46486 What Can a Regional Government Do to Improve Education? The Case of Junin Region, Peru What can a Regional Governmet do to improve education? The Case of Junin Region, Peru EQUIVALENCIES (Effective Exchange Rate, November 4th, 2007) Monetary Unit = Soles (S/.) S/. 2,99460 = US$ 1 Abbreviations and acronyms APAFA Parents' Association (Asociación de Padres de Familia) CAP Personnel Assignment Chart (Cuadro de Asignación de Personal) CEI Preschool Education Center (Centro de Educación Inicial) CNE National Education Council (Consejo Nacional de Educación) COMUNED Municipal Education Council (Consejo Municipal de Educación) CONEI Institutional Education Council (Consejo Educativo Institucional) COPAREJ Junin Regional Participation Council (Consejo de Participación Regional de Junin) CTAR Transitory Council for Regional Administration (Consejo Transitorio de Administración Regional) CVR Truth and Reconciliation Commission (Comisión de la Verdad y Reconciliación) DESCO Center for Development Studies and Promotion (Centro de Estudios y Promoción del Desarrollo) DRE Regional Education Office (Dirección Regional de Educación) EBR Basic Regular Education (Educación Básica Regular) ENAHO National Household Survey (Encuesta Nacional de Hogares) ENCO National Continuous Survey (Encuesta Nacional Continua) IE Education Institution (Institución Educativa, i.e. school) INEI National Institute of Statistics and Information Technology (Instituto Nacional de Estadística e Informática) INFES National Institute for Education and Health Infrastructure (Instituto Nacional de Infraestructura Educativa y de Salud) ISP Post-secondary Pedagogic Institute (Instituto Superior Pedagógico) LBD Law on the Bases of Decentralization (Ley de Bases de la Descentralización) LGE General Education Law (Ley General de Educación) LOM Muncipal Organic Law (Ley Orgánica Municipal) LOGR Regional Organic Law (Ley Orgánica de Gobiernos Regionales) MED Ministry of Education (Ministerio de Educación) MED-UEE Ministry of Education- Education Statistics Unit (Ministerio de Educación - Unidad de Estadísticas Educativas) MED-UMC Ministry of Education- Education Quality Measurement Unit (Ministerio de Educación - Unidad de Medición de la Calidad Educativa) MEF Ministry of Economy and Finances (Ministerio de Economía y Finanzas) MRTA Tupac Amaru Revolutionary Movement (Movimiento Revolucionario Túpac Amaru) NBI Basic Unsatisfied Need (Necesidad Básica Insatisfecha) OOII Intermediate Bodies (Órganos Intermedios) PBI Gross Domestic Product (Producto Bruto Interno) PCP-SL Communist Party of Peru - Shining Path (Partido Comunista del Perú - Sendero Luminoso) PEI Institutional Education Project (Proyecto Educativo Institucional) PRONOEI Non-formal Preschool Program (Programa no escolarizado de Educación Inicial) SIAF Integrated Financial System (Sistema Integrado de Administración Financiera) UC Pricing Unit (Unidad de Costeo) UE Executing Unit (Unidad Ejecutora) UGEL Local Education Management Unit (Unidad de Gestión Educativa Local) USE Education Service Unit (Unidad de Servicio Educativo) This study was lead by Daniel Cotlear (LCRHD) and Inés Kudo (consultant), with the participation of Cristian Aedo (LCRHD), Carmen Osorio (LCRHD), Marcela Echegaray, Enrique Prochazka, José Velásquez and Eduardo Ballón (consultants). Acknowledgments The team would like to thank the following individuals for their contributions: Livia Benavides (LCRHD), Alberto Rodríguez (LCRHD), Ian Walker (LCRHD), Erik Bloom (LCRHD), Eduardo Vélez Bustillo (LCRHD) and Luis Crouch (RTI); and the valuable support of Érika Bazán (LCRHD). We would also like to thank the following people for their collaboration in this effort: Cecilia Thorne, Tesania Velázquez and Marcia de la Flor (Psychology Department at the Pontifical Catholic University of Peru). We would also like to acknowledge the fieldwork researchers for their assistance: Mercedes Ayarza Hernández, Mariella Bazán Maccera, Arturo Cárdenas González, Inés Deustua Rossel, Andrés Figallo Brero, Luis Ernesto Fodale Vargas, Javier Hayashi Hiraoka, Inés María Lazarte Heraud, Diana Loli Valverde, Rosela Millones Cabrera, Eliana Neyra Cruzado, Alberto Piscoya Magallanes, Patricia Ramos Olivera, Jorge Rivas Rivas, Luz Uriol Gárate, María Natividad Vila Torres and Mónica Malpartida. Contents Executive summary The region of Junin: images and reality..................................................................................................................................i The education supply in Junin: achievements and problems..................................................................................................i Financing education in Junin................................................................................................................................................iii Stagnation in education ........................................................................................................................................................iii Three recommendations for a Regional Government that is committed to education ..........................................................vi Chapter 2 Introduction...............................................................................................................................1 2.1.Education Decentralization ................................................................................................................................ 1 2.2.Education in the context of Junin....................................................................................................................... 2 2.3.The legacy of violence in the education system................................................................................................. 4 2.4.Challenges for education in Junin ...................................................................................................................... 5 Chapter 3 Education Resources.................................................................................................................7 3.1.School Infrastructure and Equipment................................................................................................................. 7 Access to basic services: electricity, water and sewage systems ...........................................................................................7 Quality and salubrity of sanitary services..............................................................................................................................8 Maintenance of infrastructure and school furniture...............................................................................................................9 Available teaching materials ...............................................................................................................................................10 Why is infrastructure so poorly managed? ..........................................................................................................................11 3.2.Human resources in the schools in Junin.......................................................................................................... 11 The inefficient and inequitable distribution of human resources.........................................................................................12 Difficulties in evaluating teacher quality.............................................................................................................................13 Inadequate working conditions............................................................................................................................................14 3.3.Financing education ......................................................................................................................................... 15 Now more is spent but not well spent..................................................................................................................................17 Current expenditure on education in Junin is not equitable.................................................................................................19 Non-salary expenditure is insufficient and is not distributed based on homogenous criteria ..............................................20 Management problems are not evident when comparisons are made with other regions, except in preschool education ...21 Expenditure efficiency varies between schools with similar characteristics........................................................................23 3.4.Conclusions...................................................................................................................................................... 24 Chapter 4 Education Results....................................................................................................................25 4.1.Measuring reading skills .................................................................................................................................. 26 4.2.Fieldwork Results............................................................................................................................................. 27 Results for reading skills in Junin are generally low ...........................................................................................................27 Who is at a disadvantage? ...................................................................................................................................................31 4.3.Conclusions...................................................................................................................................................... 35 Chapter 5 The greater challenge: education management....................................................................36 5.1.The UGEL: the weakest link............................................................................................................................ 37 The field of action exceeds the UGEL's capacities.............................................................................................................38 Problems in the selection of specialists ...............................................................................................................................39 Few incentives for good performance .................................................................................................................................39 5.2.School management and participation.............................................................................................................. 41 The schools' demand for more autonomy ...........................................................................................................................42 A shared vision and clear objectives ...................................................................................................................................43 5.3.Conclusions...................................................................................................................................................... 44 Chapter 6 Recommendations to a Regional Government Committed to Education...........................45 6.1.Declare measurable goals................................................................................................................................. 45 Learning goals.....................................................................................................................................................................46 Goals for enrollment and expenditure efficiency.................................................................................................................47 Goals for operating conditions at schools............................................................................................................................47 6.2.Provide necessary pedagogical and financial support ...................................................................................... 48 Pedagogical support ............................................................................................................................................................48 Financial Support ................................................................................................................................................................50 6.3.Regular progress evaluation and dissemination of results................................................................................ 52 Reports on reading skills .....................................................................................................................................................53 Reports on minimum operating conditions..........................................................................................................................54 Reports on financial resources.............................................................................................................................................54 Dissemination of information aggregated by districts, provinces and the region ................................................................55 Annexes 1. Annexes in Chapter 1 ....................................................................................................................................... 56 Annex 1.1. Accessibility of Public Services in Junin.........................................................................................................56 Annex 1.2. The Ashaninka war and the spiral of violence in the rainforest.......................................................................57 Annex 1.3. Socio-economic and Ordering Characteristics.................................................................................................60 Annex 1.4. Dictionary of Statistical Indicators ..................................................................................................................61 Annex 1.5. Summary of Statistical Indicators....................................................................................................................64 2. Annexes in chapter 2........................................................................................................................................ 65 Annex 2.1. Status of school infrastructure and ordering ....................................................................................................65 Annex 2.2. Status of school equipment and ordering.........................................................................................................68 Annex 2.3. Status of human resources and ordering ..........................................................................................................69 Annex 2.4. Status of education financing and ordering......................................................................................................73 Annex 2.5. Expenditure executed on education and the number of students in the system (2000-2006)...........................75 Annex 2.6. Per student expenditure by level of education (% of per capita GDP).............................................................76 Annex 2.7. Expenditure executed on Education and Culture by program and budget allocation (current soles 2006) ......77 Annex 2.8. Distribution of Education Expenditure (2000-2006) .......................................................................................77 3. Annexes in chapter 3........................................................................................................................................ 