and unitary primary schools are to be upgraded to offer all grades with more teachers. At the lower secondary level (grades 7-9), increased emphasis is given to science and technology, stressing research, critical analysis, experimentation and innovation. In addition, a new curriculum is being developed, structured according to specific subject matters (geography, biology, chemistry, etc.) as opposed to areas of knowledge (social science, science, mathematics and language). /I Figure for Brazil is an average from 1975-1985, not 1965-1985.  -9- 2.29 To further improve school efficiency and the quality of instruction, the Government plans to reorganize and integrate teacher training institutions, offering revised curricula and stressing methodologies which impart skills and knowledge for analysis and creativity. New in- service programs are planned for all of the country's approximately one million teachers, and innovative distance training programs should increase access of rural teachers to training opportunities. Also included on the Government's agenda are the gradual improvement of work conditions of teachers and supervisors and the gradual introduction of incentive systems to reward the most efficient teachers or those who take on chaDlenging tasks. A first step has already been taken by providing a significant increase (25%) in teachers' salaries in May 1991. 2.30 Within this overall strategy designed to cover the whole country, priority attention is being given to its implementation in the 14 states that display educational indicators below national average. The objective of this special regional emphasis is to compensate for the lack of private spending on education by families and communities due to higher incidence of poverty in those states, and thus to achieve greater equity in the distribution of opportunities at the national level. 2.31 Administratively, SEP also plans to continue its gradual process of decentralization, including ultimately not only the transfer of operational and technical responsibility to state-level education authorities, but also of limited financial autonomy (Annex 5 describes the current state of decentralization.) Meanwhile, planning, budgeting, curriculum design, norm setting, supervision, evaluation and teacher training will remain under federal controL In particular, the Government recognizes the need to upgrade its supervision and evaluation capabilities, and has proposed a National Evaluation System, covering all aspects of the education sector. Fmally, the Government is expanding its activities in promoting community participation, especially for school supplies, construction and maintenance. Viewed in its entirety, the Government program as presented in the PME is well-targeted, ambitious, and designed to improve both the quality of and access to basic education. Through its implementation, it should also contribute to redressing the imbalances in current resource allocation within the sector. E Bank Role and Experience in the Education Sector 2.32 The nrimary focus of Bank assistance to the sector in the past has been to help the Government increase workers' productivity so as to meet manpower requirements at the middle occupational leveL The emphasis has been on technical training, reducing the shortages of skilled technicians, long considered as a major constraint to industrial growth. The Bank has financed three training projects worth a total of US$251 million. Through the First and Second Technical Training Projects (CONALEP I, Loan 2042-ME of 1981, and CONALEP II, Loan 2559-ME of 1984) a nationwide system of 252 vocational training centers, with an enrollment of approximately 200,000 students, has been built, equipped and fumished. Through the 1987 Manpower Training Project (Loan 2876-ME) additional training programs (pre-service and in-service training as well as retraining for the unemployed) were supported. These projects have met their targets, and there have been no serious implementation problems. A Third Technical Training Project (CONALEP m, Loan 3358-ME) has just been approved by the Board of Directors (June 1991), with the objectives of improving and expanding CONALEP's vocational training, and its responsiveness and links to the labor market. 2.33 In addition to these manpower training projects, the Bank provided (following the 1985 earthquake) US$55 million for reconstruction, rehabilitation and reinforcement of public schools in five states (Colima, Guerrero, Michoacan, Chiapas and Oaxaca) and the City of Mexico  - 10- (Loan 2665-ME of 1986, Earthquake Rehabilitation and Reconstruction Project). Under the 1991 Decentralization and Regional Development Project (Loan 3310-ME), the Bank is also supporting efforts to increase the access of poor and indigenous populations to basic infrastructure and social services, financing in particular the construction and rehabilitation of public primary and secondary schools in four of the poorest Mexican states. 2.34 The Bank's role in the sector, however, has remained marginal in the sense that, with the exception of the above-described support to social infrastructure, it has not addressed the principal issues related to the main subsector, primary education. Public primary education enrollment constitutes 60% of total public enrollment and covers nearly 14 milion students. It also accounts for the largest segment (23.5%) of federal education expenditures. It is the core of Mexico's educational system and the basis of human capital formation. With recent important changes in Government policies in the subsector, the Bank now has an opportunity to make an important contnbution to human resources development, in particular by addressing the critical issue of diminishing resources to primary education. 2.35 Lessons Learned. The experience acquired by the Bank in the training subsector and in other poverty-related areas does not provide for the drawing of specific lessons applicable to the primary education subsector. However, the collaborative relationship developed between the Bank and NAFIN during the implementation of the above mentioned completed and ongoing projects would be an important asset for the implementation of the proposed project. F. Rationale for Bank Involvement 2.36 Reflecting the success of the economic adjustment program in recent years, the Bank's assistance strategy for Mexico is gradually shifting in emphasis from adjustment lending to investment lending geared towards assisting sustained growth and poverty alleviation programs. In support of growth, the Bank aims to assist the Government in shifting sector policies in order to improve the efficiency of public resource allocation. In support of poverty alleviation programs, the Bank's strategy would assist the Government's efforts in expanding the coverage of targeted social programs, while improving institutional capacity and the efficiency and equity of such programs. The proposed project is an integral part of this strategy as it would assist increased investments in human beings and social infrastructure necessary for growth, and would be targeted to the poor. By addressing the basic educational needs of the poor in the most disadvantaged states in Mexico, it would also complement actions undertaken in the context of the 1991 Basic Health Care Project (Loan 3272-ME), the 1990 Low Income Housing Project (Loan 3140-ME), and the 1991 Decentralization and Regional Development Project (Loan 3310-ME); and would be supplemented by the proposed Initial Education Project (under preparation) which would emphasize early childhood development as an effective means to impiove children's readiness for learning. The proposed project would thus contribute to a more equitable distnrbution of opportunities for welfare gains within the population. Moreover, by focusing on primary education, the project would assist tb:x Mexican Government in improving the efficiency of educational expenditures which account for a large share of government spending and have the highest rate of return. Fnally, it would be the first time that the Mexican Government opens up the basic education sector for assistance from external aid agencies.  - 11- ml. THE PROJECT A. Project Objectives 3.01 The objective of the project is to assist the Government of Mexico in improving the quality and efficiency of primary education, focusing on four Mexican states (Oaxaca, Guerrero, Chiapas and Hidalgo) with the highest incidence of poverty and low education indicators. These objectives, considered as being of the highest priority within the Government's Education Modernization Program, would be achieved through: (i) reducing the high repetition and dropout rates; (ii) raising the level of cognitive achievement of chfldren; and (iii) strengthening management of the primary education system, including program design and implementation, monitoring and evaluation system. B. Project Areas and Target Population 3.02 The four project states comprise 13.2% of the total Mexican population but a much higher proportion of those living in poverty. Sixty-five percent of the population of the four states (about 7 million persons) is estimated to live below the poverty line (below one minimum wage, equivalent to about US$100 per month in the project states); they account for one third of the nation's absolute poor. In addition, these states include nearly 40%o of the total indigenous population (native ethnic groups with distinct language and culture), many of whom live in remote, isolated villages. 3.03 The high incidence of poverty which can be associated with poor health and nutrition, restricted access to land and labor markets, and inadequate economic infrastructure (roads, power lines, etc.), is also highly correlated with an extremely low education leveL The average number of years of schooling in the four project states is only 4.4 years, with illiteracy estimated at 16%. Viewed together, these states have an average primary school completion rate of 39%, with dropout averaging 9.9%. Repetition rates, as reported by school principals, average 14.7%; however, more precise estimates (see Annex 3, Table 2) indicate that true repetition rates are significantly higher, ranging, in particular, from 28% to 53% in the first grade. As a result, inefficiency costs have been estimated to amount to about 31% of total spending on primary education in those states, and about one-fourth of total inefficiency costs measured at the national level. Significant resource savings would thus be gained by improving the efficiency of primary education in this area. 3.04 In all four states, the educational system is predominantly public (only 1.4% of schools are private). The control of public schools, however, differs slightly. In Oaxaca and Hidalgo, all public primary schools are managed and operated by federal authorities through SEP's delegations (SCEPs), while in Chiapas and Guerrero, both federal and state systems coexist, although the latter is still small and relatively concentrated around urban areas. For the whole project area, the federal system is thus predominant (92% of public schools) and even more so in rural areas. Indigenous schools and CONAFE community courses, in particular, are ali federal and only operate in rural areas. It is also worth mentioning that although state and federal schools are run by different entities, the Federal Government has the prerogative to provide direct support to state schools and to coordinate actions taken under both educational systems (Ley Federal de Educaci6n, Art. 25-29).  -12 - 3.05 The project is directed towards all public (federal and state) primary schools, students, teachers and supervisors in the four states. Several of its interventions, however, would be specifically targeted towards rural areas because it is in these areas that the problems of low quality and poor efficiency are most acute. It is also in these areas that the degree of deprivation and poverty and the lack of family financial support are greatest. 3.06 Table 3.1 below provides some summary statistics of the educational system and population to be covered at the initial stage of the project (detailed statistics by state and by school modality are presented in Annex 2). The total number of beneficiaries from the project would ultimately be significantly higher as new children cohorts (of almost half a million each) enter the school system every year. Table 3.1: EDUCATIONAL SYSTEM AND TARGET POPULATION AT INITLAL STAGE OF PROJECT IMPLEMENTATION (1991) SCHOOLS STUDENTS TEACHERS SUPERVISORS (thousands) (thousands) (thousands) Rural Urban Rural Urban Rural Urban Rural Urban TOTAL 13.1 2.6 1,340 838 46 25 1,196 104 FEDERAL 12.2 2.0 1,246 680 43 21 1,136 67 STATE 0.8 1.2 95 115 3 4 60 37 C. Project Components 3.07 The project consists of two major components: (a) The Educational Services Improvement Component (72 percent of the total cost including contingencies) would improve the availability and quality of educational services for the primary school students. It would: (i) provide educational materials, books for school libraries and some bilingual reading books in primary schools, mostly rural; (ii) upgrade skills of underqualified teachers and provide in-service training for primary school teachers and teacher trainers, using traditional methods as well as distance education technologies; and (iii) rehabilitate, replace and construct educational facilities; and (b) The Institutional Strengthening Component (28 percent of total cost including contingencies) would further improve the efficiency of the primary education sector. It would: (i) provide work incentives for teachers; (ii) improve the supervision system; (iii) strengthen the organization of the support service delivery system and management capacity of SEP at the federal and state levels; (iv) support improvements in the information, monitoring and evaluation systems; and (v) conduct selected studies to assess the impact of project activities and help refine future program design.  - 13 - D. Project Description 1. Educational Services Improvement Component 3.08 Under the Educational Services Improvement component (US$255 million including contingencies) the project would provide the required resources to increase the availability and quality of educational inputs in order to improve student achievement and increase student retention and promotion rates. a. Provision of Educational Materials. 3.09 The project would provide three types of educational materials: (i) didactic materials, such as maps, geometric blocks, word games, posters, transparencies and projectors; (ii) library books, which would cover all six grades of primary school, and include pedagogical material for teachers and a variety of reading material, which would also be of interest to the community at large; and (iii) bilingual reading books for about 300,000 indigenous primary schoolchildren speaking the seven main indigenous languages in the four project states. These materials, which are known to enhance student achievement, would complement the new curriculum, which has been revised this year with the objective of providing increased continuity between preschool, primary and secondary levels of education. They would also complement the long-established National Textbook Program, in which each student is to be provided at no charge a full set of textbooks for the grade in which that student is enrolled (improved distribution of these textbooks would be addressed in the project; see paras. 3.47 and 3.48). (i) Didactic Materials 3.10 Didactic materials would be provided to all federal and state r-ral schools (about 13,100) in two different packages: (a) a package containing a set of maps, laminated posters, a slide projector, etc., which would be distributed to each school; and (b) a package containing items such as dictionaries, rulers, math games, etc., which would be distributed to each class. Packages also include school supplies, such as notebooks, pencils, etc. On average, each child would receive about $7 worth of materials per year. However, the provision of materials would be targeted according to the degree of deprivation and poverty in the area to compensate for differences in support from parents' associations. Children in extremely isolated rural schools and impoverished areas showing the lowest educational indicaturs-namely, children in schools with just one or two teachers for all six grades, and in indigenous and CONAFE schools (such students represent 42 percent of rural school enrollment in the project states)-are the highest priority for improving educational services, and therefore would receive almost three times as much educational material (in terms of value) than other rural schoolchildren. Packages of materials (based on the number of teachers per school) would be distnbuted by school supervisors at the same time as textbooks are distributed, and would benefit from the same improved distribution system (see para. 3.47 and 3.48). The precise content of these packages would be readjusted, based on a survey of school needs for educational materials currently underway with grant financing provided by the Government of Japan and to be completed by November 30, 1991. Teachers would also receive training in the use and application of these materials as part of the in-service training to be provided by the project (see para. 3.14). Responsibility for this component would rest with the directorates managing general primary education and indigenous education (DGEP and DGEI, respectively) and CONAFE. Coordination would be ensured through the project coordination units at the central and state levels. At negotiations. the Government provided assurances that the results of the survey on the needs for didactic materials  - 14- for each school currently underway would be fumished to the Bank for review no later than November 30. 1991 and would be used to adjust the precise content of the didactic materials packages (see pam. 7.01 (a)). (ii) Library Books 3.11 Library books would be provided to rural and urban public schools in the four project states under the national lbrary books program (Proerama de Rincones de Lectura), which has been in operation since 1986 on a small scale and in some urban areas; upon the request of parents' associations, approximately one third of urban schools have voluntarily adhered to the program. This program is to be extended to the whole country, with special emphasis on rural and indigenous schools. T1here would be one month of radio advertising to encourage schools to form a library committee made up of a teacher, a parent and a student, who would then become contractually responsible for establishing and maintaining the classroom libraries. 