80 Annex 3.1. Education results, learning achievements and ordering...................................................................................80 Annex 3.2. Fieldwork research ..........................................................................................................................................82 Annex 3.3. Evaluation of reading competences .................................................................................................................89 Annex 3.4. Calculating continuous values in reading performance ...................................................................................90 Annex 3.5. Comparison of Items with Curricular Objectives ............................................................................................91 Annex 3.6. Differences in the percentage of achievement on the census test (May 2007) between schools that participated in the 2006 census and those that did not.....................................................................................93 Annex 3.7. Socio-demographic and institutional factors associated with reading performance.........................................93 Annex 3.8. Travels in Junin: Fieldwork adventures...........................................................................................................95 4. Annexes in chapter 4........................................................................................................................................ 99 Annex 4.1. Location of control over key decisions for EBR according to the law.............................................................99 Annex 4.2. Comparison of functions according to Peruvian law .....................................................................................100 5. Annex in Chapter 5 ........................................................................................................................................ 108 Annex 5.1. Description of the achievement levels on the census test ..............................................................................108 References ................................................................................................................................................109 List of charts Figure 3-1. What are the minimum work conditions necessary for you to be able to do your job well at this school?*...................................................................................................................................................................15 Figure 3-2. Current expenditure on education in Junin: fiscal flows...........................................................................................20 Figure 3-3. Results for education coverage in Peru by region.....................................................................................................22 Figure 3-4. Education achievements in Peru by region ...............................................................................................................22 Figure 3-5. Efficiency-achievement matrix, Stochastic Frontier Model (Junin)..........................................................................23 Figure 4-1. Fluency and comprehension: level reached in May in comparison with the level expected .....................................29 Figure 4-2. Reading performance of Ashaninka children and non-Ashaninka students in comparison with expected results ......................................................................................................................................................................33 Figure 4-3. Student reading performance by school type based on the anticipated result ...........................................................34 List of tables Table 2-1. Socio-economic characteristics in Junin ......................................................................................................................2 Table 2-2. "Rurality" of the population and Basic Regular Education (% rural) ........................................................................3 a Table 3-1. Status of school infrastructure......................................................................................................................................8 Table 3-2. Status of School Equipment.......................................................................................................................................10 Table 3-3. Status of human resources..........................................................................................................................................12 Table 3-4. Expenditure executed in the region of Junin on Education or Culture by financing source and expenditure type (new soles 2006) ..........................................................................................................................16 Table 3-5. Financial Status of Education (2006)* .......................................................................................................................16 Table 3-6. Per student expenditure in the period 2000-2006 (new soles at prices in the year 2000)*.........................................17 Table 3-7. Expenditure executed on Education and Culture by item (new soles at prices in the year 2000)...............................18 Table 3-8. Current expenditure executed on Basic Regular Education in Junin by province (2006)...........................................19 Table 4-1. Education results and learning outcomes ...................................................................................................................25 Table 4-2. Level of fluency achieved per grade (% of students) .................................................................................................28 Table 4-3. Average number of months of delay in reading skills................................................................................................31 Table 4-4. Regression results by grade........................................................................................................................................32 Table 6-1. Report on second grade reading comprehension........................................................................................................54 Table 6-2. Measurement scale for the physical environment ......................................................................................................54 Table 6-3. Report on the school's physical environment.............................................................................................................54 List of boxes Box 1. "DOGO"..........................................................................................................................................................................27 Box 2. Education Municipalization in Junin ...............................................................................................................................40 Box 3. The journey of a broom ...................................................................................................................................................42 Box 4. Solaris: improving reading with simple strategies and everyone's participation .............................................................50 Box 5. How can we ensure good financial management in schools? International experience ...................................................52 Executive Summary At the beginning of 2007, the World Bank published the book titled Por una educación de calidad para el Perú (Towards quality education in Peru). The book was widely distributed and its main findings propitiated a series of debates. Shortly after, the World Bank received a letter from the President of the Regional Government of Junin requesting advice on how to improve education in his region. The book in question analyzes the problems affecting basic education and provides recommendations for the entire country. The President of Junin was interested in improving the quality of education in his region and was seeking concrete proposals regarding decisions that the Regional Government could make. Given the existing legal framework, the region's fiscal situation and the balance of political forces in the Education sector, the President of the Regional Government of Junin asked the following question: What can a Regional Government that is truly committed to education do? It was the first time that the World Bank had received a request of this kind, and following an internal discussion process, it decided to conduct a study. The conclusions of this study are presented in this document. This study intends to be useful not only to the Regional Government of Junin but also to other regional governments committed to education. It is important to point out two elements: first, abundant legislation exists laying out the responsibilities of different instances and levels of the Education sector. These legal documents overlap and are rife with ambiguities and a number of contradictions regarding the roles of each instance. This opens the door for the Regional Government to assume a leading role in its quest to strengthen its autonomy if and only if it has the political will to do so and is equipped with a clear design for intervention. Second, although this study explores the main challenges in terms of the quality, quantity and distribution of human resources, a conscious decision has been made not to conduct an in-depth evaluation of teachers' possible rationalizations because certain political economy factors make this an extremely sensitive subject f national relevance and long-term solution. As such, this study focuses on administrative areas that can be handled by the regional government in the short and mid-term. The region of Junin: images and reality Our initial conversations with public officials, administrators and teachers in the region provided us with an image of the region and its problems in terms of education that is quite different from the reality that we found later on. To be specific, the initial interviews suggested a rural region whose population is geographically disperse, and as such faces serious problems with access to schooling, and where parents are more often than not illiterate. Nevertheless, a statistical analysis of Junin leads to different conclusions: although poverty levels are high and similar to national levels ­50 percent of the population is poor and 17 percent is indigent­ the population is primarily urban (57 percent) and the schools are so widespread that geographic access to education is no longer a significant problem. What we did find were very marked differences between the high and low levels of the region and between the mining and farming provinces. The most significant distinction is between the area of the sierra, which is home to 80 percent of the population, and the rainforest areas, which have approximately 20 percent of the population. The latter group has more ethnic and linguistic complexity and less developed public services and infrastructure and faces challenges stemming from significant geographic dispersion, political violence and drug-trafficking. The education supply in Junin: achievements and problems The region of Junin has 1.3 million inhabitants. Its size corresponds to a small educational jurisdiction in international terms. The educational offer is high: there are 4,200 schools and 21,500 teachers for 350,000 students. This means that significant physical access to schooling exists (99 percent of the population live less than one hour from an education center) and an abundant supply of teachers: an average of nineteen ii What can a Regional Government do to improve education? The Case of Junin students per teacher, which is better than the ratio stipulated by Peruvian norms (between nineteen and forty students depending on the school type and location). Although overall educational infrastructure shows evidence of considerable investment in the past, the quality of the same was poor. This is due in part to the lack of access to public services: 80 percent of the schools have electricity but barely 57 percent have water and only 39 percent have sewage systems. Nevertheless, the majority of infrastructure problems are due to poor management. Many classrooms have deteriorated because of lack of maintenance: broken windows, abandoned bathrooms or ramshackled desks. Even the public services, when they exist, are underused: a fifth of the schools have no toilets even when the school is located in a town that has a sewage system. Poor management is also responsible for the limited supply of materials: incomplete libraries, delayed distribution of textbooks during the school year and absurd logistical problems with cleaning material distribution. Poor management isn't only rural; it's also urban. Poor management is also palpable in human resource administration. The distribution of salary expenditure favors the richest provinces and districts. This means that teachers are not assigned to compensate for higher poverty levels or rurality: salary expenditure is concentrated in richer and more urban regions. In other words, the higher ratio of teachers per student in Junin is not destined to improve coverage in poor rural areas but is instead concentrated in wealthier areas. And why is infrastructure management deficient? Why does school maintenance, and procurement in general, work so poorly? In theory, this is the Local Education Management Unit's Responsibility (UGEL) but there appear to be three fundamental reasons why this theory is not put into practice: · The UGEL has a number of responsibilities, including school maintenance via the distribution of inputs or funds. This responsibility is very low priority given that in practice, the UGEL is not accountable for this task. In fact, the UGELs are seen as agencies that administer payrolls and personnel according to a complex set of rules. Their other functions (pedagogical support to schools and provision of goods and services) are seen to only if there is enough energy and funds left after handling personnel requirements. · The UGEL's responsibility in the area of maintenance is even more diluted because most of them are not executing units (meaning that they do not have authorization from the Ministry of Economy and Finances - MEF to administer funds). Only two UGELs have this authority, but the other seven do not have the size or capacity necessary to receive this