3.12 A set of 152 books (90 for students and the remainder for teachers and parents), already selected by a national library book commission, would be distributed to schools that sign up for the program. The package would also include a cassette for training of teachers and students in the use and upkeep of classroom libraries. During the first year of the project, books would be provided to indigenous schools and rural schools with 1-5 teachers; in later years coverage would increase to urban schools and larger rural schools. In addition, all schools that later wish to expand their small libraries would be able to purchase additional sets of books at cost. The Mexican postal system, through a contract with SEP, would deliver the packages of books to post offices close to the target communities. The school library committee would ensure that the book packages reach the school. Over its life, the project would provide about 2.5 million library books at an average unit cost of US$1.25 per book, bookcases, library cards and audio cassettes for training. It would also cover the cost of packaging and mailing. Responsibility for this program would rest with the Educational Publications Unit (EPU) of SEP. (iii) Bilingual Reading Books 3.13 As stated in para. 3.09, bilingual reading books accompanied by teacher guides and workbooks would be produced in the seven most prevalent indigenous languages in the four project states. These seven languages are spoken by approximately 60 percent of all indigenous primary schoolchildren in Mexico. Overall responstbility for this component would rest with the General Directorate of Indigenous Education (DGEI). Book production would first require extensive fieldwork by ethnolinguists to revise the existing fist and second grade prototype reading books so that they cover a wider range of dialects. After revision and validation in the field, these books would be redesigned using desktop publishing capabilities, which would be provided, along with training in their use, under the project. Subsequently, reading books for grades 3 and 4 would also be developed and produced by the project. These books would be bilingual, allowing students to develop reading and writing skills in both Spanish and the indigenous language. This strategy of bilingual education in early grades reflects the Government's policy aimed at reducing repetition and dropoust rates and helping indigenous- language speaking students make the transition to Spanish instruction. Over the four years of the project, approximately 1,100,000 bilingual reading books, workbooks and teacher guides would be produced. In addition, indigenous teachers would receive instruction in the use of these reading books in the classroom, as part of the in-service training to be provided under the project (see panL 3.16). Besides being part of the overall project evaluation, the implementation of the  - 15 - program would be accompanied by a specific study that would evaluate the degree of utilization of the materials, the degree of acceptance by children, teachers and communities, and the program's impact on children's achievement (see para. 3.55). b. Teacher Training. 3.14 The project would support an extensive training program aimed at upgrading skills of teachers, school principals and supervisors. The main objective is to increase teachers' subject matter knowledge, their command of instructional techniques, and their ability to handle multigrade teaching, monitor and evaluate student performance. A second objective is to improve the technical support capacity of school principals and supervisors so th'it they can be more effective in assisting teachc. . To complement these activities, the project would also support a training program aimed at raising the quality of pre-service training and directed to current professors in teacher training institutes. 3.15 The project would provide three types of training activities, as explained below (and in more detail in Annex 6, Aetachment 1): (i) direct in-person instruction during summer months; (ii) semi-direct (or semi-schooled) courses, one course per semester; and (iii) distance training activities via radio, audiocassettes, videocassettes, television, and pamphlets during the school year. These activities would complement an optional certification program begun in 1990 through which primary school teachers have the opportunity to earn higher level degrees. The overali strategy is to reach all teachers in the four states by the end of the project by providing them access to several training sessions throughout the year, offered in these various formats. Level of attendance and intensity of training received, however, is expected to vary according to school modality (again, see below and in Annex 6, Attachment 1). Special attention would be given to those with low levels of formal general education, little teaching experience, and who, typically, are working in rural and isolated areas. (i) Direct In-Person Instruction 3.16 Direct in-person instruction would be offered in four different modalities adapted to the specific conditions and needs of the personnel belonging to the various school types: (a) training for general primarv teachers, principals and supervisors working in rural areas (31,852); (b) similar but modified training for teachers and administrators working in urban zones (24,531); (c) training for indigenous teachers, principals and supervisors, providing bilingual instruction (13,353); and (d) training for the approximately 1,200 CON,E community instructors working in extremely isolated and poor areas. Training would be provided during one-week intensive (40- hour) courses provided over the summer, except for CONAFE instructors who, because of their lack of experience and low level of qualifications, are required to attend ten weeks of intensive training immediately after recruitment. (ii) Semi-Direct (Semester-long) Training 3.17 A second type of training would be provided to rural and urban general primary teachers, in which teachers would attend semi-monthly seminar/workshops (four hours each) on Saturdays or on pay days. These seminar/workshops would be provided in tandem with anthologies, workbooks, self-learning guides and radio programs that teachers would follow at home on a daily basis in between the seminar/workshops. Four month-long modules would be presented for each course, with the core course on basic literacy and numeracy teaching techniques repeated each year during the first semester. The second semester course would vary  - 16- each year and would focus on the natural and social sciences, curriculum planning, child development and student evaluation. The semi-direct approach would provide greater continuity and in-depth treatment of subjects than would be possible with short, intensive courses. 3.18 A similar semi-direct in-service training program would be offered to professors (approxmately 860) in the following teacher training institutions: "Normal" schools, in-service training centers (CAM) and branches of the National Pedagogic University (UPN). This program would consist of six different semi-schooled, three month training courses, accompanied by monthly seminars and workshops. It would aim at upgrading the skills of these professors and at raising the quality of pre-scrvice training and would complement efforts currently underway to modify the curricula of teacher training institutions (see parm 2.15). Finally, CONAFE irstructors would receive monthly in-service training throughout the school year. 3.19 The sites used for the training sessions would vary according to the modality. Indigenous teachers, C(ONAFE instructors and professors from teacher training institutions would use existing facilities (about 42) operated by their respective General Directorates. For the general primary teachers, who are geographically dispersed and comprise the largest group, it was found to be more cost-effective to organize the sessions at the level of the supervision zone, using existing primary school facilities. Attendance would be compulsory for both CONAFE instructors and professors from teacher training institutions, but voluntary for general primary and indigenous teachers (legislation prevents SEP from obliging teachers to attend training sessions on non- working days). Incentives, however, would be offered in the form of per diems and promotion points (proportional to the length of the training period and conditional upon a positive evaluation at the end of training) (see Annex 6, Attachment 1 for additional details). At negotiations. the Government submitted a letter, dated Augut 8. 1991. confirming agreements reached at appraisal regardi.:g the mechanisms to be used to monitor teacher attendance to tra'nn sesinsand distribute incentives (per diems and promotion points) (stee a 7.01 Mb). 3.20 For all training modalities, the project would support: (i) the planning and organization of the program; (ii) the preparation of course materials; (iii) the selection and training of course instructors and of technical groups (who would design and manage the courses); (iv) the actual provision of instruction, including compensations to course instructors and per diems to cover transport (and in some cases, room and board) expenses for both trainers and trainees; and (v) program supervision and evaluation. In the case of indigenous training, it would also finance refurbishing of the 16 training centers of the DGEI. 3.21 The whole program would be managed at the federal and state levels by the General Directorate corresponding to each school type (DGEP for primary education, DGEI for indigenous education, CONAFE for community schools, and CONACEN for teacher trainers) as weli as by small technical groups established both at the central and state level (composed of specialists coming both from within SEP and from other Mexican higher education institutions specialized in teacher training). This management structure has been established. The central technical groups would be responsible for the overall planning of the program, for training of the state technical groups, for defining with them the training modules and for preparing, testing and producing the course materials. The state technical groups would be responsible for selecting and training the instructors, implementing and supervising the program, and for monitoring teachers' progress with support from the central level. Trainers would be selected among the most qualified and experienced teachers and supervisors as well as among the personnel of teacher training institutions. (Annex 6, Attachment 1 provides additional details on the organization of the program.)  - 17- 3.22 Implementation would be gradual, with focus during the first year of the project on the constitution of technical support groups and on the preparation of course materials. The implementation pace would be revised, if needed, during the annual and mid-term project reviews (see para. 5.14 and 5.16). At negotiations, the Government provided assurances that a detailed description of the teacher training courses (consisting of course objectives, subject matter, time allocation and instructional techniques) and training materials would be submitted to the Bank. for revO,w,v June 30. 1992 (St- 02Q1m ( )). (iii) Distance Education Technologies 3.23 The project would support the use of distance education technologies to complement in-person instruction. The objective is to improve the effectiveness of training by: (a) ensuring a common level of up-to-date information and knowledge to ali teachers; (b) allowing continuous, year-round access to training materials; and (c) providing additional support to teachers for classroom instruction. 3.24 To ensure a common level of knowledge and information to all teachers, the project would support the production and duplication of 60 video training programs accompanied by printed guidelines for both trainers and trainees. These materidls would be used, in conjunction with the in-person instruction programs described above, in the most intensively used training centers which would be equipped with VCRs and TV monitors. It is estimated that over 35,000 teachers would benefit from these programs over the life of the project (see Annec 6, Attachment 1 for a detailed description of the program, including contents and schedule). 3.25 To provide year-round access to training materials and additional support during teaching activities, the project would also provide, over the project life, 270 audioprograms (cassettes with instructional manuals) to be used in the classroom. In addition, 540 radio programs would be aired (approximately one 15-minute program per school day) via open broadcast, making use of free transmission time guaranteed by law to the Government by both private and public radio stations. Both instruments would cover the areas of curriculum development, instructional methods and evaluation. These programs are expected to receive wide acceptance as they would build on successful past experiments (audioprimaria) and because radio is widely used over the whole project area. All primary schools would be equipped with radio cassettes and beneficiaries would be all primary teachers as well as about three million children over the life of the project. 3.26 To reinforce indigenous teachers' skills in the use of bilingual reading books, an additional production of radio programs (40 programs per language) would be provided under the project. The DGEI would be responsible for the definition of their contents and the INI would be in charge of production and broadcast, making use of its stations which currently serve the indian populations of the four states. 3.27 The project would also support two experimental activities: (i) a pilot in-school TV educational program; and (ii) the production of a radio serial to promote community participation in educational activities. The TV pilot project would explore the possibilities of using an extensive mass media infrastructure and the many years of Mexican experience in the area (telesecundaria) to produce and transmit educational programs to provide additional support to teachers for classroom instruction. The scope of the project would be modest, concentrating on the state of Oaxaca where a local public TV and radio station operates successfully. The project would support the production of materials and the provision of TV receivers in 300 schools.  - 18- Production would be carried out jointly by the local state TV (channel 8) and UTE (the educational television unit of SEP), and would be followed by an evaluation of the impact and cost-effectiveness of these experimental programs (see para. 355). The radio serial, which would be accompanied by posters and written promotional materiaL would be designed to encourage family and community participation in school-related activities and tasks (school maintenance and repairs, support to teachers and administrators, parents' associations, community and family life, etc.). The serial would be transmitted on week-ends over the whole project area, making use of available free time guaranteed by law and would also serve as a testing ground enabling experimentation with and refining of contents and corresponding evaluation. 3.28 Responsibility for planning and organizing the distance education program would rest with the director of the Educational Television Unit (UTE) of SEP, who has appointed a general coordinator and a technical committee composed of representatives from the three different primary education modalities (DGEP, DGEI and CONAFE). This technical committee, assisted with specialists contracted under the project, would define the detailed contents of both radio and video programs. A core module common to all modalities would be designed, and supplemented by additional programs corresponding to the specific needs of each of them. These modules would complement in a cost-effective manner the formal training program described above and in Annex 6, Attachment 1. Implementation of the program would be gradual with focus on the first year of the project on the organization and design of the program contents. Production would take place over the following three years. At negotiations. the Government provided assurances that thy detailed descriptions of all distance education programs, as well as their transmission schedule in the case of orograms to be broadcas. uld be presented to the Bank. for approval. by April 30. 1992: and submitted evidence, satisfactory to the Bank. that Radio-Television de Vaxaca has_the nec:ssary human and physical reso'grces for the development and application of pilot educational ,rograms (see Larn. 7.01 (d!!- As a condition for disbursement. the Government would present a detailed proposal. satisfactory to thle Bank. for the work to be done for the pilot programs by Radio-Television de Oaxaca. snecify;iwg the implementation schedule (see para. 7.03 1. c. Improvement and Expansi- i of Educational Facilities. 3.29 The project would provide financing for tbe rehabilitation, expansion, maintenance and equipment of both federal and state public school cilities, and, in the case of expansion, for incremental staff. These actions would complement, ana would be carried out in close coordination with, those financed by the Government's Solidaridad Program and supported under the Decentralization and Regional Development project, which also aims to improve social infrastructure in these four states (see para. 331). 3.30 The total civil works requirements were estimated based on: (a) the need to improve the existing school facilities replacing rented, borrowed or dilapidated classrooms and support facilities, or those which do not conform to functional, pedagogical or architectural standards; (b) the determination of classrooms in need of repairs; (c) the need to expand schools to accommodate current enrollments (up to six classrooms if, on a one-shift system, the school has at least 25 students per classroom, and to more than six classrooms if, on a double-shift system, it has more than 45 students per classroom); (d) the need to provide structural reinforcement in highly seismic areas; and (e) the need to cover expanded enrollment expected to occur in indigenous areas and as a result of increased completion rates. Because of the sharp decline in investment expenditures in the sector over recent years, the provision and maintenance of infrastructure has been largely lagging behind the provision of teachers (see para. 2.22).  - 19- Therefore, highest priority would be given to meeting 100%o of requirements in terms of rehabilitation, repairs and expansion of existing facilities, to accommodate the current level of enrollment; about 50% of requirements to accommodate the expected future expanded enrollment would be met. 3.31 Total requirements, estimated on the basis of these priority criteria, are summarized in Table 3.2 below. They include the rehabilitation of an estimated 6,538 classrooms and 4,767 support facilities as well as the repairs of some 1,174 existing public schools and structural reinforcement of some 689 establishments in highly seismic areas. They also include the construction and equipping of approximately 2,124 new classrooms and 1,068 support facilities. The project would contribute towards meeting these needs by providing the required funding to supplement the estimated US$55.8 million primary school investment program already financed under the Government's Solidaridad program and under the Bank's Decentralization and Regional Development project (Loan 3310-ME, approved by the Board on March 26, 1991). The SEP program is presented separately in the last column of Table 3.2, and the detailed program by year is presented in Annex 6, Altathment 2. The coordination of the two investment programs (SEP and Solidaridad) would be ensured by the folowing existing practice: (a) SEP is the only institution responsible for defining sector needs and priorities; (b) aU civil works done under the Solidaridad program (type of works and specific location) have to be previously authorized by the SCEPs directors (and state authorities); and (c) no authorization for civil works can be given without all the necessary educational inputs (such as teachers, books, furniture) previously included in the detailed annual planning programs and budgets. Detailed information on civil works carried out under the Solidaridad program for the first year of the project was submitted to the Bank during appraisal. At negotiations, the Government agreed that: (i for subsequent years. a list of civil works carried out by SEP and under the Solidaridad Program would be p2rovided to the Bank during the annual work plan reviews: (ii) on this basis, classroom requirements and distribution of civil works between new construction and rehabilitation done under this project would be readjusted: and (iii) all civil works would be selected in accordance with criteria and Driorities stated in Dara. 3.30 (see parg 7.01 (e)). 3.32 The civil works program executed under this project would be carried out by CAPFCE, a federal school construction agency with long-term experience, which has already been the executing agency in a former Bank project, the Earthquake Reconstruction and Rehabilitation project (Loan 2665-ME), and whose performance in this context was judged excellent. 3.33 All new facilities would follow prototype designs that have been reviewed and accepted during appraisal and final working drawings would take into account regional climatic and construction processes. The rehabilitation of facilities would also be guided as much as possible by the prototype designs for new facilities. The acquisition of new furniture and equipment under the project would follow prototype lists per facility type which have been reviewed and accepted by the Bank. The upgrading program would repair, substitute and/or expand the existing stock to comply with the approved lists. 3.34 At negotiations. SEP submitted evidence that technical plans for at least one third of planned school rehabilitation works for the first year of the project had been prepared and it afreed to submit the remaining two thirds during the rest of the year. For subsequent years. technical plans of the following year rehabilitation work program would be submitted during the annual reviews (see paa 7.01 (f)). No new construction has been programmed for the first year of the project. For the subsequent years of the project. the Government agreed that evidence  -20-0 regarding donation or purchase of land for new construction and adequate staff would be available during the annual work plan reviews (see paa 7.0 (j)). Table 3.2: SUMMARY OF CIVIL WORKS REQUIREMENTS BY STATE AND FACILITY TYPE l l STATE | Guerrero Chiapas Oxaca Hidalgo REUIREHNI| PROG H S Now Classrooms 742 626 529 227 2,124 1250 Rehabilitatlon 1,512 2,488 2,120 418 6,538 3870 Nov Support 553 316 160 39 1,068 640 Facilities Support Facility 1,195 1,278 1,506 788 4,767 2,800 Rehabilitation School Repairs 192 258 203 521 1,174 695 Sch. 10 338 341 0 689 689 Reinforcement NOTE: This clvil works program would be carried out jointly by SEP and the Solidaridad Program (with previous authorization from SEP). The last column presents the estimated SEP program separately. 3.35 Maintenance Program. Preventive maintenance of furniture, equipment and facilities would be carried out by the SCEPs, with the participation of parents' associations and technical assistance from CAPFCE. At negotiations. the Govemment presented to the Bank the manual for preventive maintenance., repared by CAPFCE (see para 7.01 (h)). 3.36 A condition for loan effectiveness would be the presentation. for at least two of the project states, of the detailed proposals on how to implement the procedures in the preventive maintenance manual for infrastructure, furniture and equipment. specifyiog the strategy to be carried out by the SCEPs. CAPFCE the state and the communiq: the remaining two states would present the plan by June 30. 1992 (see par. 7.0Q (a)). 2. Institutional Strengthening Component 3.37 Under the Institutional Strengthening component (US$97 million, including contingencies), the project would support a set of activities aimed at improving motivation and work conditions of teachers and supervisors and at strengthening the organization of the support service delivery system and the managerial capacity both at the central and state levels. a. Provision of Work Incentives to Teachers. 3.38 The project would support the provision of incentives to general primary and indigenous teachers working in marginal and extremely isolated rural areas. The objective is: (i) to reduce the very high turnover and absenteeism rates of teachers in those areas; (ii) to facilitate  -21 - a more efficient teacher distribution within each state; and (iii) to improve access to educational services for the poor living in isolated areas. 3.39 The incentive would be equal to the base salary of a primary school teacher, and would be strictly attached to the work place so that a teacher transfer to any other school would entail its loss. Payment would be made every month and would be clearly differentiated from the normal salaries. Close vigilance with respect to teacher punctuality and regular attendance would be carried out not only by supervisors but also by the commnunity councils and parents' associations who would authorize the payment (or request its suspension). Lack of compliance would entail the loss of the incentive and subsequent transfer to another teaching position. Even a temporary absence for sickness or pregnancy would entail the transfer of the incentive to the teacher's substitute. 3.40 These incentives, which would make total salaries in these areas signiicantly higher (by about 50%-6Wo) than in other (particularly urban) areas, are expected to be effective in inducing teachers to provide the number of hours/days of teaching that they are required, but currently failing, to provide, and to accept longer term assignments. They are also expected to facilitate negotiations with the union with respect to teacher reassignment within the state (see para. 2.19). 3.41 The work incentive program would be managed by the General Directorate for Personnel at the federal level until the decentralization process is further advanced, but the identification of schools which would benefit from it would rest on requests formulated and documented by the SCEP at the state leveL The SCEPs would also be responsible for the program execution and supervision, including payment and control of attendance (with assistance from community councils). 3.42 SEP agrees that this incentive program, which has already been successfully implemented on a pilot basis, should be subject to a close follow-up and should only be expanded through careful targeting. This targeting would be based on a map of each state established by the respective SCEPs which would identify "incentive-schoolse according to various indicators (lack of basic services, difficulties of access and distance from roads, repeated problems of absenteeism and abandonment of the work place). The number of beneficiaries from the program would gradually increase from about three percent of rural teachers in the first year of the project to around 12 percent of rural teachers by the end of the project. The General Directorate for Personnel has prepared detailed guidelines specifying selection criteria for target schools, the teacher attendance control mechanism and the payment system to be used, in accordance with agreements reached during appraisal.2 At negotiations. the Government presented to the Bank. and agreement was reached on. the list of target schools and localities selected for the first project year. School selection for each subseguent proiect year would be discussed with the Bank and agreed at the annual project implementation reviews (see para 7.01 (f)). 2 SEP, "Guidelines for the Payment of Work Incentives to Primary School Teachers Working in Marginal Rural Areas (Lineamientos para el pago del incentivo a maestros de educaci6n primaria por laborar en localidades rurales de marginaci6n extrema,)" July 1991 (see Project Files).  - 22 - b. Improvement of the Supervision System. 3.43 The Mexican Government has given high priority to the improvement of the functioning of the supervision system which is currently seriously impaired by strong resource constraints, especially in remote rural areas (see para 220). Research has shown that the effectiveness of the school supervision system is positively related to teacher attendance, successful implementation of school improvement efforts, and student achievement. 3.44 lo improve the supervision system so as to allow supervisors to carry out effective technical support, the proposed project would finance transportation, pecuniary incentives, offices and equipment for the reproduction of materials and incremental salaries for about 100 supervisors. The transportation provision includes the payment of rental services and the financing of about 280 four wheel drive vehicles identified as the minimum necessary for the approximately 1,200 federal and state supervisors working in rural and indigenous areas in the four states. These vehicles would be used to allow supervisors to reach isolated areas with no alternative way of access and to facilitate distribution of educational materials to the schools. Pecuniary incentives varying between around US$600 and US$1,000 per year to compensate for additional travel and subsistence costs, would be given to supervisors who have to visit distant schools, provided they comply with their expected work plan. As in the case of teachers incentives, compliance would be controlled by community councils (comisarios ejidales) and parents' associations as well as by the school director. In addition, about 1,200 supervisor offices would be built and equipped under the project. The provision of modest office space, annexes usually attached to existing schools and strategically located in the supervision zone, is designed not only to facilitate supervision work but also to allow, at the level of school zone, storage of textbooks and didactic materials, the distribution of which would be placed under the supervisors' responsibility (see para. 347). These work incentives would complement the fairly intensive training program geared towards improving supervisor skills (see para. 3.14). 3.45 The combination of work incentives, in-service training and improved work conditions is expected to increase significantly the frequency of supervision, especially in remote areas, to improve the quality of pedagogical and technical support and to allow a more timely distribution of didactic materials. As part of an effort to increase motivation and performance, the program would also include symbolic rewards to outstanding staff, related to student achievement. 3.46 The General Directorate for Personnel has prepared guidelines to specify the norms for administering and distributing pecuniary incentives to supervisors, in accordance with agreements reached during appraisal.A At negotiations. SEP provided evidence that all sites for construction of office space grogrammed for the first year of the project have been donated to or acquired by SEP or by state governments and have the adequate staff. For the subsequent years of the project. evidence regarding donation or purchase of land for new construction and adequacy of staff would be part of the annual work plan reviews (see para. 7.01 (g)). .o/ SEP, "Guidelines for the Payment of Incentives to Primary Education School Supervisors Working in Marginal Rural Areas (Lineamientos para el otorgamiento del apoyo a la funci6n de supervisi6n escolar en la educaci6n primaria, por laborar en localidades rurales de marginaci6n extrema)," July 1991 (see Project Files).  -23 - c. Orfanization/Management: Strengthening of the Support Service Delivery and Management Capacity. 3.47 The support service delivery capacity of SEP has been seriously impaired in the past, by distribution problems which have prevented a timely distribution of textbooks and other materials to schools. The problems were caused by inadequate or unavailable storage facilities, transportation problems (including unavailability of private sector transports) and difficulties of access to rural schools, particularly acute in the project states (see param 2.16). To address this critical issue, the project would support the construction and equipping of a network of about 43 warehouses, located strategically in each state and easily accessible by 3.5 ton trucks, which would replace limited and inadequate rented facilities. Materials would be delivered to these warehouses and assembled in packages according to the needs of each school in the region. These packages would be delivered to the offices of the supervisors who would then be responsible for the distribution to individual schools, using storage space and vehicles fnanced by the project as part of the improvement of the supervision system (see para. 3.44). 3.48 Space requirements for storage have been estimated on the basis of the volume of textbooks, didactic materials, school furniture, materials for preventive maintenance, to be distributed over the year, and considering a 75 percent utilization rate of storage capacity over ten months (see Annac 6, Attachment 3 for detailed explanations). Staff requirements have been estimated at four persons per warehouse but no new hiring would be necessary as existing SCEP staff could be redeployed. The project would also support the financing of 43 (3.5 ton) vehicles to allow distribution of materials from the warehouses to the supervision zone. The reliance on a private transportation system was considered, but found less cost effective at this time for these particular states. Implementation would be phased so as to allow for testing and possible modifications and alternative approaches, if necessary. As for other civil works. SEP provided evidence at negotiations that all sites for construction of office space programmed for the first year of the project have been donated to or acauired by SEP or by state governments and have the adeauate staff: agreement was also obtained at negotiations that evidence regarding donation or purchase of land and adequate staffing would be provided during the annual work plan reviews lsee para. 7.01 (g)). As a condition for disbursement for warehouse construction. SEP would submit an inventory control system satisfactoqy to the Bank (see para 7.03 (b)). 3.49 The project would also strengthen the management capacity of SEP at both the central and state levels by financing specialized short-term technical assistance to support project coordination, monitoring, and evaluation activities of the central project coordinating unit (PCU) and state project coordinating units' (SPCUs) staff (see para. 5.10), as well as training for technical, managerial, and administrative staff, office furniture and equipment, including computers, vehicles, and operating costs. d. Improvement in the Information. Monitoring and Evaluation Systems. 3.50 Information and Monitoring System. The project would support the development of the current information system to improve SEP planning and monitoring capacity through: (a) a revision of the content of current administrative registers to include all relevant and necessary data; (b) improvements of the data collection and processing system and of the communication system between states and the central level; and (c) a more decentralized use of the information allowing for diffusion to local education authorities, including supervisors, principals, teachers and parents' associations. Funding would be provided for technical assistance, supplementary equipment at the state and federal levels and training. The information system is currently  -24- managed by the General Directorate for Planning, Programming and Budgeting (DGPPYP) at both the federal and state levels, and would therefore be readily accessible to the PCU for the continuous monitoring of project implementation. Precise terms of reference for an information needs study, as well as an action plan for the development of the system,!