authorization and their functions are under the responsibility of the Regional Education Office (which covers the jurisdictions of Huancayo, Concepción, Jauja y Chupaca), the UGEL Tarma (which manages the finances of the UGELS of Tarma, Yauli and Junin), or the UGEL Satipo (which covers Satipo and Chanchamayo). No one is accountable for maintenance tasks, given that they are considered marginal functions. The idea that the UGEL is in charge of distributing funds for goods and services is a fiction in the education sector, although MEF does not see it this way. · The parties interested in maintenance of schools are located in the schools, but the UGELs are not accountable to them for this role and are often located physically far from the schools they monitor. In the school, no one is responsible for maintenance results: if no inputs arrive ­because they are kept in the UGEL­ no one in the school takes responsibility. Given that there are no funds and no capacities to use them, no one bothers to identify opportunities to resolve the situation. This means that there are no projects or initiatives to procure goods and services that may improve learning. iii What can a Regional Government do to improve education? The Case of Junin Financing education in Junin In Junin and Peru as a whole, public expenditure on education has increased considerably over the past six years. This growth has concentrated on salary items while non-salary spending ­on investments, maintenance and educational material­ remain very low. In the year 2000, Junin spent less per student than the national average. This gap has lessened relative to 1996 due to an increase in expenditure in Junin and a faster decline in the number of students in the region compared to the country: the decline in enrollment was 10 percent in Junin and 3 percent in the country, which is in line with the demographic patterns registered in the 1993 and 2005 censuses. Unlike personnel spending, expenditure on goods and services is very low. This includes entries for water and electric bills, educational material, cleaning material and maintenance. The limited spending on goods and services explains why the supply of water and electricity services has not increased (including in areas where connections are available), sanitary services are seriously deteriorated, windows are broken, roofs are leaky, and bats have infested the rafters. Junin is the third region in the country that allots the least for goods and services per student. In fact, it allots less than poorer regions. Other countries have tried to tackle similar problems by allotting funds to the education community in each school. For example, Armenia and New Zealand delegate the majority of administrative responsibilities to school boards. El Salvador, Guatemala and Honduras have done the same for rural schools in remote areas. Many other countries have decentralized specific functions such as textbook and material purchases, renovation of school infrastructure, and school maintenance among others. To achieve this, a methodology was applied that has proven successful in various countries. We suggest applying the same in the region of Junin: · Determine the minimum operating criteria (to be realistic, they should include a baseline and determine goals that the community and the Regional Government consider feasible to achieve in two years). · Evaluate the cost needed per student to determine minimum operating criteria. · Distribute school resources in a transparent and equitable way. · Set up a public information system on progress, indicating which criteria are being met by which school and which are not being fulfilled. Contrary to the fear often stirred by these reforms each time they are implemented, international evidence suggests that they work well and do not generate more corruption than what is already present in systems like the one currently used in Peru. These models incorporate periodic evaluations on administrative, financial and educational performance, which are typically designed and conducted by the central government. In the majority of cases, the central government checks prior expenditure before authorizing new funds. Stagnation in education Junin has made significant progress in education coverage, which is almost total in Primary education and is higher than the national average in Secondary (75 percent). Consequently, the education levels among young adults (18 to 25) are on average high: 78 percent of them have completed secondary education (10 percent more than the national average). In contrast, Preschool education has less coverage in Junin than in the rest of the country (second to last among the 25 regions). The poor coverage at this level is due in part to the low budget for this area but can primarily be explained by inefficient expenditure management in comparison to expenditure on Primary and Secondary education. Without a doubt, this region has lower coverage than one would expect given the expenditure level. Finally, the illiteracy rate is similar to the national average and is less than 10 percent. iv What can a Regional Government do to improve education? The Case of Junin Despite progress in coverage and the fiscal, legislative and organizational efforts made in recent years, the quality of education remains low. The study administered a survey to a random, representative sample of public schools in Junin. This effort was conducted between May and June of 2007 in 83 schools in all nine provinces of the region. The survey included 50 directors, 118 teachers and and 181 parents; two reading tests were administered to a total of 1.684 children. One of the tests was the one used by the Ministry of Education in the census evaluation conducted in December of 2006, and the other one was the reading fluency test used by the World Bank in 2005. The main conclusions were: · Junin has similar reading performance results to those obtained in the national study: the majority of children read poorly (lack fluency for their age) and their comprehension, whether literal or inferential, is low. · As the years go by, learning lags behind expected results. In sixth grade, children are three years behind because they read at a third grade level. Without a doubt children improve but fall short in terms of an absolute goal. If parents are unaware of the goal or are unable to measure the children's progress in comparison with the goal, they may erroneously believe that education improves with time. · Children that received Preschool education read better in the early grades than others. This gap persists throught Primary school, although by sixth grade all students, regardless of whether they had a Preschool education or not, have fallen behind with regard to learning goals. · Children in urban schools read better than those in rural schools. Again, although the gap persists, as time goes on everyone falls behind. This means that over time, urban and rural students perform below standard and the rural-urban gap is eclipsed by the low student performance across the board. · If we compare single teacher schools or multi-grade schools to full grade schools, we find that students in the latter read better in the early grades. The disadvantage that students at multi-grade and single teacher schools have is sizable and parents and teachers can see it. But what is most striking is that even students at full grade schools perform below standard. This means that not even those who attend schools with more abundant and better resources are capable of reaching the target learning level. · Among the ethnic groups identified in the study, the Ashaninka students obtained results far below the rest. In fact, their results are below those obtained by Quecha speakers, who have slightly lower results than Spanish speakers. · There is a great deal of variability of average scores for schools with similar characteristics. The analysis suggests that the efficiency of school management is directly related to education quality. Although problems with education are much more serious in rural areas and at single teacher schools, our results show an even more worrying scenario: low quality education is the norm. On average, the students at urban full grade schools finish Primary with a reading level that barely meets fourth grade standards, which means that they are two years behind in terms of both pedagogical standards and the resources invested. In other words, the average results are below expectations despite the schools' locations and characteristics. But as in the national study, Junin is not just the bearer of bad news. Instead, this study found that some schools, working under precarious conditions in very isolated areas, have good learning results (measured in terms of comprehension and fluency). For example, surveyors were impressed by the efficiency of a single teacher school in Pichanaki, Chanchamayo, which, despite its location in an area with high poverty levels and a history of political violence, obtained excellent results on the reading tests. v What can a Regional Government do to improve education? The Case of Junin The primary implication of these findings is that the deficiencies in learning in Junin are not limited to rural areas and cannot be blamed on either distance or lack of resources given that urban schools suffer from the same difficulties. This suggests that the root of these deficiencies is poor management. Inefficient management is visible not only in the poor results obtained in schools that concentrate the largest investment of the sector's resources (urban full grade) but can also be evident in the fact that some schools receive substantially more resources than others and it is precisely those that receive less that serve the poorest children and the most vulnerable populations. The national study conducted an in-depth analysis of why efficient schools achieve good results even under precarious conditions and concluded that these schools have three characteristics that should be generalized to improve education results throughout the system: the existence of standards, support for capacity development, and accountability. Next, we will comment on the main impressions of Junin with regard to these characteristics. Standards and goals: International literature suggests that efficient schools have clear, measurable standards or goals. The General Law of Education stipulates that each school must produce an Institutional Education Project (PEI) that serves as the basis for the teaching plan. In order to ensure that the PEI is participative and known to all the community, the law also indicates that it must be signed by members of the Institutional Education Council (CONEI). The region must also prepare and publish a regional education plan, which will be used as the basis for the Regional Government's actions in the education sector. In practice, the law's good intentions for the PEI have resulted in a bureaucratic process that has not necessarily translated into improved management practices. During the visits conducted for the study, we found that only a third of the schools in Junin have a PEI. The worst news is that the schools that have fulfilled this requirement have not developed comprehensible objectives or quantifiable goals. The PEIs are mandatory for certain procedures and the basis to evaluate school principals. This has generated a small industry for PEI reproduction where the school names are merely changed in the copid documents and no participative process is conducted with the school community. The PEIs in the schools visited are excessively complex, abstract and ambitious (for example, one school proposed "generalizing equity of opportunities for all Peruvians to access a quality education"). These statements do not lead to specific goals and there are no plans for actions whose progress can be measured. Participation is also fictional given that general statements do not mobilize individuals to engage in concrete actions that influence teaching or student learning. Although the law's intention of adapting plans to each school's necessities is laudable, in practice this is not what is happening in Junin. It would be preferable to establish clear goals for reading fluency and comprehension to evaluate how the student group is progressing month-to-month and determine how much a school improves from year-to-year. Support for capacity building: As we have already mentioned, the UGEL fulfills diverse administrative functions that significantly limit its action in pedagogical matters. The supervisory and accompaniment functions of UGEL specialists are hindered by the number of schools that the UGEL must serve, many of which take several days to reach. For example, in Chanchamayo, four specialists must oversee 378 primary schools. These individuals indicate that it takes five days round trip to visit the most distant schools. In the past year, only 30 percent of the schools were visited, falling far short of the goal of visiting each school twice a year. On the financial side, serious deficiencies exist that generate an inequitable distribution of resources. First, there is no fixed expenditure that guarantees appropriate maintenance of each school, and second no single, technically adequate mechanism has been developed to allot budget resources according to variables such as school size, the student body's special education needs or conditions related to access, dispersion, etc. Accountability. Given that the schools lack goals that can be regularly and easily measured and reported, it is impossible to follow-up on the students' educational results. There are also no minimum standards for vi What can a Regional Government do to improve education? The Case of Junin educational resource provision with which we can compare current provisions to schools with regard to their infrastructure, furniture and working materials. Additionally, the competences of actors responsible for providing and demanding information have not been clearly defined. Currently, no one is responsible for providing information on the state of a school and its educational results and parents make no requests for this information. Three recommendations for a Regional Government that is committed to education Different management problems were identified in the Junin region, which are reflected in education stagnation and the poor quality of infrastructure. As such, our recommendations are focused on: (i) establishing (few) quantitative goals that allow us to evaluate if, for example, the curriculum is being fulfilled or not in the classroom or if schools meet or exceed minimum operating standards (ii) establish the necessary pedagogical support to facilitate effective supervision of and feedback for teachers, and (iii) measure progress and inform the community, propitiating the population's active participation in improving the quality of education. This requires delivering more resources to schools and allowing for more administrative autonomy but not without first establishing clear goals and a