-Y which were prepared with financial support from the Japanese Grant Facility, have already been discussed and found acceptable to the Bank. At negotiations, the Government proided assurances that the results of the information needs study would be submitted to the Bank. for review, no later than December 31. 1991. and modifications to the action plan-(mentioned above), satisfactoLy to the Bank. made by February 29. 1992 (see para 7.01 ()). 3.51 Evaluation (or quality-assessment) Sgstem. The capacity for sustainable system-wide evaluation is essential for assessing and monitoring the level of quality of education and identifying changes over time. It is also essential for providing relevant information to those in a position to improve educational practice in the countty. The project would improve this capacity by instituting a system of assessment of educational quality and students' performance. 3.52 To assess student performance and educational quality, the project would institute in the project states a system of tests on a sample of students in all primary education grades. Samples would be designed to cover the different types of school and education modalities and a variety of geographical areas. Appropriate tests measuring cognitive achievement by subject (language, math, natural and social sciences) would be designed by the DGEIR and administered to students. Additional data relative to school inputs, family and community characteristics and social and economic environment would also be gathered to control for school and non-school factors. This system would allow for precise comparisons of student achievement by geographical area, gender, type of school and school inputs, and would become, in the long-run, an extremely useful source of periodic information on sector performance. This program would be managed by the DGEIR at the federal and state levels. The project would finance data collection and analysis, distribution of information, some equipment, and technical assistance. A detailed proposal was discussed at appraisal,3 and agreement was reached on the action plan and implementation schedule. At negotiations. the Government: (i) confirmed its agreement that questionnaires and administration manuals for the first round of field work would be sent to the Bank. for review, by June 30. 1992: and (ii) provided assurances that it would carry out the tests and other data-gathering, satisfactory to the Banl. Evaluation of subsequent program implementation progress would be part of the annual reviews (see parm 7.01 (k)). e. Selected Studies. 3.53 The proposed project would support operational research and studies which would be coordinated by the PCU at the central level in collaboration with the DGEIR and carried out by selected research institutions. The project would finance technical assistance, survey design, data collection and processing, and analysis. The studies are designed to: (i) evaluate the impact of fl/ SEP, 'Terms of Reference for the Improvement of the Information System (Terminos de referencia del proyecto Banco Mundial-M6xico - Tema: Mejorar el sistema de informaci6n)," May 1991 (see Project Files). j./ SEP, 'Proposal for a Quality Assessment System for Primary Education in the States of Chiapas, Guerrero, Hidalgo and Oaxaca (Proyecto de evaluaci6n de la calidad de la educaci6n primaria en los estados de Chiapas, Guerrero, Hidalgo y Oaxaca)," February 1991 (see Project Files).  -25 - project activities; (ii) measure the cost-effectiveness of the various components of education policies; and (iii) enhance planning and policy formulation capacity. 3.54 The principal and most extensive study woulJ be an evaluation of the impact of the project on children's learning achievement and probabilities to repeat and/or dropout from primary school. This evaluation would be based on longitudinal data gathered on two student cohorts (including a control group) and would not only measure the overall effect of the project but would also make separate analyses for each of the project components. It would thus aim at measuring the effects of: (i) increased availability of didactic materials; (ii) use of library books; (iii) teacher training; (iv) use of distance education technologies; and (v) work incentives for teachers and supervisors, controlling for the effect of non-school factors such as family and community characteristics. This impact evaluation would be complemented with a cost- effectiveness analysis based both on the impact of the project on student learning and promotion and on disaggregated cost analysis for each of the project components. The gathering of data for this study, which would be initiated during the first year of the project, would be closely linked to the setup of the evaluation system (see para. 3.51) and would be done in collaboration with the DGEIR. 3.55 Other studies include the following: (a) a specific analysis designed to evaluate the rates of utilization and acceptance of bilingual books by indigenous teachers, students and community as well as their impact on children achievement; (b) a follow-up of the impact of the use of work incentives on teacher turnover, attendance and dedication and an assessment of the efficiency of the community participation mechanism; (c) an evaluation of the impact of the TV pilot experiment; and (d) an analysis of the project impact on other behavioral and psychological traits of children. 3.56 Discussion on draft terms of reference for the studies was initiated during appraisal (see Annex 6, Autachment 4). At negotiations. the Government presented. and agreement was reached upon. terms of reference for the studies. implementation schedules, short list of consultants, the collaboration of foreign consultants. and model of consultant contracts, the latter to follow those used by the Secretary of Labor (STPS): and provided assurances that the data collected for the studies would be made available to the Bank (see para 7.01 (1)). Consultants' proposals for the studies on impact evaluation, cost-effectiveness, and work incentives for teachers will be sent to the Bank, for review, by October 31, 1991. Interim reports on all studies would be part of the project mid-term review (see para. 5.16). IV. PROJECT COST. PROCUREMENT. DISBURSEMENT, AND AUDITS A. Project Costs and Financing 4.01 Project Costs. The total project cost, including physical and price contingencies, and taxes (estimated at US$39 million equivalent), is estimated at about US$352 million equivalent, of which US$44.4 million would be foreign exchange. Table 4.1 below summarizes the estimated  - 26- estimated costs by project component. Investment costs amount to 67% of base costs and incremental recurrent costs amount to 33% of base costs (see Anne 7). Project costs are estimated at July 1991 price levels and include physical contingencies of 10% for civil works and 5% for equipment, furniture, vehicles, and educational materials. Price contingencies are calculated at 3.4% per year for 1991-96. These rates have been applied to local and foreign costs as well, assuming that periodic exchange rate adjustments would compensate for the difference between projected US dollar inflation and local inflation rates. Table 4.1: SUMMARY OF PROJECTS COSTS BY COMPONENT (Us$ MILLION) TOTAL S LOCAL IOJ IGN TOTAL S FOREIGN BASE . ~~~~~~~~~~~~~~zxcH"GE COSTS A. Educational Services Education Materials 49.7 11.1 60.8 17 20 Teacher Training 32.8 6.1 38.9 15 13 Educational Facilities 101.7 17.1 118.8 14 39 Sub-Total 184.2 34.3 218.5 15 72 B. Institutional Strengthening Teacher Work Incentives 67.4 3.3 70.7 4 23 and Supervision Organisation/Management 7.8 1.2 9.0 13 3 Monitoring, Infor. & Eval. 5.1 1.7 6.8 24 2 Sub-Total 80.3 6.2 86.5 7 28 TOTAL BASEZLIN COSTS 264.5 40.5 305.0 15 100 Physical Contingencies 12.0 2.0 14.0 17 5 Price Contiagencies _31.0 2.0 33.0 6 11 TOTAL PROJECT COSTS 307.6 44.4 352.0 14 115 Note: Figures may not add up due to rounding. 4.02 Estimated construction costs for the project area are based on unit prices derived from current contracts and costs analyses for similar standards of construction in corresponding facilities. Estimated educational facilities rehabilitation costs are calculated based on unit costs presented by CAPFCE (see Anex 6, Atachkment 2). Estimated costs of equipment, furniture and vehicles are based on current c.i.f. unit prices, adjusted to include local transportation costs. Maintenance costs are estimated at 0.5% of investment costs for civil works and at 2% for furniture and equipment. Estimates for technical assistance and studies are based on current rates for local and foreign experts. Cost estimates for training, salaries and operating costs are based on current costs or costing standards used by SEP. 4.03 Incremental Recurrent Costs. During the implementation period, the project would add an average of about US$20.4 million per year to the recurrent expenditure budget of SEP. Incremental operating costs generated by the project are estimated in 1991 prices to grow from US$8.9 million in the first year of project implementation to US$28.1 million in the last year of  -27 - project implementation, which represents only 0.4% of SEP's 1991 recurrent budget. The incidence on SEP's budget would thus remain very small and manageable. 4.04 Recurrent costs under the project, in 1991 US dollars, are estimated as follows: 1991/92 1992/93 1993/94 1994/95 1995/96 ---------------------------------- in US$ million a/- 8.9 15.0 22.0 27.9 28.1 a/ In 1991 pnces. These incremental operating costs would be generated by incentives for teachers and supervisors, salaries for new teachers and supervisors, operation and maintenance of infrastructure, equipment and vehicles, school supplies, and travel and subsistence costs for monitoring and evaluation activities. 4.05 Project Financing. The loan would be made to NAFIN, which would make the funds available to SEP for the execution of the project. The proposed loan of US$250.0 million would finance 80% of project costs, net of taxes and duties of about US$39.0 million. The loan would finance 100% of foreign exchange expenditures and 67% of local expenditures. The remaining US$102.0 million would be financed by the Federal Government Incremental recurrent costs would be financed by the Bank on a declining basis (see para. 4.14), along the lines already established under the Basic Health Project (Loan 3272-ME). The relatively high percentage of local cost financing under the project is the consequence of the project's small foreign exchange component. It is justified by the need to ensure a significant impact on social sector development and to facilitate desired policy changes. Moreover, the fact that the project targets poor states with low savings capacity, and furthermore, within these states, focuses its most significant interventions on the poorest segments of the population, also justifies this relatively high percentage of local cost financing. To ensure timely and efficient project implementation, various start-up activities, including those related to technical and bidding specifications for civil works, and employment of consultants, are now being undertaken. Retroactive financing of up to US$6.0 million equivalent would be made available to cover eligible start-up expenditures made on or after March 22, 1991, the appraisal completion date. B. Procurement 4.06 The project would include procurement of new construction and rehabilitation of educational facilities, furniture, equipment, audio visual equipment, educational materials, printing of library books, vehicles, school supplies, studies and consultant services. Standard bidding documents have been agreed upon by the SPP and the Bank in Cuernavaca in April, 1990 and would be used during project implementation. All international competitive bidding (ICB) and all first local competitive bidding (LCB) contracts per state, both for the acquisition of goods and for construction of civil works, would be subject to prior review by the Bank. All other procurement documentation would be subject to selective ex-post reviev by the Bank and would be kept by SEP/CAPFCE for periodic review during supervision. 4.07 Packaging of contracts would be encouraged. All contracts for goods that can be packaged to exceed the equivalent of US$250,000 would be procured on the basis of ICB procedures in accordance with Bank guidelines. For the purpose of comparing foreign and local bids for goods during ICB evaluation, domestic manufacturers would be allowed a margin of preference equal to 15 percent or the prevailing duty, whichever is lower.  -28- 4.08 Furniture and equipment for educational facilities amount to about US$17.0 million, about US$10.0 million of which would be procured by SEP/CAPFCE at state and central levels using ICB procedures. Packages of goods, fumiture and equipment valued at less than US$250,000, up to an aggregate value not to exceed US$6.0 million, not expected to attract foreign bidders, could be procured by the SCEPs using LCB procedures acceptable to the Bank. Procurement on the basis of comparison of price quotations solicited from at least three suppliers eligible under Bank guidelines, in accordance with procedures sa-tisfactory to the Bank, would be permitted for small, urgent and readily available off-the-shelf furniture and equipment in packages estimated to cost US$25,000 or less, up to an aggregate amount of US$1.0 million. 4.09 The total cost of 323 vehicles (including 280 four-wheel drive and 43 3.5 ton trucks) needed for the project is estimated to be US$5.0 million, all of which would be procured through ICB. 4.10 Educational materials (excluding printing of library books) and school supplies for the project amount to about US$59.5 million. About US$17.5 million of educational materials would be procured by SEP and CONAFE at the central level and by the SCEPs at the state level using ICB procedures. Packages of educational materials (excluding printing of library books) and school supplies valued at less than US$250,000, up to an aggregate value not to exceed US$36.0 million, not expected to attract foreign bidders, could be procured by the SCEPs using LCB procedures acceptable to the Bank. Procurement on the basis of comparison of price quotations from at least three suppliers eligible under Bank guidelines, in accordance with procedures satisfactory to the Bank, would be permitted for packages of readily available, off-the-shelf educational materials (excluding printing of lbrary books) and school supplies estimated to cost US$25,000 or less in the case ot 4ucational materials and US$35,000 or less in the case of school supplies, up to an aggregate amount of US$1.9 million and US$3.0 million, respectively. With regard to library books, the Bank will only finance the printing of these books using the following procurement procedures: (i) LCB procedures acceptable to the Bank would be used for contracts costing less than US$250,000 up to an aggregate amount of US$1.5 million; and (ii) comparison of price quotations obtained from at least three suppliers eligible under the Guidelines would be used for contracts costing US$35,000 or less up to an aggregate amount of US$500,000. Acquisition of certain educational materials and of certain library books, up to an aggregate amount of US$1.1 million and US$1.5 million, respectively, will be made through direct contracting and will not be financed by the Bank. 4.11 The construction of new facilities and the rehabilitation of the existing network, including repairs and structural reinforcement, would be conducted by SEP/CAPFCE through local representatives at the state level. Since all works are small and scattered throughout rural communities in the four project states and are therefore not likely to attract foreign bidders, works costing less than US$2.0 million, up to an aggregate amount of US$126.0 million would be carried out using LCB procedures, acceptable to the Bank, without excluding the possible participation of foreign firms in the bidding process. For small, simple works costing US$50,000 or less, up to an aggregate amount of US$3.0 million, shopping procedures, based on at least three price quotations from qualified contractors, would be permitted. No contract for civil works is expected to exceed US$2.0 million; however, if such contracts were to occur, ICB procedures would have to be used. About US$10.0 million, equivalent to about 7% of the total construction program, for small and scattered works with average costs of, or below US$5,000, would be implemented through force account by SCEPs/CAPFCE with no Bank financing. These works would be performed in very remote rural areas where the interest of local contractors is very unlikely.  v29- 4.12 The selection and appointment of consultants for studies and technical assistance would be consistent with the August 1981 Bank Guidelines for the Use of Consultants by World Bank Borrowers and by the World Bank as Executing Agency. 