system for providing information on progress and which facilitates more accountability to the community and Regional Government. Specifically, there are three recommendations: State quantitative goals The diagnostic suggests that the education system in the region is going adrift. Neither the region nor the school has real goals (when goals have been formalized, they constitute general statements that are ambitious in scope and impossible to measure and understand and as such are not useful as goals). The region should propose goals that can be reached in a few years and which are relevant and understood by parents. Progress regarding these goals should be measurable at least once a year. The Regional Government should identify its goals. The following are examples of the type of goals we recommend: (i) By the end of the 2008 academic year, there will be a 5 percent increase ­ relative to 2007­ in the proportion of students that reach a sufficient level of proficiency in the Ministry of Education's census test for both communication and Logic-Mathematics, and none of the students will score zero. (ii) In three years, there will be a 15 percent increase in the proportion of 3 to 5 year olds that have access to preschool education without exceeding the maximum number of students allowed per class at this level. (iii)By the inauguration of the 2010 academic year, all schools will have adequate and sufficient sanitary services and sinks; the classrooms will have their roofs, windows, chairs and desks in good condition and these inputs will be properly maintained throughout the school year. With regard to the first goal, our results show that 4 percent of students in Junin currently enrolled in the third grade are in Group 0 ­meaning that they cannot recognize words and isolated sentences­ and 17 percent are in Level 3 ­meaning that they are able to meet the reading goals set for their grade. According to the results of the census test administered by the Ministry in December of 2006, by the end of second grade 18 percent of students were in Group 0 and 24 percent had reached Level 3. The first goal seeks to ensure that the majority of students that finish the second grade achieve the reading goals set for their grade and none of the students completes the grade without the ability to recognize isolated words or phrases. This goal is merely an example of learning goals that are based on the competences measured in the census for each grade (in 2007 third graders were also evaluate and the test included Mathematics). Ideally, goals should be set for each grade at least in the areas of Communication and Mathematics. Meanwhile, we must work with what we have. vii What can a Regional Government do to improve education? The Case of Junin About the second goal, according to our estimates the total investment in Preschool Education in Junin should achieve a net enrollment rate of at least 56 percent; yet, it stands at less than 42 percent of the children between the ages of 3 and 5. This goal focuses on ensuring that the expenditure on Preschool Education is efficient given that a 15 percent increase over three years will bring coverage to 57 percent. Nevertheless, this implies that a cost-efficiency study must be conducted on service provision for Preschool Education in Junin to identify the most efficient ways of allocating current resources before increasing spending at this level. This goal will be hard to reach without an awareness campaign to encourage parents to send their children to Preschool Education Centers (CEI) or Non-formal Preschool Education Centers (PRONOEI). The third goal has a double purpose. First, it focuses on ensuring that all of the students and teachers learn and teach in dignified and adequate conditions. The information gathered in the field visits suggests that the most deficient areas in schools pertain to the sanitary infrastructure (drinking water, toilets and sinks), classrooms' roofs and windows, and ­to a lesser degree­ students' chairs and desks. Second, the goal entails putting in place a monitoring system that works throughout the school year. This system should consist in a combination of self reports, perception surveys and random verification visits and should also articulate with a wider system for accountability on the funds delivered to schools for classroom maintenance (mandated by the 2008 Budget Law). This means that he Regional Government must also work to set minimum operating standards for schools and later ensure that these are met. This consists of a set of essential inputs, school resources and minimum conditions that the education community agrees are necessary to allow the school to offer students adequate learning conditions. This design should also include basic pedagogical inputs (books, materials for students), school furniture (desks, chairs and shelves), the physical characteristics of the education center (basic infrastructure; for example, the state of roofs and windows, square meters per student in each classroom) and sanitary and maintenance conditions in schools (classroom and school clean up). Offer support to schools to reach these goals Obviously goal setting is not enough; it is necessary to support schools with financial and pedagogical resources. In practice, schools have not received these resources from the UGEL, and we believe that alternatives to this system should be sought so that the UGEL can concentrate on the important task of managing personnel. Financial support · Each school should receive a budgeted amount per student. The proposal consists in increasing the average current annual non-salary expenditure per student in the Junin region and transferring this amount directly to each school. These funds will be allocated to goods, services and maintenance and will be distributed by applying transparent criteria that include compensatory mechanisms for the neediest schools in order to correct the inequalities found in the quality of infrastructure, furniture, maintenance and learning experiences. Peru has already made progress in this direction. According to the General Budget Law, 270 million soles will be transferred directly to schools to finance classroom maintenance during the 2008 fiscal year. Given this legal framework, the Regional Government should be required to redefine its role to ensure that the resources allocated to the schools are used in an efficient and pertinent manner. · Strengthen the capacities of school personnel in charge of the administration of funds and maintenance tasks. viii What can a Regional Government do to improve education? The Case of Junin Pedagogical support · Call on qualified amautas (teachers of teachers, in Quechua). Give schools resources so that they can hire specialists ­amautas­ that can provide support to reach reading and comprehension goals in second grade. A diverse supply of specialists should be sought to determine what work best (institutes, NGOs, universities, others). · Design and distribute support materials to achieve learning and comprehension goals. Conduct regular measurement of progress and inform the community The politicians and technical specialists responsible for education in Junin and the authors of this report share the conviction that it is possible to improve the quality of education. Thus, we suggest that Junin adopts instruments to help measure progress in all the aforementioned areas and facilitates the implementation of accountability systems based on active participation from stakeholders in a context of transparency, consensus and cooperation propitiated by the Regional Government. Such accountability system should contemplate delivering school-level information on goal achievement as well as information on the financial resources given to each school. We also suggest creating and adopting periodic reports on the education and financial results obtained by district, UGEL and the Regional Education Office for Junin to evaluate the degree of compliance with the Regional Government's priorities for education in the region. Chapter 2 Introduction At the beginning of 2007, the World Bank published the book titled Por una educación de calidad para el Perú (Towards quality education in Peru). The book was widely distributed and its main findings propitiated a series of debates. Shortly after, the World Bank received a letter from the President of the Regional Government of Junin requesting advice on how to improve education in his region. The book in question analyzes the problems affecting basic education and provides recommendations for the entire country. The President of Junin was interested in improving the quality of education in his region and was seeking concrete proposals regarding decisions that the Regional Government could make. Given the existing legal framework, the region's fiscal situation and the balance of political forces in the Education sector, the President of the Regional Government of Junin asked the following question: What can a Regional Government that is truly committed to education do? It was the first time that the World Bank had received a request of this kind, and following an internal discussion process, it decided to conduct a study. The conclusions of this study are presented in this document. This study intends to be useful not only to the Regional Government of Junin but also to other regional governments committed to education. 2.1. Education Decentralization The issue of education decentralization has been on the agenda in Peru for a long time but for different reasons, such as downsizing bureaucracy, promoting community participation, and fiscal constrictions. In July 1986, in the context of a National Program for Debureaucratization, the Ministry of Education (MED) proposed restructuring its Intermediate Bodies (OOII) by creating Education Service Units (USE) to replace the Education Zone Offices and the Education Supervisors (Supreme Decree 12-86/ED). The objective of this change seemed to be none other than facilitating administrative management in the zone offices. The USEs were defined as "deconcentrated bodies of the Ministry of Education responsible for achieving quality and efficiency in education services..." (Ministry resolution 105-89-ED, article 3). In the 1990's, Fujimori's government witnessed ­and cancelled­ two successive attempts to change the inefficient education administration system through municipalization: the first attempt was during Gloria Helfer's ministerial period and the second was with the creation of Municipal Education Councils (COMUNEDs) in 1992 and 1994. The second "municipalist" attempt was led by a desire to generate fiscal savings and legitimation. Its cancellation meant, during the central years of Fujimori's government, the end of the political "legitimacy" of a decentralization policy directed toward empowering democratically elected authorities. The regional governments that had been elected ­which were dissolved in the self coup of 1992­ were later replaced by the Transitory Councils for Regional Administration (CTAR). The latter were decentralized bodies of the Ministry of the Presidency with technical, budgetary and administrative autonomy to perform their functions. During this period, some of the Education sector's 2 What can a Regional Government do to improve education? The Case of Junin Table 2-1. Socio-economic characteristics in Junin Indicator Peru Junin Ranking Economic Context Per capita GDP2005 (S/.) 5,665 3,947 9 Per capita household income (S/.per month) 286 307 10 Total Poverty (%) 44.5 49.9 12 Extreme Poverty (%) 16.1 16.5 10 Geographic Distribution R. coast (% pop.) 52 0 R. sierra (% pop.) 35 80 R. rainforest (% pop.) 13 20 Source: INEI, ENAHO 2006 ; for family income: UNDP 2006b. Note: For more details, see annex 1.3. For a definition of variables, see annex 1.4 and for a summary of indicators, see annex 1.5 administrative functions were delegated to the Regional Education Offices (DRE), which were discredited and inefficient and were mapped administratively to the CTARs. Once democracy was restored, different laws were passed to regulate the decentralization and modernization process in the public sector. In 2002, Peru embarked on an ambitious regionalization process as part of a broad agenda for the democratization and modernization of the State, which was extremely weak and discredited. Between July and August of 2002, laws were approved for Regional Elections (27683), the Bases of Decentralization (27783), and Constitutional Reform regarding Decentralization (27680). The decentralization policy was based on the redistribution of state functions and resources to sub-national governments. This did not change significantly the intergovernmental relationships already in place, and although a degree of harmony was achieved between the sector and the CTARs, tensions began to mount when the regional governments were elected. In November of 2002, the first election of sub-national authorities was held, even though the institutional and legal framework for decentralization had not been completed. Disorganization was exacerbated when money began to flow directly from the Ministry of Economy and Finances (MEF) to the Regional Governments without going first through the Ministry of Education (Budget Line 010). In January of 2003, the Organic Law of Regional Governments (27867) was approved, followed by the Organic Law of Municipalities (27972). On July 28th of this same year, Congress approved the General Law of Education (28044), which incorporated a decentralized perspective on sector management. 2.2. Education in the context of Junin The region of Junin has close to 1,3 million inhabitants and is situated in central Peru to the east of Lima. Its 9 provinces and 123 districts cross through the natural regions of the sierra and rainforest as well as five ecological sections ranging from 400 to 6,000 meters above sea level. Although the majority of Junin's population lives in the sierra (80 percent), the proportion of the population that lives in rainforest areas is higher in this region than in any other in Peru. Junin has significant livestock, agricultural, mining, and touristic resources and its geographic location facilitates commercial flows to the capital. As such, Junin exceeds the average for household income in Peru and ranks tenth out of twenty four regions (PNUD, 2006). With regard to per capita GDP, the most recent available figures (2005) show that Junin holds the 9th of 24 places with an average of 3,947 soles a year, which is still below the national average. The parents' education levels exceeded expectations: less than 9 percent of adults are illiterate. Poverty levels in Junin are similar to national averages. Junin ranks 12th of 24, with 50 percent of the population living in poverty and 17 percent in extreme poverty. The incidence of poverty is greater among 3 What can a Regional Government do to improve education? The Case of Junin Table 2-2. "Rurality" of the population and Basic Regular Education (% rural) a Indicator Peru Junin Students enrolled 30.8 43.0 Preschool 32.3 50.0 Primary 37.2 50.2 Secondary 19.6 29.2 Teachers 27.2 38.0 Preschool 19.5 32.7 Primary 34.4 46.2 Secondary 21.0 30.3 Schools 55.7 65.5 Preschool 48.4 58.6 Primary 68.4 76.2 Secondary 38.8 46.0 Rural population b 35.0 43.0 aSource: MED-Unit of Education Statistics , School Census 2007. bSource: INEI, ENAHO 2006. children: 73.7 percent of Preschool age children live in poverty along with 64.9 percent of primary school age children and 65.6 percent of children between the ages of 12 and 16 (INEI, 2005a). Although the proportion of the rural population in terms of the total population (43 percent) in Junin is higher than the national average (35 percent), it is basically an urban population. The education system appears to correspond to demographic characteristics given that 57 percent of the students enrolled in Basic Regular Education live in urban areas. Two out of three primary schools are urban as are 54 percent of secondary schools and approximately half of Preschool education centers. Although considerable demographic dispersion exists in rural areas, the vast majority of Junin's population lives close to basic social services. According to ENCO 2006, 99 percent of Junin's inhabitants declare that they one hour or less from the closest state primary school and 75 percent live less than ten minutes away. Moreover, 91 percent live less than one hour from a state secondary school (see annex 1.1). Data analysis of other services such as medical posts and police stations shows similar results. This suggests that the social service infrastructure covers a large portion of the regional territory. These figures are encouraging and show that significant progress has been made in the geographical coverage of education services although it is important to keep in mind that the dispersion may be underestimated.1 There is sizable inequality in the region. Poverty and precarious living conditions are worse in rainforest areas, particularly in the provinces of Satipo and Chanchamayo. In Satipo, 86 percent of the population lives with at least one basic unsatisfied need (NBI), which contrasts with Huancayo (47 percent) and Yauli (38 percent) (INEI, 2007). The seven districts with the lowest household income in Junin are located in Satipo, that is, all the province's districts except its capital. A family in the district of Rio Tambo, Satipo, has an average per capita monthly income of 205 soles: almost half of what a family in the district of Yauli earns. 1 In the first place, the sample excludes the districts of Mazamari and Pangoa that- according to information gathered during fieldwork- have above average dispersion for the region. The National Census of Population and housing of 2005 (INEI, 2006b), the main source for the selection of the ENCO sample, did not cover the districts of Mazamari and Pangoa, both in the province of Satipo. These were the only districts that were not included in the census (INEI, 2005b) and, as a set, represent 5.1 percent of the population of the entire region (INEI, 2005b). In the second place, the variables for the accessibility of services in the ENCO 2006 are particularly susceptible to sample design effects and have a higher standard error than the rest of the variables. For Junin, the standard error for school enrollment and attendance represents barely 0.008 percent of the median. Equally, the sampling effect on the variables for accessibility to a Primary and Secondary School corresponds to 54.3 percent and 29.1 percent of the respective median while school enrollment and attendance is barely 3.1 percent a 3.0 percent (INEI, 2006a, p. 186). 4 What can a Regional Government do to improve education? The Case of Junin Today Junin is one of the regions with the best education results for young people: 78 percent of students between 18 and 25 have finished secondary education, which is far above the national average of 67 percent. Net enrollment rates for primary and secondary education are above the national averages, with 95 percent of children between 6 and 11 enrolled in Primary and 76 percent of adolescents between 12 and 16 enrolled in secondary. Compared to the rest of the country, a child in Junin has more possibilities of finishing primary and continuing on to secondary. In fact, Junin is the sixth region with the highest percentage of students that finish secondary school. Additionally, the incidence of illiteracy in Junin is less than 10 percent. 2.3. The legacy of violence in the education system The political violence in the region of Junin had a considerable impact in the Region, both in terms of the loss of human life and the weakening of public institutions and grassroots organizations. Junin is the second region ­after Ayacucho­ with the highest number of casualties between 1980 and 2000 due to political violence. Overall, the Commission for Truth and Reconciliation (CVR) received reports of 2,565 cases of death and disappearance, of which 2,102 (82 percent) occurred in rural areas. The provinces of Satipo and Chanchamayo suffered the greatest losses, with 1,482 cases reported. Additionally, Junin has been one of the regions with the highest rates of forced migration due to the armed conflict and also received refugees from neighboring regions such as Ayacucho, Huancavelica y Pasco (CVR, 2003). The presence of the Communist Party of Peru Shining Path (PCP-SL) and the Tupac Amaru Revolutionary Movement (MRTA) can be traced back to the early 1980s. The areas with the highest number of registered subversive actions were the provinces of Satipo and Chanchamayo (see annex 1.2), the districts of Comas, Cochas, Mariscal Castilla and Andamarca in Concepcion and the districts of Santo Domingo de Acobamba and Pariahuanca to the east of Huancayo. Active terrorist cells still exist in some of these areas and are generally associated with drug trafficking. Even worse, the coexistence of terrorism, drug trafficking and illegal timber exploitation has stimulated the evolution of mafias that obliterated the grassroots organizations and exacerbated the population's feelings of insecurity. Family and social disintegration due to armed conflict is considerable. In some of the native communities visited, the majority of families are made up of widows and orphans. This ineludibly affects children's education opportunities, given that they are forced to assume economic responsibilities at an early age and lack family support to complete school assignments. 2 In this context, native populations and colonos3 developed different forms of coexistence. Learning and teaching were tainted by mistrust, uncertainty and the reproduction of patterns of abuse and violence that translate, for example, into the frequent use of corporal punishment to discipline students, which teachers consider a valid means of reaching their education objectives. Many teachers indicated that family and social disintegration constitute one of the main problems that schools face. During fieldwork, we collected ample testimony from students regarding the physical and emotional punishments from their teachers. These children indicate that whips, sticks, rope frays or chalanka (tree branches), pulling and insults are typical means of exercising discipline. These "weapons" are kept in the students' sight to intimidate them. In one classroom, we saw a teacher giving a lesson with a brigadier's baton in one hand and a belt in the other as students sat timidly, evidently afraid of being called to the board. 2 The impact of violence was also evidenced in the reiterated plea for emotional support from teachers and families during the interviews held for this study, particularly when they found out that the surveyors were psychologists. In many cases, requests were made for workshops for parents or individual sessions, which unfortunately could not be met. 3 Inhabitants from the highlands that migrate to the rainforest area looking for better economic opportunities. 5 What can a Regional Government do to improve education? The Case of Junin The persistent use of violent disciplinary methods is also the result of the lack of information regarding the psychological consequences of these measures and ignorance of alternative methods to exercise authority and discipline. Some of the teachers who were interviewed indicated that they knew of no other way to discipline students. There were those who complained because parents "know" that they can file a complaint if their sons or daughters are abused. One teacher asked, with genuine concern, that we explain what harm was caused by corporal punishment because this was a new concept to him and he wanted to know why he wasn't allowed to hit his students anymore. This is evidence of misconceptions that transcend the classroom and the school and it has to do with flaws in initial and in-service training. Additionally, children's fantasies provide evidence of the marks left by armed conflict that are expressed, for example, in their artwork. In the school in Mazaroveni (Rio Tambo), the students carved machine guns out of wood to hang in the classrooms as decoration. Nevertheless, contrary to what one might think -and despite the fact that residual terrorists movements still exist and are allied with drug trafficking- Junin, as a region, has had one of the lowest incidences of social conflict over the past few years. According to the Ombudsman's Office, between January of 2006 and May of 2007, the region had only four latent conflicts, all associated to mining activities. This can be Classroom decoration in Mazaroveni, Satipo (Photo: Mariella Bazán) explained, among other reasons, by the weaknesses and limitations of the population's organizations. A recent study on the region found that there were less than eight hundred grassroots organizations (DESCO, 2007). It is important to point out that between May and June of 2007, during the time we conducted our fieldwork evaluation, the teacher's union had instigated a series of regional and national strikes to oppose the Law on Teaching Careers among other discrepancies. Some of these strikes had considerable turnouts. 2.4. Challenges for education in Junin This study analyzes and systematizes the financial and statistical information available,4 complementing it with data gathered in the fieldwork conducted between May and June of 2007. As part of this fieldwork, the team met with public officials from the Regional Government (including the president and the regional director of Education), as well as general directors, area directors and specialists from all the UGELs in the region, and mayors from the two districts in which pilot programs for education municipalization are underway. Additionally, we interviewed 50 school principals, 118 teachers, 181 parents and 1164 boys and girls in our random, stratified sample of 83 public primary schools in the region's provinces. In these schools, the reading performance of 1,687 Primary school students was evaluated (primarily from second, third and sixth grade), using an oral reading test and a silent reading exam.5 4 The main sources of information used in this report are the following: the School Census 2006 of the Education Statistics Unit (MED); the National Performance Evaluation 2004 of the Unit to Measure Quality (MED); the National Household Survey 2006 by INEI; and the Integrated System for Financial Administration- National, Regional and Local 2007 (last up-date) by MEF. 5 The difference between the number of students evaluated and those interviewed is explained by the fact that not all of the children who were evaluated were interviewed. Only those who were given the oral test were interviewed. The written test was administered collectively to all third grade students in the classroom chosen. The oral test was administered individually in a random sample among the selected classrooms (including second, third and the highest grade available). The sample includes 475 third grade students that took both the tests (written and oral), 686 students from second to sixth grade that took only the oral exam, and 526 students from the third grade that took only the written exam, for a total of 1,687 students evaluated. Additionally, three students were interviewed but for one reason or another they were not evaluated. In total, 1,164 students were interviewed (475 + 686 + 3). For more details on the sample design, see annex 3.2. 6 What can a Regional Government do to improve education? The Case of Junin The study identified three large areas that require attention: the resources invested in education, the learning results obtained, and the education management, which determines the relation between the first two. Next in this report, Chapter 2 provides more information on the challenges pending with regard to the quantity, quality and distribution of human and financial resources, comparing the situation in Junin with that of other regions in the country. We found that the problem is not that resources are scarce but that they are distributed inefficiently and inequitably given that no adequate mechanisms exist to ensure minimum quality levels. This sense, many of the problems facing schools in Junin could be substantially reduced --or even overcome-- by means of more efficient management and accountability.6 Chapter 3 explores the issue of education results in general, with specific attention paid to learning results. It was found that over the last few years, important progress has been made in terms of coverage and in general young people in Junin receive more years of education than the national average although the quality of the education is poor. It was not surprising to find that profound inequalities exist in the performance of students in single teacher and multi-grade schools, rural and urban schools, and between speakers of native languages and those whose mother tongue is Spanish. It was surprising however that not even schools that receive more and better resources were able to offer an education that allows students to reach minimum learning standards. In other words, low quality education is the norm. This points once again to management issues given that resources are invested without clear education goals, accountability mechanisms, quality control through periodic evaluation and access to information. Chapter 4 focuses on analyzing the biggest challenge: managing the education sector in Junin in the context of a decentralization process. It offers a comparison between what is stipulated in the legislative framework and what is practiced at different levels of education administration: regional, district/local, and school. It summarizes the regionalization process and the obstacles found in Junin, the municipalization pilot, the role that UGELs fulfill and progress towards strengthening school autonomy, which, according to the law, if the main objective of all education administration. Finally, based on the study's findings, Chapter 5 proposes adopting a management focus that grants more autonomy and resources to schools with an emphasis on measuring results, pedagogical and technical support to teachers and principals as well as accountability. In this new administrative framework, we suggest three measures that the Regional Government of Junin can take to improve basic education. First, set quantitative goals with regard to: learning in Primary, coverage in Preschool and operating conditions in schools. Second, to offer financial and pedagogical support to schools so that they can achieve these goals. Third, make regular measurements of progress and inform the community. 