4.13 Table 4.2 showing project cost breakdown by procurement method is presented below. Table 4.2: PROCUREMENT METHODS BY CATEGORY (USS million) CATZGORY ICB . NA. TOTAL Civil Works - 126.0 3.0 a/ 10.0 b/ 139.0 (102.0) (2.0) (0.0) (104.0) Equipment and 10.0 6.0 1.0 a/ 17.0 Furniture (10.0) (5.0) (0.8) - (15.8) Vehicles 5.0 - - - 5.0 (5.0) (5.0) Educational 17.5 9.0 3.0 a/ cl - 29.5 Materials (17.5) (7.7) (1.6) (26.8) Library Books - 1.5 2.0 a/ d/ - 3.5 (1.5) (0.5) (2.0) Training, Studies and - - 45.0 - 45.0 Technical Assistance (45.0) (45.0) School Supplies - 27.0 3.0 a/ - 30.0 (13.5) (1.5) (15.0) Tniacher Incentives - - 83.0 - 83.0 Incremental Salaries (36.4) (36.4) and Other Operating Costs TOTAL 32.5 169.5 140.0 10.0 352.0 (32.5) (129.7) (87.8) (0.0) (250.0) Note: Figures in parenthesis are the amounts estimated to be financed by the Bank. a/h Works and goods to be procured through price quotatione from at least three suppliers or contractors. b/i Force account, to be used in some cases where works are small and scattered and in remote rural areas. c/s Includes US$1.1 million of direct contracting, not to be flianced by the Bank. d/s Includes US$1.5 million of direct contracting, not to be financed by the Bank. C. Disbursements 4.14 The proceeds of the loan would be disbursed during a six year period as follows: (a) civil works, 75% of expenditures for new construction and rehabilitation and maintenance of existing educational facilities; (b) equipment and furniture, 100% of foreign expenditures, and 85% of local expenditures; (c) vehicles, 100% of foreign expenditures, and 85% of local  -30- expenditures; (d) educational materials, including printing of library books, and distance education materials, 100% percent of foreign and local expenditures, excluding taxes; (e) training, technical assistance, consultancy services, and studies, 100%o of total expenditures; (f) work incentives for rural teachers, at a rate of 100% until an aggregate amount of US$10.0 million has been reached; thereafter, disbursements for this category would be made at a rate of 60%o until an aggregate amount of US$16.0 million has been reached; thereafter, disbursements would be made at a rate of 40% until an aggregate amount of US$22.0 million has been reached; and the remainder would be disbursed at a rate of 20%; and (g) incremental salaries, school supplies, and other operating and maintenance costs, at a rate of 80% until an aggregate amount of US$12.0 million has been reached; thereafter, disbursements for this category would be made at a rate of 50% until an aggregate amount of US$1&0 million has been reached; thereafter, disbursements would be made at a rate of 25% until an aggregate amount of US$22.0 million has been reached; and the remainder would be disbursed at a rate of 10o. 4.15 Proceeds of the proposed loan would be disbursed against expenditures that would be submitted to the Bank and fully documented for contracts valued at more than US$250,000 equivalent for goods and US$1.0 million for civil works. Claims for expenditures of lesser amounts would be disbursed against presentation of Statements of Expenditure (SOEs), for which the supporting documentation wouid be retained by SEP and each SCEP for periodic inspection by the Bank and by external auditors. All expenditures and SOEs would be transmitted to the Bank through NAFIN. The Government would set up a Special Account in US dollars in Banco de Mexico with an initial deposit of US$12.0 million, which is an estimate of the average disbursement per quarter of the loan over the project cycle. Withdrawals from the Special Account would be supported by the required documentation. Separate project accounts would be established and monitored by PCUs in SEP and each of the four states. D. Accounts and Audits 4.16 SEP would maintain records and separate accounts adequate to reflect, in accordance with sound accounting practices, the operations, resources and expenditures in connection with the execution of the project, including the expenditures with respect to which withdrawals from the loan account were made on the basis of statements of expenditures. Central records, maintained by the central PCU, would reflect the resources and expenditures in connection with the execution of the entire project. Other sets of records would be maintained through the respective SPCUs, under supervision by the central PCU and would reflect the resources and expenditures in connection with the execution of the project in or for the respective project state. Data from these state centers would be sent to the central PCU for consolidation purposes. NAFIN would keep records in separate accounts with respect to the Special Account. At negotiations. the Government and NAFIN provided assurances that they would: (i) have the records and accounts for each fiscal year audited by independent and qualified auditors. in accordance with generally accepted auditing standards and procedures and the specific current understanding with the Mexican authorities: (ii! furnish to the Bank as soon as available, but in any case not later than six months after the end of each year. a certified copy of the audit reports: and (iii! furnish to the Bank such other information concerning the records and accounts as well as the audit as the Bank would from time to time reasonably request (see para. 7.01 (m).  -31 - V. PROJECT IMPLEMENTA7IONAND A. Project Readiness for Implementation 5.01 The project was jointly prepared by SEP at the central level and the SCEPs, with periodic inputs from Bank missions. The preparation team was placed under the General Director for Planning, Programming and Budgeting (DGPPYP) and the Undersecretariat for Educational Coordination (UEC). The DGPPYP team, staffed with experienced personnel and with strong support from the Secretary of Education: (i) defined the project scope and objectives in accordance with the overall strategy defined in the Government 'Education Modernization Program"; (ii) coordinated the work performed by the other general directorates and entities involved in the project (DGEP, DGEI, CONAFE, CONACEN, DGEIR, UPE, UTE); and (iii) served as liaison between the central level entities and the state delegations for the gathering of relevant statistical information and the discussion of all relevant issues, mainly personnel and implementation issues. This organizational setup would serve as a basis for the implementation arrangements. 5.02 Project preparation is in an advanced state of readiness for implementation. The survey of school needs is underway to help define the precise contents of the educational materials packages and other school needs. The library program is already being implemented. Prototype designs for new facilities have been reviewed and accepted by the Bank and technical specifications for the first year rehabilitation program are under preparation. Draft guidelines for the distribution of incentives to teachers and supervisors have been prepared. Technical assistance has already been provided under grant financing by the Government of Japan for the definition of terms of reference for the development of the information and quality assessment systems. Technical groups responsible for the implementation of the teacher training programs, including distance education programs, have been formed. B. Proiect Organization and Management 5.03 SEP would be responsible for overall coordination and attainment of project objectives. Implementation responsibilities would be shared by SEP at the central level and its state delegations (SCEPs). As part of the policy aimed at strengthening the institutional capacity at the state level, it is expected that the share of responsibilities held by the SCEPs, in terms of management and monitoring of the various components, would gradually increase over the project life. The SCEPs would also '- responsible for supervising and monitoring the support provided under the project to state ss: --is. At negotiations, the Government agreed to prepare and submit to the Bank an addendum. bv Februar 29, 1992. to the existing agreement between the federal and state educational authorities to ensure that the state authorities would collaborate with SEP's state-level entities (SCEPs) to supervise and monitor project implementation (see para 7.01 (D)). 5.04 No change in implementation arrangements is expected to occur if the decentralization process is pursued further over the life of the project. The project would help prepare and facilitate the decentralization process in two ways. First, the implementation and monitoring capacity at the state level would be strengthened; second, by providing support to both federal and state schools, better coordination between state and federal authorities would be achieved as a preliminary step towards a merging of both systems. In accordance with the Mexican Constitution, however, even in the event of a totally decentralized school system, SEP  -32 - would still keep its normative and monitoring functions as well as the prerogative to provide direct technical and financial support to state schools. 5.05 Overall authority for project management would be vested in the Undersecretariat for Educational Coordination. The Undersecretary would be assisted in the general management of the project by a project coordinating unit (PCU) directly associated with the General Directorate for Planning, Programming and Budgeting (DGPPYP) (Annex 6, Chart 1 provides a detailed description of this organizational structure). 5.06 The PCU would report to the Undersecretary and to the General Director for Planning, Programming and Budgeting and would be responsible for the folowing functions: (i) overall coordination of project activities; (ii) consolidation of the work plans prepared by the executing agencies for the annual and mid-term implementation reviews (see paras 5.14 and 5.16); (iii) managing project resources, monitoring expenditures and arranging for audit accounts; (iv) contracting of technical assistance and studies; (v) liaising with project coordinating units established at the SCEP level; and (vi) continuous monitoring and evaluation of the project. 5.07 The PCU would coordinate, monitor, and evaluate the activities of the following participating entities: (i) lhe General Directorate for Primary Education (DGEP), the General Directorate for Indigenous Education (DGEI), CONAFE, CONACEN and UPE for (a) the production and/or acquisition of educational materials and their distribution to the project states; and (b) the design, planning and implementation of teacher training programs; (ii) The Educational Television Unit of SEP (UTE) for the design, planning and implementation of all distance education programs; (iii) CAPFCE for the improvement of educational facilities; (iv) The General Directorate for Personnel for the provision of work incentives to teachers and supervisors; and (v) The General Directorate for Evaluation (DGEIR) and the DGPPYP for the improvement of the monitoring and evaluation systems and the studies. 5.08 These entities, which operate both at the central and state levels, would appoint representatives to form a technical committee which would serve as liaison with the PCU. 5.09 In each of the states, a small PCU (SPCU) would also be established and its head would directly report to the director of each SCEP. It would be responsible for coordinating activities of all participating agencies and monitoring project implementation at the state level with special emphasis on: (i) the implementation of the civil works program; (ii) the timely acquisition and distrbution of educational materials; (iii) the proper functioning of the supervision system; (iv) the monitoring of the teachers' incentive program; and (v) the coordination and supervision of support provided to both federal and state schools. 5.10 The central PCU and the technical committee referred to above would be formed by experienced staff who have been involved in the project preparation since the earliest stage. The  - 33 - central PCU would be staffed on a fMl-time basis by an operating director, a financial controller a procurement specialist, an information specialist, an administrative coordinator and four program directors, each one responsible for the follow-up of specific project components (see Annex 6, Chart 1). A similar structure would be repeated at the level of the SPCUs. At negotiations. the Government: (i) confirmed agreement reached dung appraisal on the staffing plan for the project 'zoordinating units at both the central (PCU) and state levels (SPCUs): (ii! provided evider-e that the PCU operating director, and several key staff to the PCU and SPCUs. had been appointed: (iii) agreed to finalize appointment. bv the date of Loan effectiveness, of all operating directors of the SPCUs. all other staff, and appoint staff for the technical committee associated with the PCU: (iv) agreed that all aforementioned appointments should clearly specifv responsibilities: and (v) provided assurances that the PCU and SPCUs would be maintained during the project implementation with the characteristics agreed upon with the Bank (see para, 7.01 (o)). As a condition for loan effectiveness. SEP would provide evidence that the central and state PCUs are adequately staffed in accordance with the staffing plan agreed upon during negotiations (see para 7.02 (b)). 5.11 Because of SEP's lack of experience with Bank projects, a project launch seminar would be conducted prior to loan effectiveness to increase the familiarity of PCU and SPCU staff with Bank procedures. 5.12 Implementation Schedule. The proposed project would be implemented over a period of approximately five years and is expected to be completed by June 30, 1996, with the closing date set for December 31, 1996. The cost estimates prepared for the various components (see Annex 7) and implementation schedule (see Annex 9) would serve as a basis for project implementation planning and would be updated annually by SEP and the state SCEPs in preparation for the annual progress and work plan reviews (see para 5.14). During the annual reviews with the Government, the Bank would give particular attention to the effectiveness and efficiency of implementation of project components and changes would be agreed as needed. In addition, an in-depth mid-term review (see para 5.16) would provide an opportunity for a preliminary assessment of the project impact and performance and for the definition of necessary adjustments. 5.13 Monitoring Targets. To assist in the monitoring of project implementation, particularly in the context of the annual and mid-term reviews (see paras 5.14 and 5.16), a comprehensive and well-focused list of monitoring indicators would be used (see Annexes 9 and 10 for a list of such indicators). At negotiations. the Government submitted a letter, dated August 13. 1991. containing performance indicators and Droiect progress review criteria. satisfactory to the Bank. regarding Project monitoring and evaluation to be used for annual and mid-term reviews (see para 7.01 (j)L 5.14 Annual and Mid-Term Project Implementation Reviews. At negotiations, the Government provided assurances that SEP and its state delegations (SCEPs) would conduct. iointly with the Bank. SHCP. SPP and NAFIN annual project implementation reviews (see para. 7.01 ()). The review would: (a) analyze the previous years performance and budget allocation; (b) analyze the proposed work plan and budget for the following year; (c) discuss adjustments to be made in project's implementation schedule; and (d) assess changes in budgetary allocations at the state leveL The proposed work plan and budget for the following year would have to be previously defined at the central and state levels and contain a detailed description of targets and planned activities. Plans for civil works would specify type and schedule of construction and provide evidence of ownership or donation of land and of adequate staffing.  - 34 - 5.15 At negotiations. the Government provided assurances that the PCU director and the SPCUs' directors would. by September 30 of each year. beginning in 1992. prepare and submit to the Bank annual progress reports on the implementation of the project in accordance with targets established in the previous annual work and imnestment plans (see par. 7.01 (r)). 5.16 At negotiations. the Government provided assurances that: (i) an in-deDth mid-term review would be conducted by SEP. jointly with the Bank and with the assistance of consultants employed under terms of reference satisfactory to the Ba,nk. to assess project performance and initial impact. and focusing on the effectiveness of teacher and supervisor incentives, the library and bilingual books program. the distbution sb tem. teacher trainin& the Rilot TV program. and interim results of the studies mentionedin pai 35-55: (ii) results would be submitted to the Bank by September 30. 1994 for its reaiew and comment: and (iii) on the basis of this review. action plans. satisfactory to the Bank. would be grepared and initiated by November 30. 1994 (see para. 7.01 (ID. This review would assess project performance and initial impact in the light of project objectives and SEP policies. The PCU would be responsible for hiring consultants and preparing the terms of reference, which would be submitted to the Bank for comments, and for preparing the report with recommendations. VI. PROJECT BENEFIT AND RISKS A. Project Benefits 6.01 The proposed project would support the Government programs for improving both the quality and efficiency of primary education, focusing on four states with high incidence of poverty and low education indicators. The level of cognitive achievement of children as well as their probability to complete the primary education cycle are expected to increase as a result of the project. Benefits would thus consist in a higher rate of human capital accumulation contributing to economic development. They would also consist in a more equal distribution of economic opportunities among Mexican children as the large proportion of poor children living in the project states would have greater chances to increase their level of educational attainment and their earnings potentials. In addition, the project would improve the efficiency of public resource use by reducing repetition and dropout rates and by increasing the amount of investments with high social rates of return. Institutional strengthening would contribute to a better allocation and increased productivity of human resources within the sector; and to an enhanced capacity to evaluate and monitor educational policies and programs. B. Impact on Women 6.02 Approximately half of the children who would benefit from an improved quality of educational services in the project area would be girls. Their opportunities to have access to higher levels of education and better work opportunities would thus improve. Further benefits can be expected as research has shown that increasing the level of education of women has a significant positive effect on the health status and on the educational attainment of their children and contnbutes to lower fertility. The impact may be particularly significant for girls from the indigenous population.  -35 - C. Project Risks 6.03 The main project risks are associated with possible delays in project execution due to: (i) possible teachers' union opposition to Government policy changes; and (ii) limited implementation capacity at the state leveL The risk of teachers' union opposition is difficult to address. However, the improvement of work conditions for teachers and supervisors that the project would bring about and the recent success of the Government (and in particular, of the SCEP directors of the states involved) in hardling and solving union and employment issues provides some confidence that this risk can be minimized. The project attempts to reduce the second risk through: (a) the provision of training and technical assistance to the states; and (b) close monitoring from the central level. To increase SEP staff familiarity with Bank procedures, a project launch seminar to cover implementation aspects would also take place soon after loan effectiveness. 