6 Education administration is understood as the organizational process geared towards ensuring continuous improvement in education quality by developing sub-processes such as the definition, planning, execution, follow-up, evaluation, and recognition of meaningful experiences with regard to education and institutional strengthening. We can identify three basic management issues: school leadership, academic management, and efficient administration of resources. All administrative efforts should contribute to ensuring the quality of institutions. From the perspective of school leadership, good management requires well functioning government bodies and the implementation of planning and evaluation models that allow institutions to make progress towards achieving their strategic goals. Academic management focuses on designing innovative and adequate curricular structures. Finally, efficient administration of resourcer requires taking into consideration everything that can contribute substantially to improving the physical and technological infrastructure, developing and managing human talent to serve schools, as well support systems for more efficient financing, budget allocation, and other issues key to improving institutional performance, such as accountability to citizens and the central level. Chapter 3 Education Resources In Junin, the education supply is elevated: there are 1,629 Preschool Education centers, 2,045 primary schools and 543 secondary schools that serve 42,114 students at the preschool level, 191,221 primary school students and 119,987 students in secondary respectively. These education centers employ 1,909 preschool teachers, 9,205 primary school teachers and 8,479 secondary school teachers. What are the operating conditions at these education centers? Is there an undersupply of teachers? What is their level of preparation and how are they distributed? Which financial resources are destined to education and how are they allocated? This chapter analyzes the availability and management of resources assigned to education in Junin, ranking this regionagainst all other regions in the country as a reference. 3.1. School Infrastructure and Equipment Two out of three of the school principals interviewed and three out of four teachers in single teacher schools indicate that the main problem at their institutions is infrastructure. The central problem in this area is no longer the lack of schools ­the majority of the population lives not more than ten minutes away from the nearest public school-, but the quality of education these institutions provide and the physical conditions they operate under. In our visits, we found that many classrooms are deteriorated due to lack of maintenance (broken windows, worn-down desks, bathrooms in ruins), which is primarily the result of problems with management. Poor management is also partly responsible for the under-utilization of existing sanitary infrastructure (e.g., schools with water and sewage connections but no toilets), and the lack of materials (incomplete libraries, absurd difficulties in the logistics to distribute cleaning materials, inadequate pedagogical materials). These problems affect rural communities and are also visible in urban schools. Access to basic services: electricity, water and sewage systems Junin is well above the national average in terms of electrification of schools: it ranks sixth; almost 80 percent of schools have electricity. But Junin lags behind other regions in terms of sanitary infrastructure: a mere 39 percent of schools in Junin have public sewage connection, compared to 45 percent at the national level. The schools' connection to potable water networks is also more limited than the national average but more than half have access to this service. 8 What can a Regional Government do to improve education? The Case of Junin Table 3-1. Status of school infrastructure Category Peru Junin Ranking Location relative to the population Population close to an Preschool (%)1 99.0 99.1 13 Population close to a Primary School (%)1 93.6 91.2 14 Population far from a Secondary school (%)2 0.82 0.90 14 Basic services Schools with all services (%) 42.4 37.2 11 Schools with electricity (%) 62.7 79.7 6 Schools with potable water (%) 62.6 56.8 12 Schools with sewagesystems (%) 45.2 38.8 12 Schools with no services (%) 26.3 16.4 7 Sanitary Services Schools with sanitary services (2004. %)3 87.5 89.8 12 Schools with toilets (%) 49 39 12 Schools with latrines (%) 29 39 10 Schools with septic tanks or other sanitary services (%) 9 12 6 Schools with sanitary services in use (2004. %) 87.1 89.4 12 Toilets in use (%) 97 99 4 Latrines in use (%) 96 98 5 Septic tanks in use (%) 96 98 11 Schools with toilet bowls (2006. %) 54.9 45.3 13 Maintenance Classrooms that do not require major repairs (%) 78.7 78.4 10 Sources: School Census 2004, School Census 2006, MED ­ Statistics Unit. 1An hour or less for travel time. 2More than four hours from the district capital. 3Data from 2004 because from 2005 only toilets are recorded. Note: For further details, see annex 2. 1. For the definition of variables, see annex 1.4. Inequality in the region runs high, as schools in the rainforest provinces have less access to basic services. In Satipo, for example, only one out of 617 schools has access to sewage systems and only 2 percent have running water while in Yauli, 83 percent of education centers have sewage connection and 92 percent have potable water. In rural or marginal urban areas, which have more access to the aforementioned services, we found that 10 percent of the schools in Junin (and 15 percent of the schools in Peru) lack at least one of these basic services despite the fact that they exist in the town in which the school is located. Silo at a school in Pomamanta (Photo: Luis Enrique Fodale) Quality and salubrity of sanitary services The principals, teachers and parents interviewed frequently referred to the poor condition of sanitary services: "There are no restrooms and there are only three sinks for two-hundred students," said the principal of a school in Curimarca, Jauja. Currently, more than 10 percent of schools in Junin and Peru do have any kind of sanitary service. Although Junin is ranked slightly above the national average in this area, the quality of sanitary services is low: the majority of these services are latrines and septic tanks and barely 39 percent of the schools have toilets (in comparison to 49 percent at a national level). According to the data gathered in the fieldwork survey, calculations indicate that only 19 percent of the primary schools in Junin have sanitary services that are connected to a sewage network. 9 What can a Regional Government do to improve education? The Case of Junin Interviewed subjects coincide that latrines are not the best alternative for children. In addition to the inadequate sanitary infrastructure, maintenance and cleaning of such sanitary facilities is poor: "Because the restrooms are latrines, they provide no safety for boys and girls," indicated a teacher at San Cristobal de Alto Cheni, in Satipo. A parent in Yanamarca claimed that "the sanitary services don't work, so children have to go to the bathroom on the patio." A teacher in Los Angeles Ubriki in Chanchamayo added that "the sanitary services are a source of infection: no one cleans them." In some of the schools visited, strong odors from poorly maintained silos (holes in the ground that are used and then covered) interfered with the normal progress of classes, creating sanitary risks. The best-case scenario, both in terms of salubrity and dignity, is for each school to have sanitary services with toilets that are connected to a sewage network and sinks which have potable water. The use of latrines or septic tanks is a lower quality alternative for areas where it is not possible to install water and sewage networks. Nevertheless, 17 percent of schools already have water and sewage installations and yet lack toilets. This figure almost doubles the proportion found in the rest of Peru (9 percent). In Chanchamayo, one out of every two schools with water and sewage installations has no sanitary services for its students; the percentages are around 40 percent in the provinces of Jauja and Concepcion. In Satipo, the situation is worse: the only school that declares having water and sewage connections has no toilets. In addition, another 4 percent of schools lack sewage connections despite the fact that the towns in which they are located have this service. This means that close to 21 percent of school facilities are under- utilizing the sanitation infrastructure that is already in place, and the students are the ones suffering the most. Roof of a classroom in San Cristóbal de Cheni (Photo: Andrés Figallo) Windows without glass or screens, Los Angeles de Ubiquirri (Photo: Andrés Figallo) Maintenance of infrastructure and school furniture In Junin, 22 percent of the classrooms require major repairs such as building walls, replacing floors or roofs or installing/replacing electrical connections. The region's performance in this area is similar to the national average: it ranks 10th in comparison to the rest. According to fieldwork results, we estimate that a third of primary schools in Junin (50 percent in the case of Satipo) have roofs with holes, leaks and occasional bat nests. Half of the classrooms that have windows have no glass or broken glass and 35 percent of schools have peeling walls that show evidence of filtrations and cracks. Practically all of the education centers that suffer from these problems are located in rural areas. Poor maintenance of school infrastructure is a frequent problem in the provinces of Satipo, Chanchamayo and Concepcion and in the districts of Santo Domingo de Acobamba and Pariahuanca in Huancayo. 10 What can a Regional Government do to improve education? The Case of Junin With regard to school furniture, we found that the problem lies more with maintenance than with provision. The proportion of students without desks is relatively low in general and in Junin: less than half of the national average (7 percent versus 15 percent). But the condition of the furniture is less encouraging. According to our fieldwork, 37 percent of schools have desks that are in poor condition and 36 percent have chairs that are deteriorated or falling apart. The situation would be different if these schools had an adequate maintenance system for infrastructure and furniture, run by individuals that are accountable for their actions and who have the funds to cover operating expenses. Available teaching materials When asked about the minimum conditions needed to do a good job, the first thing that most teachers mention is teaching materials and textbooks. About 72 percent of teachers in single teacher schools and 51 percent of all the teachers interviewed indicate that, in order to teach well, they need adequate textbooks and education materials. For example, a teacher in Pampa Americana indicated: "Government-issued textbooks leave out certain things because they are not consistent with this reality; there are no textbooks that are adapted to the rainforest." Rural teachers often mention the lack of appropriate materials for the students' reality as an obstacle to learning, given that lessons are based on foreign situations and objects. Teachers in the rainforest indicated that the textbooks available were a holdup for teaching and as such they were forced to look for more appropriate texts. Junin is below the national average at all three education levels with regard to school libraries. Only 26 percent of preschools, 35 percent of primary schools, and 53 percent of secondary schools have a library, , compared with 31 percent, 48 percent and 62 percent at a national level respectively. Junin is definitely among the regions with the lowest provision of school libraries at a primary and secondary level. Inside Junin, the differences are also notable. While in Satipo and Chanchamayo barely 6 percent of schools have a library, in the province of Junin this rate is 96 percent. Additionally, none of the Preschool or Secondary schools report having a library. Reading hut with no books, Satipo (Photo: Mariella Bazán) During fieldwork, we found that 62 percent of the primary schools visited have small classroom libraries, Table 3-2. Status of School Equipment Category Peru Junin Ranking Furniture Students with a desk (%) 92 93 9 Reading materials Preschool library (% IIEE) 31.2 26.3 14 Primary school library (% IIEE) 48.3 35.4 20 Secondary school library (% IIEE) 62.3 52.9 21 Information and Communication Technologies Student ratio per computer1 27 24 8 Internet connection1 15.7 10.0 13 Source: School Census 2006, MED ­ Statistics unit. IIEE: Education Institutions (schools). 1 Excludes Preschool Centers. Note: For more details, see annex 2.2. For a definition of variables, see annex 1.4. 