6.04 The high priority given by the Government towards improving economic and social conditions in the poor southern project states on the one hand and, on the other hand, the strong commitment expressed in repeated occasions by SEP towards the project objectives as well as the dedication and work already invested in the project preparation, also contribute to reduce the risk of slow implementation.  - 36 - VII. AGREEMENTS REACHED AND RECOMMENDATIONS 7.01 During negotiations the Government: (a) provided assurances that the results of the survey on the needs for didactic materials for each school currently underway would be furnished to the Bank for review no later than November 30, 1991 and would be used to adjust the precise content of the didactic materials packages (see par. 3.10); (b) submitted a letter, dated August 8, 1991, confirming agreements reached at appraisal regarding the mechanisms to be used to monitor teacher attendance to training sessions and distribute the incentives (per diems and promotion points) (see para. 3.19); (c) provided assurances that a detailed description of the teacher training courses (consisting of course objectives, subject matter, time allocation and instructional techniques) and training materials would be submitted to the Bank, for review, by June 30, 1992 (see para. 3.22); (d) provided assurances that the detailed descriptions of all distance education programs, as well as their transmission schedule in the case of programs to be broadcast, would be presented to the Bank, for approval, by April 30, 1992; and submitted evidence, satisfactory to the Bank, that Radio-Television de Oaxaca has the necessary human and physical resources for the development and application of pilot educational programs (see para. 328); (e) agreed that: (i) a list of civil works carried out by SEP and under the Solidaridad Program would be provided to the Bank during the annual work plan reviews; (ii) on this basis, classroom requirements and distribution of civil works between new construction and rehabilitation done under this project would be readjusted; and (iii) all civil works would be selected in accordance with criteria and priorities stated in para. 3.30 (see para. 331); (f) submitted evidence that technical plans for at least one third of planned school rehabilitation works for the first year of the project were previously prepared and agreed to submit the remaining two thirds during the rest of the year; and agreed that for subsequent years, technical plans of the following year rehabilitation work program would be submitted during the annual reviews; (see para. 3.34); (g) provided evidence that all sites for new construction programmed for the first year of the project have been donated to or acquired by SEP or by state governments and have the adequate staff; agreed that evidence regarding donation or purchase of land for new construction and adequate staffing would be provided during the annual work plan reviews (see paras. 3.34, 3.46 and 3.48); (h) presented to the Bank the manual for preventive maintenance, prepared by CAPFCE (see para. 3.35);  - 37 - (i) presented to the Bank, and agreement was reached on, the list of target schools and localities selected for the first project year; school selection for each subsequent project year would be discussed with the Bank and agreed at the annual project implementation reviews; (see para. 3.42); (j) provided assurances that the results of the information needs study would be submitted to the Bank, for review, no later than December 31, 1991, and modifications to the action plan (mentioned above), satisfactory to the Bank, made by February 29, 1992 (see para. 3.50); (k) confirmed its agreement that questionnaires and administration manuals for the first round of field work would be sent to the Bank, for review, by June 30, 1992; and provided assurances that it would carry out the tests and other data-gathering, satisfactory to the Bank; evaluation of subsequent program implementation progress would be part of the annual reviews (see para. 352); (I) presented, and agreement was reached upon, terms of reference for the studies, implementation schedules, short list of Mexican consultants, the collaboration of foreign consultants, and model of consultant contracts, the latter to follow those used by the Secretary of Labor (STPS); and provided assurances that the data collected for the studies would be made available to the Bank (see paras. 355-56); (im) provided assurances that NAFIN and the Guarantor would: (i) have the records and accounts for each fiscal year audited by independent and qualified auditors, in accordance with generally accepted auditing standards and procedures, and with the specific current understanding with the Mexican authorities; (ii) furnish to the Bank as soon as available, but in any case not later than six months after the end of each year, a certified copy of the audit reports; and (iii) furnish to the Bank such other information concerning the records and accounts as well as the audit as the Bank would from time to time reasonably request (see para. 4.16); (n) agreed to prepare and submit to the Bank an addendum, by February 29, 1992, to the existing agreement between the federal and state educational authorities to ensure that the state authorities would collaborate with SEP's state-level entities (SCEPs) to supervise and monitor project implementation (see para. 5.03); (o) (i) confirmed agreement reached during appraisal on the staffing plan for the project coordinating units at both the central level (PCU) and the state level (SPCUs); (ii) provided evidence that the PCU operating director, and several key staff to the PCU and SPCUs, had been appointed; (iii) agreed to finalize appointment, by the date of Loan effectiveness, of all operating directors of the SPCUs, all other staff, and appoint staff for the technical committee associated with the PCU; (iv) agreed that all aforementioned appointments should clearly specify responsibilities; and (v) provided assurances that the PCU and SPCUs would be maintained during the project implementation with the characteristics agreed upon with the Bank (see para. 5.10); (p) submitted a letter, dated August 13, 1991, containing performance indicators and project progress review criteria, satisfactory to the Bank, regarding Project  - 38 - monitoring and evaluation to be used for annual and mid-term reviews (see para. 5.13); (q) provided assurances that SEP and its state delegations (SCEPs) would conduct, jointly with the Bank, SHCP, SPP and NAFIN annual project implementation reviews (see para. 5.14); (r) provided assurances that the PCU director and the SPCUs' directors would, by September 30 of each year, beginning in 1992, prepare annual progress reports on the implementation of the project in accordance with targets established in the previous annual work and investment plans (see pam. 5.15); and (s) provided assurances that: (i) an in-depth mid-term review would be conducted by SEP, jointly with the Bank and with the assistance of consultants employed under terms of reference satisfactory to the Bank, to assess project performance and initial impact, and focusing on the effectiveness of teacher and supervisor incentives, the library and bilingual books program, the distribution system, teacher training, the pilot TV program, and interim results of the studies mentioned in paras 3.53-55; (ii) results would be submitted to the Bank by September 30, 1994 for its review and comment; and (iii) on the basis of this review, action plans, satisfactory to the Bank, would be prepared and initiated by November 30, 1994 (see para. 5.16). 7.02 As conditions for loan effectiveness, the Government would: (a) present, for at least two of the project states, the detailed proposals on how to implement the procedures in the preventive maintenance manual for infrastructure, furniture and equipment, specifying the strategy to be carried out by the SCEPs, CAPFCE, the state and the community, the remaining two states would present the plan by June 30, 1992 (see para. 336); and (b) provide evidence that the central and state PCUs are adequately staffed in accordance with the staffing plan agreed upon during negotiations (see par. 5.10). 7.03 Conditions of disbursement will be: (a) for the pilot programs, presentation of a detailed proposal for the work to be done by Radio-Television de Oaxaca, specifying the implementation schedule (see para. 328); and (b) for warehouse construction, submission by SEP of an inventory control system satisfactory to the Bank (see para. 3.48). 7.04 Subject to the above conditions, the proposed project would constitute a suitable basis for a Bank loan of US$250.0 million equivalent to the United Mexican States, to be repaid over a period of 17 years, including 5 years of grace.  -39. MEXICO - PRIMARY EDUCATION PROJECT ANNEX LIST ANNEX 1: Comparative Education Data in Selected Countries Table 1: Comparative Educational Data ANNEX 2: Mexico: Educational Statistics Table 1: Primary Education Statistics by National and State Levels ANNEX 3: Mexico: Efficiency Indicators Table 1: Efficiency Indicators for Primary Education by State Table 2: Estimated Repetition Rates in Primary Education by Project State and Inefficiency Costs Figure 1: Primary School Completion Rates, National Level by School Modality and by Percent ANNEX 4: Public Education Expenditures and Financing Table 1: Recent Trends in Federal, State and Private Education Expenditures Table 2: SEP Actual Recurrent and Investment Expenditures by Project State Table 3: Federal, State and Private Education Expenditures by Project State Figure 1: Federal Recurrent and Investment Education Expenditures ANNEX 5: Sector Organization Current State of the Decentralization Process Box 1: Mexico's Decentralization Process Chart 1: Organizational Structure of the Secretariat of Public Education (SEP) Attachment 1 Teacher Appointment and Reassignment Policies and Practicts ANNEX 6: Project Description Chart 1: Organizational Structure of the PCU at the Central Level Attachment 1 Description of Teacher Training Program Matrix 1: Matrix of Training Modalities and Groups of Teachers to be Trained and Corresponding Unit Costs Attachment 2 Table 1: Civil Works Program for Improvement of Educational Facilities Attachment 3 Distribution System: Space Requirements for Storage Table 1: Warehouse Space Requirements Attachment 4 Terms of Reference for Studies ANNEX 7: p:nce Costs Table 1: Project Cost Summary Table 2: Project Components by Year Table 3: Projects Components by Year (including contingencies) Table 4: Summary Accounts Cost Summary Table 5: Summary Accounts by Year (cont'd)  -40- MEXICO - PRIMARY EDUCATION PROJECT ANNEX LIST (cont'd) Table 6: Summary Accounts by Year (including contingencies) Table 7: Summary Account by Project Component ANNEX 8: Disbursement Table 1: Estimated Schedule of Disbursements ANNEX 9: Implementation Schedule ANNEX 10: Key Monitoring Targets Table A: Quantitative Indicators Table B: Qualitative Indicators ANNEX 11: Key Documents Available in the Project File  MEXICO PRIMARY EDUCATION PROJECT Table 1: COMPARATIVE EDUCATIONAL DATA Tablo A: PRIMARY GROSS ENROLLMENT RATIOS Table D: AVERAGE PRIMARY TEACHER SALARY Toal and Femal (in parenhse) Raios and As a Percentage By Year, In 1985 Constant US Dollars, and As a Mulbple of GNP Per Capita ..~~~~~ . i ~~~~1965 1970 1975 1980 1985 1YEAR SALARY GNP MULTIPLEI CHILE 119.0 104.8 112.2 108.8 105.7 CHLE 1981 2845.0 2.0 (117.7) (104.7) (111.5) (107.7) (104.3) COSTA RICA 107.6 109.7 107.0 104.8 97.4 COSTA RICA 1986 4760.3 3.2 (106.1) (109.1) (106.5) (103.8) (96.2) NICARAGUA 67.1 80.0 82.0 99.0 101.5 NICARAGUA 1986 3523.5 4.5 (67.2) (81.0) (84.5) (101.8) (107.0) HONDURAS 77.4 88.2 87.6 92.8 101.9 HONDURAS 1982 3307.3 4.2 (77.2) (88.6) (87.0) (93.6) (103.0) GUATEMALA 50.5 58.3 63.2 70.8 76.5 GUATEMALA 1979 2489.6 1.7 (45.6) (52.4) (57.5) (64.9) (70.2) BRAZIL 107.2 143.1 87.8 98.9 100.9 BRAZIL - - - (106.9) (143.7) (87.0) (96.6) (96.5) Table C: PUBLIC RECURRENT EXPENDITURE PER PRIMARY STUDENT As a Parcentge of GNP Per Capita and In 1985 Constant US Dollua l~~~~~~~A a Pffeitgo of GNP For Capiita 1985 cotm US Dolus 0 1 196~~iS 197 1975 198 198S i 96 1970 1975 198 1985 CHILE 63 - 6.1 9.S 13.2 72.4 - 64.2 134.9 155.2 CSTA RICA 13.6 12.5 13.0 12.7 10.4 137.9 154.8 18S.4 204.2 149.4 NICARAGUA 9.3 9.6 7.5 7.4 14.9. 114.9 122.0 10S.7 73.7 119.3 HONDURAS 15.6 13.1 - 11.1 12.2 104.0 94.3 - 94.5 91.S GUATEMALA - 10.0 7.8 a/ Sb - - 110.7 103.2 a/ 74.0 bl - BRAZIL - - 7.6 af 8.5 8.3 - - 110.1 a/ 142.1 133.2 Source: M. Lockbeed & A. Verspoor, Improving Primary Education in Developing Countries, World Bank, Tables 4; 20; and 22, 1990. Notes: a/ 1976 figures; bl 1979 figures.  - 42 ANNEX 2 Papg 1 of 3 MEXICO PRIMARY EDUCATION PROJECT Table 1: PRIMARY EDUCATION STATISTICS By National ad State Lovls (1989-1990) AREA LEVEL OF STUDENTS TEACHERS SCHOOLS EDUCATION Rural TOTAL Rural TOTAL Rual| TOTAL Federal 4,353,781 10,389,248 153,902 341,252 46,815 64,246 General 3,716,618 9,752,085 126,666 314,016 34,183 51,614 NATIONAL Indigenous 564,698 564,698 21,193 21,193 6,589 6,589 Community 72,465 72,465 6,043 6,043 6,043 6,043 Couris State 757,390 3,239,381 26,965 99,732 6,797 13,015 Private 28,721 865,134 879 25,548 189 3,375 LEVEL OP STUDENTS TEACHERS SCHOOLS -_ El)UCATION Rora TarAL Ruml | TOAL Rwrl TaTAL Federal 1,245,731 1,925,624 43,197 64,247 12,249 14,273 General 869,268 1,549,161 28,692 49,742 7,219 9,243 PROJECr Indigenous 360,993 360,993 13,353 13,353 3,878 3,878 AREA Community 15,470 15,470 1,152 1,152 1,152 1,152 Courses State 94,687 210,09S 2,946 6,641 818 1,204 Private 2,987 4S,662 105 1,438 29 223 (coat)  - 43 - ANNEX 2 Page 2 of 3 MEMCO PRMMARY EDUCATON RELCT (CsL) | STATE LEVEL OF STUDENTS T_CH SCHOOLS l EDUCATION Rum TOTAL Rwg TOTAL _ Rud| TOTAL| Federal 366,758 S14,412 12.017 16,461 3,931 4,401 General 233,465 381,119 7,241 11,685 1,919 2,389 CHIAPAS Indigenous 125,463 125,463 4,210 4,210 1,446 1,446 Community 7,830 7,830 566 566 566 566 Courses State 35,715 99,087 1,046 3,148 305 541 Private 1,002 11,711 33 395 7 68 STATE ~LEVEL OF STUDENTS TEACHER SCHOOLS EDUCATION Rwd TOTAL I Rum TOTALI Rw| TOTAL Federal ~~248,672 447,017 9,817 16,606 2,406 3,016 General 185,397 383,742 7,020 13,809 1,669 2,279 GUERRERO Indigenous 61,767 61,767 2,680 2,680 620 620 Community 1,508 1,508 117 117 117 117 Courses State 58,972 111,008 1,900 3,493 513 663 Private 231 11,296 6 345 1 46 (coaL)  - 44- ANNEX 2 Page 3 of 3 M1EXCO PRIMARY EDUCATION PROJECT (COQL) STATE LEVEL OF STUDENTS TEACHERS SCHOOLS EDUCATION Rurd TOTAL Rural TOTAL Rural TOTAL Federal 411,918 611,368 13,631 19,597 3,740 4,355 Generl 278,603 478,053 8,707 14,673 2,176 2,791 OAXACA Indigenous 129,273 129,273 4,622 4,622 1,262 1,262 Community 4,042 4,042 302 302 302 302 Courses state 0 0 0 0 0 0 Private 1,754 9,997 66 298 21 54 STATE ~LEVnEL OF STUJDENTS TEACHEIERS SCHOOLS EDUCATION | RAwml TOTAUL Rural TOTAL Ruxol | TOTAUL Federal 218,383 352,827 7,732 11,583 2,172 2,501 Genord 171,803 306,247 5,724 9,S7S 1,455 1,784 HDALO I bidigenous 44,490 44,490 1,841 1,841 550 550 Cnommunity 2,090 2,090 167 167 167 167 Cours Stat. 0 0 0 0 0 0 _Pivate 0 12,658 0 400 0 55 Source: DOPPYP, SEP, AprU 1990.  - 45 - Page I of 3 PRIMARY EDUCATION PROJECT Table 1: EFFICIENCY INDICATORS FOR PRIMARY EDUCATION By State (1988-89) | ~~STATE COMPLEIMON DROPOUr PREPETMlON | INDEIX l ~~~~ ~ ~~~~RATE RATE RA TE1 Distrito Federal 80.3 1.8 6.5 49.2 Morelos 72.3 1.7 7.0 43.4 Baja Californa Sur 86.8 2.0 8.4 41.2 Nuevo Leon 72.3 3.3 5.5 38.2 Baja Califonia 73.0 1.9 8.S 37.6 Tlaxcala 71.3 2.4 7.4 37.0 Caohuila 66.9 2.6 6.9 34.8 Mexico 68.3 2.3 9.0 30.6 Aguascalientes 66.9 3.3 7.8 27.4 Sonora 62.8 4.2 8.2 18.0 Queretaro 65.6 2.7 11.7 17.7 Tamaulipas 65.3 4.6 8.5 16.0 Nayant 58.1 4.9 7.3 13.7 Jalisco 56.4 5.6 6.6 10.5 Colima 56.2 5.1 7.9 9.3 Chihuahua 57.2 4.6 9.2 8.8 Sinaloa 54.3 5.3 8.4 5.3 Guanajuato 56.5 5.5 9.4 2.1 Hidalgo 54.7 4.0 13.1 -1.7 Durango 54.7 5.7 10.1 -2.6 San Luis Potosi 54.0 4.8 12.5 -5.2 Quintana Roo 54.0 4.1 13.9 -5.3 Zacatecas 53.5 5.3 11.7 -6.0 Tabasco 50.1 5.2 12.4 -9.7 Puebla 53.1 5.7 12.2 -10.4 Veracrz 43.6 7.8 11.6 -27.4 Michoacan 44.0 6.8 13.6 -27.4 Yucatan 43.2 7.3 14.0 -32.3 Campeche 43.1 8.3 12.5 -33.7 Guerrero 41.1 9.2 12.7 -41.3 Oaxaca 40.7 7.5 17.0 -44.9 Chiapas 27.1 15.7 14.7 -97.1 Source: Indicadores Educativos, DGPPYP, 1989. Note: The index is based on an average of the percentge deviation from the national average for the three indicators.  - 46 - Page 2 of 3 PRIAMAYEDCTOPRJT Table 2: ESIMAATD REPE ON RATES IN PRIMARY EDUCATION By Project State and Inefficiency Costs (198B89) PROJEC STATE MEXICO COUNTRY Chiapas Guerero Hidalgo Oaxaca TOTAL GRADES REPETITON RATES (%) La 1 53.0 43.2 28.5 38.3 31.8 2 40.8 33.4 21.2 28.7 17.4 3 36.7 29.4 20.1 28.4 14.0 4 34.4 284 20.9 30.0 21.2 5 35.6 29.6 23.6 32.1 15.8 6 34.3 32.7 25.7 37.3 (na.) Total # of Repeaters 260,092 196,949 86,069 201,942 2,745,300 1988 Unit Costs (in Pesos) k 245,026 292,554 247,743 228,095 259,364 Source: E Schiefelbein, Output of the SMMG model with data from DGPPP. UNESCO- OREALC, SIRI System, Santiago, June 1991. La These repetition rates were estimated by E. Schiefelbein from detailed data on total enrollment (public and private) by grade and disaggregated according to student age, following the methodology used in E. SchiefJbein, 'Repetition: The Key Issue in Latin American Primary Education," LATHR, Worling Paper No. 1, June 1989. They indicate that repetition rates reported by school principals are underestimated (cf. Table 1). k The 1988 unit costs (in Mexican pesos) were used to estimate the total inefficiency costs. The exchange rate used was M$2,273.1 = US$1.  - 47 - ANA3 Page 3 of 3 MEXICO P-RIMARY EDUCATIONPRJC Figure 1: PRIMARY SCHOOL COMPLETION RATES COHORT 1981-1986 National Level By School Modality and By Percent /0100 120,. n 1981 1982 1983 1994 1995 1996 -. 1 Gr. 2 -. 3 -. 4 It. 5 G-. 6 | _ CONAFE _ il E3 ngual I General | Source~ Oreccion de Anolgim y Sictt XO IrWoanelofl. OGPR, SEP.  MEXICO PRIMARY EDUCATION PROJECT Table 1: RECENT TRENDS IN FEDERAL, STATE AND PRIVATE EDUCATION EXPENDITURES (Absolute Amount and as Percentage of GDP) (1982-1990) FED)ERAL f {TOTAL YEAR SEP OTHER STATE PRIVATE MILLIONS OF MILLIONS OF AS % OF ISECRETARIATS NOM. PESOS 1982 PESOS GDP (MilL of Nominal Pesos) % % X 1982 368,608 46,819 80.6 14.3 5.1 515,327 515,327 5.3 1983 488,667 64,400 82.0 13.4 4.6 674,436 3S2,371 3.8 a.0 1984 826,712 101,510 74.5 16.6 8.9 1,245,464 395,134 4.2 00 1985 1,332,034 148,510 77.1 15.1 7.8 1,918,991 403,234 4.0 1986 2,112,674 271,339 76.3 14.3 9.4 3,125,720 362,612 3.9 1987 5,034,274 484,806 78.3 12.0 9.7 7,046,794 336,266 3.7 1988 10,120,060 956,954 80.0 11.0 9.0 13,842,634 342,775 3.5 1989 12,998,225 1,754,249 76.5 14.9 8.6 19,281,682 398,885 3.9 Notes: State expenditues include municipalities. Expenditures reflect executed budgL eP Source: DOPPYP, 1990.  MEXICO PRIMARY EDUCATION PROJECr Table 2: SEP ACTUAL RECURRENT AND INVESTMENT EXPENDITURES By Project State (Mil1ions of Current Peso) (1987-1991) PROJECr 1987 198 1989 1990 1/ 1991 1/ STATE Recurret Ivest TOTAL Recurret Invest. TOTAL Recurrnt invest. TOTAL Recumet Ives TOTAL R crrent NesL TOTAL Chiapas 110,769.1 77.4 110,846.5 I8,878.9 X76.4 189,155.3 281,71S.4 252.2 281,967.6 273.326.9 S89.3 273,916.2 373,360.1 1.351.0 374.711.1 Guarro 123,224.1 92.3 123,316.4 235,44.4 244.4 235,78.8 334,530.3 238.7 334,769.0 345,588.2 611.5 346,199.7 456,743.4 1,443.6 458,137.0 Hidalgo 80,012.4 140.2 80,152.6 IS7,78.8 334.3 1S8,117.1 218,867.9 297.9 219,165.8 231,487.4 496A 231,93.8 303.666.4 1,5.2 301,891.6 t Oaxaca 129,804.4 14S.6 129,950.0 250,937.6 427.7 251,365.3 314,511.9 479.5 314,991.4 350,607.7 806.4 351,414.1 458,627.5 1,942.6 460,570.1 <.-.m. I III I .. 1/ Budgeted Amount. N.A.: Not Available. Sources: SEP, Compendio Estadisdico del Gasto Educativo, 1990; and State Budgeo for 1990 and 1991.  MEXICO PRIMARY EDUCATION PROJECT Table 3: FEDERAL, STATE AND PRIVATE EDUCATION EXPENDITURES BY PROJECT STATE (19861989) 1986 TOTAL 197 TOTAL 1988 TOTAL 1989 TOTAL AREA (Millios (Millions (Mlios (Miions Foderal State Private of current Fedeal State Private of current Federl Stae Printe of curent Fedl Ste Prte of curet (%) (%) (%) pesow) (%) (%) (%) pesos) (SU) (%) (X) peos) (X) (U) (%) peos) Chiapas 78.5 17.8 3.7 80,632 73.2 23.4 3.5 200,402 70.2 26.5 3.3 385,093 75.5 21.3 3.2 532,435 Guerrero 91.6 6.4 2.0 79,672 92.2 5.8 1.9 188,003 92.1 6.1 1.8 365,997 92.8 S.S 1.7 50S.799 Hidalgo 90.6 5.2 4.2 S5.208 92.7 3.0 4.3 122,768 92.8 3.4 3.8 257.173 93.8 2.6 3.6 353.382 Oaxaca 95.9 2.2 1.9 82.825 98.0 0.0 2.0 181,294 98.1 0.0 1.9 356.427 98.2 0.0 1.8 492,259 Source: Compendio Estadistico del Oasto Educativo, 1990; DOPPYP, SEP, Cuadro 2.1, 1990. Note: State expenditures inchude municipalities.  - 51 - A4A Page 4 of 4 MEmCO PRIMARY EDUCATION PROJECI Figure 1: FEDERAL RECURREN4T AND INVESTMENT EDUCATION EXPENDITURES (Millions of Constant 1970 Pesos) (1970-1987) Recurrent Recurrent 6,35S 81X Investment 1,462 19X I nvestment 1, 992 t1X 1970 1976 Recurrenvt 26.058 93x RecLrrent 14,771 922 5x I nieutant 1987 1.873 7X 1982 Source: J. Preawd, barch 1IM9.  - 52 - Page 1 of 3 MEXaCO PRIMARY EDUCATION PROJiCI CURRENT SITUATION OF THE DECENTRALTATION PROCESS FUNCTION BEFORE 1978 CURRENT SITUATION Central (Secretariat of PubUe Central (SEP) Educational policy Education - SEP Financial resources mobilization Budget based on fiscal historic Budget based on historical fiscal trends. Financial resourees trends FPinanial resources allocated by central goveronot allocated by SEP to SEP's State to SEP Delegations (SCEPs) with final approval from SPP. Sources of funding Central goverment Central govermment Curriculum development Central. One curriculum for the Central. One curriculum for the entire country entire country, but complemented with regional and ethnic teaching materials for corresponding regios Teachers hiring and firing SEP with unions approval Combined procedures between center and SEP's State Delegations. Teachers union greatly Influences outcomes Teachers salaries Central governmsnt after National scale fixed by central negotiated with teachers union govermnent and negotiated with unions. AlibJ for regional and herd working conditions salary differences Pre-service and in-service special teacher tr ainig schools special teacher training schools training and upgrading (Eecuela Nomsles) mnaged by (Escuelas Normales) operated by SEP. Important state and private SEP$s State Delegations, State participation goverments and the private sector Textbooks Designed by SEP and publised and Designed by SEP and published and distributed, free of charg by distributed, free of charge, by sZP's paratatal agency CONALITSG S8P's parastate agency CONALITEG Supervision Centralized at SEP Decentralized at the State level Student certification Centralized at SEP SEP's State Delegations (SCEPs) School calendar setting SEP SEP Infonmation system Centralized at SEP Gathered by the SEP's State Delegations and integrated by SEP School building and maintenance Centralized by SEP and its State representations of building parastate agency CAPFCE parastate building agency (CAPFCE) incompliance with SEP'S State Delegations under broad regulations by SEP and CAPFCE. There is involvement of Parents- Teachers Association, specially in rural areas  - 53 - ANNEXS Page 2 of 3 MEXICO PRIMARY EDUCATION PROJECT .Box 1: MEXiCYS DECENTIRAIZAION PROCSSa In 1978, the Mexican federally spod educational systcmprisingmore than 11 million students, 320 thousand hersand r 4 t , was to.all centralized. Power was distnbuted by educational t eteL According to tis cenall managed model, the Director General of Primaiy d ', for exhle, thi in the hierarichal ':. ranking in the Secretariat of Public Educadion (SE?),was accountable for all the federafly sponsored primary education m the entir ry,hand a u dget oPfUS708 million, larger than the combined stat more inpa 'stat f the'e's .t country. The Mexico City office 'planned' het education system in the country'and directly contrd t a i t * ' . : 'The decentralization s'tra'tegyiiu undertke'nsm'shifte the p e b educationalJevelsto, geopolitical. entities-- Id" "tha'~l;zS'Sae)lg.i (SCEP, - "Servcios Cordin ado s' MOWdcai6 Pb" a)wr salshda h einn of 1978, itthe objective- of p lannn,mngn n prtn h eeal p oe .pre-schoo, pmary secdary d iacher i sport Sys'tms in'ther.rspective state. The S s .s aaet the same responsibilities for Mexico" Cit:y i diint 6enraieadeauto functions for their respective eduCational tevl..v' ,:To'ficilitate the'process, finacia-iaat ica a all the necessary managerial support were m'ade: -a able Iotthe'SCEP' and a.cntinuous.: technical monitoring' o progress .wa-set inple.by~ehgetS?ofca&(h ertr .conducting periodic short-term retreats. vihtA minatr fth eon -After.12 years of contin'u'ous upgrdn f ths.d.ns .atv reom the..d....ion ysehas girown to 25 mlinsuet,oe iUb e.hz n 4 htsn col Source: '. Prawda, "Decentralization and Edcatiodal c W a . A view from LAT,, - No., No. N 99  MCOM PRMARY EDUCATION PROJECT ORGANIZA1TIONAL STRUCTE OF THE SECRETARIAT OF PUBLIC EDUCATION (SEP' COODIATION f UNDERSECETARIAT WUDSC IAT UDSECRETARIAT UNDESECRETARIAT UNDERSECRETARIAT COIEF DECENTRALIZED F0 EDUCATIONAL 01 PRE-PRIMARY & OF SECONDARY OF BIGHER EDUC&- OF VOCATIONAL hD11IISTRATOR EDUCATION COORDINATIGE PRIHARY EDUCATSION ISDUCATION TION AND SCIENTI- EDUCATION AND l l B~~~~~~~~~~IC INTiGATON inRZ8 CS GENERGALMA GlNZRAL .N . DIRECTOMATE 0DDRECT1 DIRECTORATE DIRCTORU PLANIn , OF PRIM -SRVIAD OF m PROGRA 9 EDCATION IN8SRVICZ UD 1 AND BUDGETING CEIDt llUllD RSAID IGE NERAL GENERAL NATIONL r L LDIzCToT0 or DIRECTORATE OF PED AGOICAL -| DI1C I | EVALUNTIO L IIIDICEIIW UllIVlRSlTY |oF FLoUCAL| EDUCATION l m | 8SERVICES8 OP | I I PIJBLIC I |EDUCATION1 I  - 55 - ANNEX 5 Attachment 1 Page 1 of 2 MEXICO PRIMARY EDUCATION PROJECI TE-ACHER APPOINTMENT AND REASSIGNMIEN POLICIES AND PRACTICES 1. The number of teaching positions available for appointment or reassignment is determined by: a. the detailed annual programming which determines the number of new teaching positions to be created (creation of new schools or addition of new classrooms to existing schools); the number of incremental staff is based on a request from the SCEPs presented in February/March to DGPPYP, which revises it on the basis of resource availability and sector priorities and communicates final approval to SCEPs in June; and b. the number of vacant positions because of death, retirement, abandonment, or transfer to another state. 2. Lists of teachers desiring a transfer to another school are established between January and April. One third of the candidates for transfer are proposed by SEP; another third are proposed by the teachers' union (SNTE); and the final third are determined by pure seniority criteria. 3. Transfers are requested for changes within a given state or from one state to another. Within a state, transfer is normally requested for moves from an urban to a rural/urban school, or from a rural/urban school to an urban school; direct transfers from rural to urban schools are not allowed. No pecuniary benefit is attached to such moves; hard work and living conditions in rural areas appear to be sufficient incentives for such requests. A transfer from one state to another implies a reassignment to a rural school in the destination state, regardless of the type of school in which the teacher was working in the state of origin. For the project states, the number of teachers wishing to exit the state is typically larger than the number of teachers from the rest of the country wishing to enter these states. The reverse phenomenon is observed in all northern Mexican states. 4. In each state, teacher reassignment is decided, on the basis of existing vacancies and the proposed list of candidates, by a SEP-SNTE committee (Comision Mixta) that meets at the level of the SCEP. As requested moves are normally from rural to urban areas, available urban positions are all filled while rural positions remain vacant after the reassignment process is over. The difficulty of filling rural positions with experienced teachers is further compounded by the union's opposition to the suppression of any vacant position in urban schools even if changes in enrollment size would justify such suppression. S. Remaining vacant positions (all rural, as explained above) are filled through new apsointments from the pool of students who have completed their studies in the teacher training  - 56 - ANNEX S Attachment 1 Page 2 of 2 institutions (escuelas normales). Priority is given to students from the federal 'normar schools, to whom employment is guaranteed. The position granted will be, if possible, in the same state as the school; if this is not possible, the position will be in the same region; if this also is impossible, then the position wilt be wherever there is a vacancy. Positions remaining vacant after all students from federal teacher training institutions have been appointed are filled by students from state institutions sc;eened through a selection exam, and exceptionally by students trained in private institutions. SNTE does not play any role in the appointment system. 6. The teacher incentive program supported under the project is expected to have a significant impact on reassignment by inducing teachers to accept longer-term assignments in rural schools and therefore reducing the number of requested moves from rural areas.  MEXICO PRIMARY EDUCATION PROJECr ORGANIZATIONAL SMUCR OF THE PCU AT THE CENTAL LEVEL GENER^L 11| COORDITR rTE ICAL 21 COORDINATOR DIRECTOR TECMNIC"L 3t COODIMWTOR COORDINATOR COORDIVATOR COORDIVATOR | NIORN&TIOK ADIITRAT. STATE 41 CWITlRE 1 1 2 3 4 SYS I COORDI1NATO PCU I _ L _ , SPECI*LIST -PrImary -1dactic -Teacher -Training LCvil, Education Materil Incentives Urban, Works General -Lbrary Supervision Rural, Directorate Books Normales, fINANCIAL P)IOCURNhEN Indigenous -Distrlbution -Evaluation Indigenous COTROKLLER SPECIALIST Education System General -Indigenous -Information _Distance Directorate Reading System Education F E ~~~~Book. IcONACEN It Under-Secretary for educational coordination of SEP. 2/ General Director of Planning, Programing, Budgeting of SEP. ,_ 3/ Techical Committee would be fonred by representatives of the different directorates and/or units iavolved in the project. a 4/ At the state level, there ould be a similar structure, but the Operating Coordinator of the SPCUs would report to th General ,q Director of Coordlnated Services. PA  - 58 - AnX 6 Attachment 1 Page 1 of 13 MEXCO PRIMARY EDUCATION PROJECT PROJECI DESCRIPTION: TEACHER TRAINING I. SUMMARY A. Target Trainees The project would provide training to five different groups of teachers working in federal and state schools in the four project states: (1) General Primary teachers working in rural areas (31,852); (2) General Primary teachers working in urban schools (24,531); (3) Indigenous teachers providing bilingual instruction (13,353); (4) "Community Instructors" (CONAFE) working in extremely isolated and poor areas (1,152); and (5) Teacher Trainers working in the Normal Schools, In-Service Training Centers and state branches of the National Teachers College (862 trainers). B. Training Activities Three types of training activities would be provided: (a) direct in-person instruction during the summer months; (b) semi-irect (or semi-schooled) courses, one course per semester; and (c) distance training activities via radio, television, audiocassettes, videocassettes and pamphlets during the school year. The overall strategy is to reach ail teachers in the four states by the end of the project by providing them access to several training sessions throughout the year, offered in these various formats (see the matrix at the end of this annex for a summary table). Level of attendance and intensity of training received, however, is expected to vary according to school modality, as explained below in more detail. IL. DESCRIPTION OF TRAINING BY TARGET GROUP A Teacher Training for General Primary Teachers 1. Training Stratev (i) Week-Long Intensive Training For 31,852 rural and 24,531 urban primary teachers, the project would provide two types of person-to-person training. The first would be an intensive week (40 hour) of training over the summer. One course would be offered in every school zone throughout the four states (each  - 59 - Attachment 1 Page 2 of 13 zone comprises 10-20 schools), starting in 1992. (Note: Summer-long courses are not possible because of the lack of training facilities and the great dispersion of rural teachers.) (ii) Four-Month Semi-Direct Training The second type of training would be semi-direct (or semi-schooled) classes, in which teachers attend semi-monthly seminar/workshops (four hours each) on Saturdays (or on paydays). These seminar/workshops would be provided in tandem with anthologies, workbooks, self-learning guides and radio programs that teachers would follow at home on a daily basis in between meetings. Courses would last for three and one-half months (one per semester) and would consist of four modules presented during seven workshops. At the end of each month (or module) a small exam would be given during the seminar/workshop to assess progress in mastering the training material and to provide immediate feedback to the teacher and trainer as to what additional measures need to be taken. The semi-direct method means teachers will not have to miss any school days to participate in the training.1 Both of these courses would be voluntary, so 100% participation is not expected. On the other hand, incentives for attendance would be offered: per diems during training and "promotion points' required for advancement to school directorships. Based on these incentives, and the anticipated high quality and interest of teachers in taking the courses, the Secretariat of Education expects attendance to gradually increase over time, from about 20% in the first year (1992) to 30-35% in the last year in each of these courses. The core course in teaching techniques for literacy and numeracy (Semester 1) will be repeated each year, while the summer and second semester courses will vary. By the end of the project, it is expected that all teachers will have at least taken the core course. (iii) One-Time New Rural Teachers Orientation In addition to the strategy described above for all General Primary teachers, new rural school teachers (approximately 5,700, a high figure because of the huge rural turnover rate) would receive a one-time, two-week orientation course at the beginning of the 1992i93 school year. The objectives are: (1) to help inexperienced and young teachers to be effective in multigrade schools with high absenteeism (e.g., by using self-study guides and Audio Primaria cassette tapes, and by working with the community to increase attendance, improve health practices, etc.); and (2) to help them adjust to the rural environment (thereby reducing the extremely high rural teacher turnover). This course will only be given once during the life of the project, as the curriculum of the Normal Schools is expected to be modified to include all elements necessary to prepare new teachers for work in rural areas. 1 An exception to this might apply to rural teachers, if it is decided that the most practic-al time to offer training is on payday (every other Friday), when teachers travel from their schools to the towns in which training centers are located to cash their checks.  - 60 - ANNEX 6 Attachment 1 Page 3 of 13 2. Training Togics Training topics would include: new methods of teaching reading, writing and arithmetic; multigrade teaching techniques; the use and elaboration of educational materials; remedial teaching methodologies to reduce repetition and dropout; educational evaluation and planning; the social and natural sciences; child development; group dynamics; the new pedagogical model being introduced in schools nationwide; and the Education Modernization Program. Each topic would be presented as a module, and the basic training modules for literacy and numeracy would be repeated each year. Course contents would vary by modality (urban or rural) so that they correspond to the actual environments teachers work in. At the same time, since many common themes exist across urban and rural environments, it will not be necessary to design two entirely different courses for each semester, but rather one course which is then adapted to teachers' realities. 