11 What can a Regional Government do to improve education? The Case of Junin which in general are the result of combined efforts by teachers and families. Finally, access to computer labs and Internet is limited. Throughout the country, only 16 percent of primary and secondary schools have their own Internet connection while in Junin this ratio is below 10 percent. On average, the country has one computer for every twenty-six students; in Junin there is a computer for every twenty-four students although in Huancayo, the levels are similar to those in Callao and exceed Lima's regional average (9.1 students per computer). In contrast, in the province of Satipo, there are no computers at all for pedagogical purposes. Why is infrastructure so poorly managed? In theory, distributing goods and resources to guarantee the maintenance of school infrastructure is the UGEL's responsibility, but in practice this task has very low priority. In essence, this is just one of many responsibilities assigned to the UGEL. The UGEL acts mainly as a personnel management agency, administering payrolls and personnel flows according to the complex rules that govern these processes. Its other functions (providing pedagogical support to schools and administering goods and services) are conducted only if energy and funds are left over after completing personnel management tasks. No one is accountable for any other task because they are considered marginal. Second, the UGEL's responsibility in the area of maintenance is even more diluted if we consider that of the nine UGELs of the region, only two are "executing units" (meaning that they have MEF's authorization to administer funds); the other units have neither the volume nor the capacity necessary to receive this authorization and their functions are delegated to the Regional Office in Huancayo (which covers the jurisdictions of Huancayo, Concepcion, Jauja and Chupaca), the UGEL Tarma (which covers Tarma, Yauli and Junin) or the UGEL Satipo (which covers Satipo and Chanchamayo). The idea that the UGEL is in charge of distributing funds for goods and services is a myth created by the Education sector although MEF does not see it this way. Third, the schools are the most interested parties when it comes to maintenance but the responsibility lies, in theory, in the UGEL. Yet, the UGEL is not accountable to the school for this role and it is physically distant from the school site. In the school, no one is responsible for maintenance results. We also can't expect the teachers or students to clean the latrines because this is not- and should never be- their role. 3.2. Human resources in the schools in Junin7 Poor management is also evident in human resource administration. In Junin, there are an average of eighteen students per classroom in preschool and twenty-one in primary and secondary, which is below the range established by the norm.8 Nevertheless, there are significant differences within the region depending on the type of school. Although Junin is the sixth region with the lowest number of students per classroom in full grade schools, it ranks 16th in the case of single teacher schools where the average number of students per classroom exceeds the maximum number allowed. Junin is one of the regions with the highest rate of overcrowded classes in rural schools and has the most underpopulated urban classes simultaneously. There are also variations between provinces: in Satipo a primary school class has an average of twenty-nine students, compared to sixteen in Tarma. Consequently, the distribution of salary expenditure favors the 7 In this section, we will explore the main challenges in terms of quality, quantity and distribution of human resources. We have chosen not to study the labor aspect in depth given that it involves a number of extremely sensitive factors relative to political economy that make it a national issue. Instead, we have chosen to focus on issues that can be handled by the regional government in the short and medium term. 8 The referential values of class size stipulated in the norm on pedagogical orientations, which are used to determine the number of teaching spots that an education institution must cover, are twenty students per teacher for single teacher schools, thirty for multi-grade schools and 40 to 35 in full grade schools for an urban or rural school respectively. 12 What can a Regional Government do to improve education? The Case of Junin Table 3-3. Status of human resources Category Peru Junin Ranking Quantity Students per teacher 19 19 14 Preschool 18 18 10 Primary 21 21 10 Secondary 14 12 8 Average class size (students) 21 21 14 Preschool 19 19 10 Primary 22 22 15 Secondary 22 21 11 Distribution Teachers in the classroom (% teachers) 89.8 90.7 16 Overcrowded classes in rural schools (%)1 12.4 13.6 18 Overcrowded classes in multi-grade schools (%)1 2.6 2.1 12 Overcrowded classes in single teacher schools (%)1 22.9 22.6 15 Underpopulated classes in urban schools (%)2 40.5 49.0 23 Primary school teachers in multi-grade classrooms (%)3 43.6 50.3 13 Primary school students in multi-grade classrooms (%)3 39.9 47.1 13 Quality Teachers with a title (%) 84.4 93.3 12 Teachers evaluated in the last census (%) 50 48 19 Source: School Census 2006, MED ­ Statistics Unit. 140 percent more than stipulated in the norm. 250 percent less than stipulated in the norm. 3Includes single teacher schools. Note: For more details, see Annex 2.2. For the definition of variables, see Annex 1.4 wealthiest and most urban provinces and districts. This demonstrates that the problem with human resources has more to do with distribution than scarcity. The inefficient and inequitable distribution of human resources In Junin, half of the primary school teachers are in charge of classes where two or more grades are taught simultaneously. Multi-grade classes respond to the fact that the school-age population in remote areas is limited and having a teacher for each grade is unjustifiable. Nevertheless, this adversely affects student learning and generates an excessive workload for teachers in multi-grade classrooms. This is even more evident in single teacher schools that attend to six grades in one class. Currently, 47 percent of primary school students in Junin attend multi-grade or single teacher schools. Fieldwork shows that the teachers at these schools are overworked and many do not have the time to do everything they need: they must make up lesson plans for various grades simultaneously and handle students with different learning needs during class time. This means that teachers rarely teach in-depth on any topic and are forced to use the short time available to resolve students' individual questions one-by- one or focus on specific groups. In practice, we found less effective time use in these classrooms, given that the teacher is either implementing or supervising activities with a group of students while others wait idly until the teacher finishes to receive their next assignment. Teachers at these schools express frustration because they feel that they do not dedicate enough time to their students. They often ask that at least one more teacher be brought on, or express a desire to work at another location. In many cases, this situation is aggravated by the profound emotional needs of students in remote areas, whose parents are either absent or violent, and many of whom were orphaned by the armed conflict. The study has found that the workload and emotional demands from students in multi-grade classrooms further complicate teaching and learning processes. 13 What can a Regional Government do to improve education? The Case of Junin Currently, 22 percent of single teacher schools (and in Satipo 50 percent) are overcrowded, meaning they have enough students to divide the classroom at least in two and be multi-grade schools (Table 3-3). This is a matter of management efficiency and equity rather than a lack of resources, as evidenced by the fact that 49 percent of urban school classes have less than half the students allowed by norm. Even in Satipo ­where human resources are scarcer- the distribution could be much more equitable: One out of four rural classes is overcrowded and one out of four urban classes is underpopulated. In other words, if two overpopulated urban classes were merged and one overpopulated rural class were to be divided in two, it would result in a more equitable assignment of these three teachers. Inefficiency in human resource distribution is unacceptable because it produces a negative impact on learning. Of course, the problem is more complex given that teachers prefer to work in a full grade urban school rather than single teacher rural schools (primarily due to the location, work conditions and workload) and use an infinite number of strategies to achieve this objective. One common strategy is permutation, which consists of two teachers who mutually agree to "swap jobs." Teachers who wish to interchange jobs post an ad at the UGEL describing the teaching position they want to swap with. According to testimony gathered by several UGEL officials, some teachers working in remote areas use permutations to "buy" a post from teachers who are close to retirement and work in less remote areas. The latter agrees to work in a school in a remote area for the months prior to his or her retirement in exchange for money and the former has the right to stay on at the more centric school once retirement takes place. Another example involves demands made by UGEL employees for sexual favors from teachers in exchange for particular assignments. Some recently appointed directors at UGEL indignantly reported that they had detected cases where young teachers -preferably married and with children- were assigned to very remote schools on purpose, to subsequently be asked for sexual favors in return for a transfer to schools that were closer by. As a result of the rationalization process in rural schools, some of the individuals interviewed also mentioned cases of teachers at small schools who encouraged students to drop out in order to subsequently request that the school be closed due to lack of students and obtain an assignment at a less remote school (since by norm they cannot be assigned to a more remote school). Good human resource management should create conditions under which the best and brightest agree to work in single teacher and multi- grade schools; these incentives include providing transportation and better housing conditions, alternative work schedules that allow teachers to spend more time with their families, additional economic compensation, pedagogical training for multi-grade classrooms and constant accompaniment. Difficulties in evaluating teacher quality Information regarding the professional quality of teachers is still very incomplete and limited. In Junin, only 48 percent of the teachers agreed to participate in the teachers' census evaluation in 2007.9 It is possible that with the new Teacher Career Law some progress in this area may be made. For now, the only information available on teacher quality refers to the proportion of teachers with titles although in truth, without an accreditation system that guarantees the quality of the training offered in Post-secondary Pedagogical Institutes (ISP) and Schools of Education at univerities, a title is not a reliable indicator. In any case, we found a higher proportion of teachers with titles in Junin than in Peru in general (93 percent versus 84 percent) and considerable progress has been made in hiring teachers with titles. There is no reliable information either on teachers' attendance and punctuality. Although principals must report unjustified absences -which are deducted from the teacher's pay- many opt not to do so in order to 9 On January 8 of 2007, the Ministry of Education conducted an evaluation of all teachers working in the public education system. The exam evaluated mastery of specific knowledge, ability to manage the curriculum, knowledge of their students' physical, psychological and social development, along with reading comprehension, the ability to analyze selected texts, and logical reasoning. This initiative met with strong opposition from organized teachers' groups that urged teachers not to show up for the evaluation. 14 What can a Regional Government do to improve education? The Case of Junin impose a sanction of their choice (rarely monetary). Some of the principals interviewed were of the opinion that a salary cut generates more conflict inside the school and prefer to handle teacher absences internally. In any case, rural schools tend to have fewer school days and effective hours of class time than urban schools. During our research, the probabilities of finding a rural school that holds classes on Fridays was very low10 because teachers travel to the city for the weekend and sometimes leave beforehand. The UGEL informed us that specialists do not make visits on Mondays or Fridays because these days are reserved to work with principals and teachers who come to the UGEL to get help with paperwork. This means that local management also lacks clear mechanisms to ensure compliance with class hours. Moreover, its office hours force teachers and principals from rural schools to skip class hours to take care of paperwork because the UGEL is only open from Monday to Friday and transportation is scarce. According to the responses gathered during our research, parents, teachers and school principals agree that for a teacher evaluation system to be pertinent, it must be based on students' learning achievements, measured against prior performance. A second area of evaluation is related to teachers' compliance with responsibilities such as attendance, punctuality and lesson preparation. Finally, other indicators that should be evaluated on a regular basis, although the challenge is greater, are classroom processes and dynamics, in particular with regard to physical and psychological abuse of students. In the opinion of school principals, good teaching performance means that most students meet the objectives planned for the year, understand the material taught, are able to interact properly in the classroom, and have the necessary skills to be promoted to the next year. In the words of the parents, a teacher is good at his or her job when the "children learn," when "they teach well and don't abuse students" and "make sure that everyone learns." Teachers, principals and parents agree that good performance requires the teacher to be punctual and not miss classes without justification, which is a frequent problem in the rural areas of Junin. A parent in Perene, Chanchamayo said that bad teachers are those who "miss an entire week of classes to go get paid;with the excuse that the offices are far away, they don't come in or arrive late to school." A mother in Cerrito de la Libertad, Huancayo, added: "They give assignments to the kids and go to lie out in the sun." Another mother in Huancayo added that the teachers are not good "when they are irresponsible, arrive at the time they please, and ask for bribes: Your child is not doing well in school, give me twenty soles and I'll give him classes but if we don't give anything, they give him a bad grade and that's it." Inadequate working conditions If teachers are to perform well in the classroom, in addition to improving early education, it is necessary to offer them the basic conditions to do their jobs. When questioned about these conditions, half of all teachers asked for good teaching material and textbooks. Additionally, 27 percent of teachers and 40 percent of principals indicate that ongoing teacher training is a requirement for good classroom performance. In their opinion it is also important for schools to have better infrastructure and equipment. While some teachers from rural schools request "appropriate furniture, desk materials and cleaning products, their peers in urban schools already have basic infrastructure and ask that "every child has access to a computer and adequate classroom conditions." One out of every five teachers interviewed indicated that, in order to do a good job, they needed more support from parents. Without a doubt, in the opinion of the majority of teachers interviewed, the main problem in their schools is the lack of support from parents ("Everything is left to the teacher and they wash 10For example, during the first few weeks of fieldwork, we encountered the most difficulties (school closings, teacher absenteeism, local celebrations and sporting events, etc.) on Friday May 25th (five out of eight teams could not complete their evaluations) and Friday June 1st (five out of eight teams reported difficulties). Problems were also recorded on Wednesday May 23rd (four teams) due to a strike called by SUTEP. 