3. Program Implementation For both training modalities the project would support: (i) the planning and organization of the program; (ii) the preparation of course materials; (iii) the selection and training of central and state techaical groups, as well as of course instructors; (iv) the actual provision of instruction, including compensations to course instructors and small per diems to cover transport costs of teachers attending the course; and (v) pro,gram supervision and evaluation. (i) Course Planning and Organization For the planning and organization of the training course, SEP will form two central technical groups (GTCs) with 12 specialists in each. One group wil work on rural teacher training and the other on urban, although considerable coordination and cross-fertilization in their work is expected. The specialists will come from within SEP (e.g., Directorate of Primary Education, Directorate of In-Service Training) and from other educational institutions in Mexico (e.g., the National Teackers University, the Latin American Institute for Education Communication, etc.). All specialists will be paid a special (though minor) compensation for the extra time they dedicate to this activity. For both the intensive and semi-direct courses, the GTCs will design the course material (see below) and provide training to the state-level technical groups (GTEs), who will have full- time primary responsibility for organizing the courses. In addition, the GTC will produce and distnrbute a basic training course preparation manual for use in organizing training courses in the states ("la normativa de capacitacion"). In the summer of t . conference between the GTC and four GTEs will be held in Mexico City to initiate p ogram preparation, explain timetables and responsibilities, and provide a mini-workshop on how to plan, organize, implement and evaluate the training program. Subsequently, the GTCs will provide two weeks of training to the GTEs each year in Mexico City.  - 61 - ANNEX 6 Attachment 1 Page 4 of 13 For the design of course materials the GTC will participate in a week-long fact-finding mission ("diagnostico") to define the specific training needs as expressed by teachers and to design rough outlines of the contents of the training modules. The GTC will then prepare the training material (see below), as well as evaluation instruments. Following this, pre-testing ("validacion") of training material will be necessary, for one week in each state. At this time the 8 training modules for the semi-direct courses and the course contents for the week-long intensive course will be presented to "focus groups" composed of teachers and members of the GTEs. Training material will then be revised and mass produced based on the results of the pre-testing. For its part, the GTE will be responsible for: selecting the training sites; ensuring proper logistical support at all training centers (i.e., the supervisors' schools); inscribing interested teachers (through their supervisors) into the intensive course; selecting and training course instructors; and monitoring teachers' progress throughout the school year. The GTEs will provide one week of training to instructors in early July, before the intensive week-long course, and another week will be provided in January before the start of the second semester semi-direct course. Approximately half of all instructors will be trained for the intensive course, with the other half trained for the semi-direct courses. At the same time, it is assumed that (if required) instructors will be capable of providing both types of training. In the last week of August (at which time teachers are obligated to attend "school orientation"), instructors will conduct a two-day induction seminar, explaining the semi-direct training method, the location of training centers, the payment of per diems, the "promotion points" to be awarded for passing the course and other logistical matters. At this time all interested teachers will inscribe in the course and receive the training anthologies, self-instruction guides and other necessary materials (which will have been distnbuted to all supervisors before this time using the normal distribution channels). For the second semester semi-direct course, an initial seminar/workshop will be added (so that there will be 8 in all, instead of 7, in the second semester) which will repeat these induction steps. In September the semi-direct courses will begin, with program monitoring provided by the GTEs. Progress in program implementation, numbers of teachers inscnbed and resolution of logistical problems (e.g., payment of per diems) will be closely followed and corrective measures will be taken if needed. (ii) Preparation of Course Materials Course contents and materials would be designed by teams of specialists contracted for this work from within SEP and from other educational institutions in Mexico such as the National Teacher's University, the Latin American Institute of Education and Communication or the Center for Economic, Social and Anthropological Research. Each course would be composed of four modules, corresponding to the topics described above. Thereafter, the "cascade" technique would be applied, by which this team of specialists (Central Technical Group, or GTC) would train teams of technical specialists in the states (GTEs). In turn, these state technical teams would train the instructors, who would have final responsibility for training the nearly 60,000 teachers.  _ 62 - A6 Attachment 1 Page S of 13 The first semidiect coure would be considered the basic literacy and numeracy course and would be repeated each year. An example of the course structure over the three years of training would be: COURSE 1 - 1992 COURSE 2 - 1993 Techniques for Rading Techniques for Reading and Writing I and Writing E Mathematics I Mathematics H Use of Educational Multigrade Teaching Materials Techniques Evaluation Educational Planning COURSE 1 - 1993 COURSE 3 - 1994 Techniques for Reading Natural Sciences and Writing I Social ScieTices Mathematics I Group Dynamics Use of Educational Child Development Materials Evaluation COURSE 1 - 1994 COURSE 4 - 1995 Techniques for Reading Regional Studies and Writing I Educational Mathematics Administration Use of Educational Multigrade Teaching Materials Techniques Evaluation Education and Society The courses would have anthologies, self-instruction guides, workbooks and training manuals for instructos The anthologies would include bibliographical material and pedagogical theory with respect to the module, followed by practical applications of this theory at the various grade levels. The anthologies would be closely linked to the radio programs which would be broadcasted daily throughout the four states These programs would cover the same subject matter as the module beig covered, and the anthologies (or workbooks) would include exercises pertaining to the radio programs. The semi-schooled courses would be supplemented by sets of audiocassettes (distributed to each school by the project). Ihese cassettes wold be designed by the teams of specialists for use in the classroom and would provide examples of how to apply the training course material in the classroom. In addition, n 50 selected training centers video training programs will be shown during the semina/workshopL These are the same programs as those to be used by Indigenous  - 63 - ANNEX6 Attachment 1 Page 6 of 13 teachers and CONAFE instructors, with some programs specially designed for General Primary teachers (see Part Im, Distance Education and Training Technologies, for a full description of the distance methods to be applied). (iii) Selection and Training of Instructors Both types of training would be provided by highly qualified primary teachers working in the four states, technical support staff, underutilized Normal School teachers and school supervisors. Instructors would have to meet certain qualifications, such as holding a university degree and having at least 5 years of teaching experience. All primary teachers would be invited to apply for positions as instructors ("asesores") after which those best meeting the qualifications would be selected for training. Initially, an extra 10% of required instructors would be selected. After receiving a week of specialized training in July, enabling instructors to provide both the intensive courses and the semi-schooled courses, the top 90% would be selected (based on a final exam). Before the start of the second semester semi-direct course, another week of intensive training would be given to those instructors selected the previous July. This process would be repeated each year, ensuring motivated, qualified trainers. The total number of instructors to be selected and trained is based on an estimated 35:1 ratio of teachers to instructors in the courses. For urban teachers 701 instructors would be required, whereas for rural teachers 911 instructors would be trained, calculated in the following manner: URBAN TEACHERS: Teachers and State Directors Instructors (35:1) Chiapas 6,332 181 Oaxaca 5,966 170 Guerrero 8,382 240 Hidalgo 3.851 110 Total 24,531 701 RURAL TEACHERS: Teachers and State Directors Instructors (35:1) Chiapas 8,501 243 Oaxaca 8,707 249 Guerrero 8,920 255 Hidalgo 5,724 164 Total 31,852 911  - 64 - ANNEX 6 Attachment 1 Page 7 of 13 (iv) Actual Instruction Considering 7 semi-monthly seminar/workshops per semester, including a final exam at the seventh meeting, a total of 9,814 seminar/workshops would be provided each year by the project for urban teachers (701x7x2), with 12,754 seminarAworkshops per year for rural teachers. Each teacher would attend 56 hours of instruction each year, in addition to putting in an estimated 40 hours of self-instruction at home (5 hours per month for 8 months) and listening to 45 hours of radio training programs (180 programs, 15 minutes each). This organization is based on expected high attendance and would require adjustments (on the basis of discussion during project implementation reviews) if expectations were not being met. As with the intensive week-long summer courses, seminar/workshops would be held in every school zone. The instructor would be the facilitator of these sessions, reviewing the self- study teaching material, proposing discussion topics, answering teachers' questions, suggesting ways to apply the training material in the classroom, and providing immediate feedback to teachers on their performance based on monthly module quizzes. These quizzes would provide important information as to the quality of the course, the extent of mastery of training material, and what measures teachers need to take (e.g., increased study-time) to improve their performance. At the end of each course a final exam will be given, passage of which will earn teachers "promotion points" ("puntos escalafonarios"), necessary for advancing to school directorships. Lastly, the instructor would certify teacher attendance with a SEP stamp in the training schedule provided in the beginning of the anthology. This would authorize payment of small per diems to teachers. For each class rural teachers would receive 25,000 pesos for transportation and token incentive (about US$8), while urban teachers would receive 20,000 pesos (about US$7). These per diems are expected to induce regularity of attendance. (v) Supervision and Evaluation As stated above, the GTEs would have primary responsibility for program supervision, although the course instructors and GTC would also play important roles. At the end of each course, the instructors would sub.nit the results of the monthly quizzes and the final exam to the GTC and GTEs, who would then analyze this information and use it in the revision of existing training materials and the design of new courses. In addition, at least one week per year, the GTC would visit schools in all four states (3 specialists for rural teachers and 3 for urban in each state), assessing the impact of the training in the classroom and getting first-hand feedback from the teachers as to the quality of the training courses and ways to improve them. B. Teacher Training for Bilingual Teachers 1. Training Strategy A high priority is given to upgrading the skills of indigenous teachers, as over half of them have not received formal teacher training courses and over a third have at most only a lower  - 65 - ANNEX 6 Attachment 1 Page 8 of 13 secondary education. The project would provide one week of training during the summer months to every indigenous teacher in Oaxaca, Guerrero, Chiapas and Hidalgo. The attendance rate is expected to reach practically 100% of teahcers by the second year of program implementation. Training would be conducted for eight weeks each summer at 16 training centers (14 Centros de Integracion Social and 2 Albergues) maintained by the Department of Indigenous Education (DGEI). Courses would be given by quaUfied teachers and technical support staff of the DGEI, who would receive two weeks of specialized training prior to providing courses themselves. This training would complement two other existing programs: (i) a six-month pre-service training provided once at time of recruitment; and (ii) the on-going distance education certification program begun in 1990, in which teachers study to earn their primary, secondary and baccalaureate degrees (as needed), so that all Indigenous Primary teachers will eventually have their licenciatura. (Approximately 7,000 teachers are enrolled in these programs in the four states, which include 4 weeks of direct instruction over the summer and during school vacations.) lherefore, most Indigenous Teachers will receive 5 weeks of direct classroom instruction each year, although all teachers will receive at least the one week each year for three years to be provided by the project. As with General Primary teachers, the basic technique to be used is the "cascade" approach, in which a central technical group trains state technical groups, who then train the course instructors. Again similar to General Primary training, instructors will be highly qualified teachers who share the same concerns and teaching realities as the indigenous teachers who will attend the courses. 2. Training Topics In the first year, the application of bilingual reading materials to be provided by the project would be emphasized, which would include training in the basics of reading and writing in indigenous languages. In addition, training in the use of educational materials and library books to be distnbuted by the project would be provided. In the second year, multigrade teaching techniques would be emphasized, as the vast majority of indigenous schools lack teachers for all six primary grades. In the third year curriculum development and lesson planning, specially adapted for bilingual teachers, would be emphasized. As a complement to direct instruction, video training programs to be designed and produced by the project would also be shown during the weeks of intensive training (see below for description of distance training activities). 3. Program Implementation As described above regarding General Primary teachers, the project would support: (i) the planning and organization of the program; (ii) the preparation of course materials; (iii) the selection and training of central and state technical groups, as well as of course instructors; (iv) the actual provision of instruction, including compensations to course instructors; and (v) program supervision and evaluation. The basic structure and procedures as those described for General  - 66 - ANNEX 6 Attachment 1 Page 9 of 13 Primary teachers would also be used for Indigenous teachers, except there would only be the intensive week of training to organize (for eight consecutive weeks). In addition, because the courses will be given at 16 regional centers with fu&l lodging and eating facilities, the project would also finance the rehabilitation and re-equipping of the 16 centers. This wifl avoid payments of per diems to teachers and result in considerable savings (room and board costs are estimated at only US$2.20 per day compared to US$20 per day for rural teachers during intensive classes). Also different from the training strategy proposed for General Primary teachers, all supervisors and sector chiefs would attend the training of the course instructors. This is so supervisors can provide continuous pedagogicaltechnical support to teachers throughout the school year as part of their school visit routine. In addition to the direct training descnbed above, indigenous teachers would have several opportunities for distance education. Radio programs would be produced by the project for broadcast in seven different languages (40 programs per language). They would provide both teachers and students with special didactic and linguistic support in the classroom. The radio stations of the National Indigenous Institute (with stations and transmitters covering the four project states) would be used for these broadcasts. Secondly, all indigenous teachers would have access to the daily radio program (broadcast in Spanish) which is to complement the semi-direct courses offered to rural General primary teachers (see above). These programs, while designed to follow the curriculum of the semi-direct course, would offer important training lessons on their own. Additional radio broadcasts promoting community participation and proposing practical applications of academic subjects to daily rural life would be transmitted, one program per week for three years (156 programs overall). Third, all indigenous schools will receive a set of audiocassettes which cover the full range of subjects to be taught at the primary level. These audiocassettes are intended to provide direct support to the teacher in the classroom, particularly tc those working in multigrade schools, and will be directed at students. In this sense, the audiocassettes will not have an explicit training function for teachers, although quality programmed teaching materials typically have a training element for teachers in that they provide teachers with new ways to present material.! Finally, bilingual teachers working in the state of Oaxaca will be able to watch educational television programs produced as part of the pilot Tele-Primaria project in that state.