15 What can a Regional Government do to improve education? The Case of Junin Figure 3-1. What are the minimum work conditions necessary for you to be able to do your job well at this school?* Maestros (N=118) Directores (N=50) Libros de texto, materiales educativos 50,8% 32% Capacitación 27,1% 40% Infraestructura y equipamiento 23,7% 40% Apoyo de los padres 22,0% Otro 12,7% 26% Remuneración 8,5%14% Más personal docente 5,9% 24% Alimentación a los niños 4,2% Transporte 3,4% 8% Asesoría, acompañamiento 2,5% 14% Vivienda 1,7% 6% Clima laboral 1,7% 10% 0% 10% 20% 30% 40% 50% 60% * Teachers answer this question based on the conditions that are lacking. their hands of it."), their low expectations regarding their children's education ("They are happy if their children can read and write."), their low education level ("the majority of people here are illiterate, they don't know how to help their children") and scarce resources ("We ask the students for materials and the next day only two or three have brought them"). We found that 44 percent of teachers in general and 58 percent of single teacher schools in particular would rather work somewhere else, which indicates a significant level of job dissatisfaction. The location of the school is the main reason for such discontent, particularly among teachers in multi-grade schools and single teacher schools who indicate that access is difficult and costly, which means more time away from their families, fewer opportunities for professional development, more difficulties with parents, and more challenges with the students. A teacher told us that the first time that she traveled to the school she cried because of the long way, and decided she would stay there for two years at most, after which she would even pay to be reassigned, but seven years have passed and she is still there. Teachers assigned to rural and remote schools must make many personal sacrifices: "I would like to be closer to my family, it's been 18 years," said one teacher. "I am a mother ­remarks another teacher­ and I leave my kids, I leave them for a week." Another teacher said: "I have a family outside of this community and I cannot communicate with them, I am constantly worried about my children." This is a clear indication that teachers that work in single teacher and multi-grade schools need to be offered better working conditions so that they feel motivated by and satisfied with their work. 3.3. Financing education Peru spends close to 20 percent of total public expenditure on Education and Culture. This expenditure is executed through three budget centers: the Ministry of Education, regional governments and the National Institute of Education and Health. In all regions, expenditure on Education and Culture represents more 16 What can a Regional Government do to improve education? The Case of Junin than half of the regional budget but Junin is among those that spend the least in this area. In 2006, Junin spent 59 percent (352 million soles) of its regional budget on Education, compared to the 86 percent that all the regions spent on average (Junin ranks 19th). Additionally, expenditure on education per student in Junin represents 22 percent of per capita GDP, which exceeds the average of all the country's regions 17 percent. Nevertheless, Junin spends 60 soles less per student than in the rest of the country: it ranks 12th with regard to per student expenditure, lower than spending in regions with higher poverty indices and lower per capita GDP, such as Puno and Ayacucho.Junin spends 659 soles per student at the preschool level (ranks 14 in the country), 775 soles at the primary education level (ranks 17) and 1,147 soles for secondary education (ranks 14). In Junin, 77 percent of expenditure of Education and Culture is assigned to Basic Regular Education (EBR) and the main executing agency is the Regional Government (99.5 percent of funds allocated to EBR). Of the education budget, 8 percent is spent on preschool education, 49 percent on primary education and 43 percent on secondary (Table 3-5). This follows patterns similar to those found in other regions although resource allocation for preschool education in Junin is notably lower than in other regions (ranks 23). Almost all salary expenditure in Junin is financed with ordinary resources (99.8 percent) and practically all of the ordinary resources are allotted to personnel expenses (94.4 percent). Of the 24.5 million soles assigned to non-salary expenditure in 2006, 14 million were directly collected by executing units (Table 3-4). In Education, all amounts financed through donations and transfers are allocated to primary education (547 thousand soles approximately); this corresponds to only 11 percent of the donations and transfers received by the Regional Government of Junin. Table 3-5. Financial Status of Education (2006)* Category Peru Junin Ranking Quantity Total expenditure on Education and Culture (% of regional budget) 86 59 19 Per student expenditure (% of regional Per capita GDP) 17 22 13 Per student expenditure (S/.) 943 885 12 Preschool 705 659 14 Primary 836 775 17 Secondary 1.236 1.147 14 Distribution Expenditure executed per item Non-salary expenditure per capita (S/.) 88,4 18,8 22 capital expenditure per capita (S/.) 32,3 3,4 16 Expenditure on Basic Regular Education by level (% total) Preschool 11 8 23 Primary 47 49 14 Secondary 42 43 11 Note: MEF - SIAF. *Includes national and regional sources. Note: For more details, see annex 2.4. For a definition of variables, see annex 1.4. Table 3-4. Expenditure executed in the region of Junin on Education or Culture by financing source and expenditure type (new soles 2006) Other Investments and Goods and current other capital Personal and Financing Source services expenditures expenditures social obligations Total per source Canon and other royalties 257,442 0 1,607,971 0 1,865,413 Donations and transfers 546,750 0 3,770,096 0 4,316,846 Direct revenues* 10,697,330 3,254,406 1,527,293 513,854 15,992,882 Ordinary resources 7,993,228 1,737,944 8,709,843 311,560,876 330,001,891 Total 19,494,750 4,992,350 15,615,202 312,074,729 352,177,032 Source: SIAF - MEF. * Resources collected by each executing unit: universities, institutes, Regional Government, among others. 17 What can a Regional Government do to improve education? The Case of Junin Table 3-6. Per student expenditure in the period 2000-2006 (new soles at prices in the year 2000)* Preschool Primary Secondary Year Junin Peru Junin Peru Junin Peru 2000 358 481 449 524 685 771 2002 440 531 517 540 764 830 2004 455 577 576 655 869 995 2006 587 627 690 744 1,021 1,100 % annual average growth** 8.6 4.5 7.4 6.0 6.9 6.1 Source: SIAF and MED. * This calculation includes the students enrolled in formal and non-formal preschool, primary and secondary education centers for minors. ** Using the CAGR formula (Compound Annual Growth Rate). In addition to the resources that the Central Government transfers to the regional government, some funds for education come from and are controlled by local instances. In 2006, around 17 million soles were spent through local governments (provincial and district municipalities), which represent approximately 5 percent of the total of resources allocated to Education and Culture in Junin by the Central and Regional Government. Of this expenditure, 96 percent is allocated to investments and is financed mainly by the Municipal Compensation Fund (FONCOMUN), resources from the canon and other royalties (sobrecanon), resources that are collected directly such as municipal license fees, among others. With regard to the canon and "sobrecanon," of a total of 7,115,034 soles executed in 2006 in Junin, 26 percent was destined to Education and Culture (1,865,413 soles). However, 80 percent of this amount was allocated to Higher Education and the remaining 20 percent (390,949 soles) was used for education infrastructure purposes; two thirds of this amount was destined to goods and services. At the end of 2006, Peru had executed 95 percent of its budget for investments while Junin only reached 85 percent. In contrast, regions such as Lima, Pasco, Piura, Ucayali and Tacna spent above the budget amounts given that they received increases to the initial budget from canon resources. In Ucayali, the government spent six times the amount initially budgeted for investments in EBR programs. Other regions such as Ancash, Huancavelica and Loreto also had high percentages of budget execution and in Loreto additional expenditures to the initial budget were financed with resources directly collected by the Regional Government, which decided to assign them to investment in Education. In the particular case of Junin, under-execution was due primarily to the amount that INFES did not manage to spend. Now more is spent but not well spent Spending on Education11 in Junin, similar to national trends, has experienced substantial increases over the last six years at all three EBR education levels. Nevertheless, growth in expenditure does not correlate with an increase in the number of students enrolled. Without a doubt, total student enrollment has gone downcompared to the year 2000, and this decrease is more pronounced in Junin than in the rest of the country (see annex 2.5). This apparent discrepancy between the evolution of education expenditure and the reduction in the school population is in part due to the fact that the average per student expenditure in Junin in the year 2000 was significantly less than the national average. In preschool education, for example, Junin spent 358 soles per student while the national average was 481 soles, meaning 25 percent less. In primary education, this difference was 14 percent and in secondary education, it was 11 percent. Compensating for these gaps, per student expenditure in Junin grew at a faster rate than the national average, particularly at the preschool level, where it increased by 8.6 percent a year on average (at year 2000 prices ) compared to 4.5 percent 11Included within Education and Culture budget line in the functional program classifier of the Integrated System of Financial Administration (SIAF). 18 What can a Regional Government do to improve education? The Case of Junin Table 3-7. Expenditure executed on Education and Culture by item (new soles at prices in the year 2000) Junin Peru Annual average growth (%) Item 2000 2006 2000 2006 Junin Perú Compensation and salaries 194,268,152 277,795,054 3,875,095,524 5,792,140,018 6.1 6.9 Goods and services 14,365,678 17,353,360 769,223,789 941,652,117 3.2 3.4 Other current expenses 3,619,891 4,443,969 110,369,944 181,757,022 3.5 8.7 Investments** 4,198,614 10,901,025 527,186,228 562,667,886 17.2 1.1 Financial investments** 0 57,290 13,015,401 13,970,234 43.8* 1.2 Other capital expenditure** 1,416,700 2,941,643 61,573,458 164,184,938 12.9 17.8 Amortization 277,804 141,234 ­10.7 Interests and debt charges 768,942 3,304 ­59.7 Total 217,869,035 313,492,342 5,357,511,091 7,656,516,754 6.3 6.1 Source: SIAF. * Since 2002. ** Refers to Central and Regional Government investment; excludes municipal sources. on a national scale (Table 3-6). This allowed the per student expenditure in Junin to come significantly closer to the national average. In 2006, the gap had been reduced to 6 or 7 percent at all education levels. This shows considerable progress but in light of the region's income, one would expect education expenditure to reach or exceed the national average. Spending on Education and Culture as a percentage of the regional Gross Domestic Product (GDP) also grew from 5.1 percent in the year 2000 to 7.6 percent, which is the opposite of what occurred with the country's average expenditure, which decreased from 2.9 percent of the national GDP in the year 2000 to 2.8 percent in the year 2006. Per student spending as a percentage of per capita GDP experienced a slight increase between 2000 and 2002 (1 percent) but from this point on it has maintained more or less the same level. In 2006, the per student expenditure in preschool education in Junin was the equivalent of 6 percent of the regional per capita GDP; primary education was 7 percent and secondary education was 10 percent, which were below the national averages in all cases (see annex 2.6). The main source of budget growth in the sector was due to the salary increases over the past few years. Expenditure on compensation and salaries has grown sustainably in real terms over time at a rate of 6 percent per year on average since 2000, while expenditures on goods and services have grown barely 3 percent per year (Table 3-7). In 2006, expenditure on salary and compensation on Junin exceeded 88 percent of the region's total education expenditure, which exceeded the national average of 76 percent. It should be noted that, according to the World Bank (Banco Mundial, 2006b), this percentage should be in the 65-70 percent range. It is also evident that the expenditure on goods and services in Junin represents around 6 percent of the total expenditure for the 2000-2006 period, while in Peru it was around 13 percent. In any case, this figure should be between 19 to 24 percent (Banco Mundial, 2006b). This means that Junin assigns a considerably less to goods and services than what would be needed to ensure a quality service. Without a doubt, Junin is one of the four regions that spend the least on goods and services per student. Of the 885 soles that Junin spends on students on average per year, only 15 soles are assigned to non-salary current expenditures while Huancavelica spends 33 soles, Ayacucho 30 soles, Pasco 27 soles and Madre de Dios 67 soles. Non-salary current expenditure ­in particular in the area of goods and services­ includes payments for basic services (e.g. electricity, water), education material for classroom use and cleaning supplies, among others. As such, it is indispensable to allocate a greater proportion of the budget to this item to ensure that education centers enjoy minimum operating conditions: are clean restrooms in working order, no electricity cuts due to lack of payment, windows with glass, roofs that are not leaky or bat-infested, 19 What can a Regional Government do to improve education? The Case of Junin Table 3-8. Current expenditure executed on Basic Regular Education in Junin by province (2006) Average per Average Average student expenditure on Class expenditure on Household expenditure compensation size (300, per capita Population per capita on goods and and salary per 301, investments density income(S/. services (300, student (300, 302)* (Regional (hab/km2) ) 301, 302)* 301, 302)* Government, S/.)** Huancayo 19.96 883.60 18 9.30 126 3,885.86 Concepcion 6.60 826.40 21 0.96 20 3,541.10 Chanchamayo 4.94 759.53 23 5.99 32 3,352.80 Jauja 5.58 974.90 20 10.19 27 3,720.76 Junin 7.28 958.47 18 17.75 14 3,380.29 Satipo 14.96 718.71 27 2.35 7 2,925.18 Tarma 15.87 924.12 19 0.94 38 3,728.57 Yauli 6.75 944.31 19 9.61 14 4,539.24 Chupaca 6.27 1,111.55 19 3.64 45 3,560.42 Total Junin 12.94 865.80 21 6.84 29 3,679.46 Coef. Correlation/Household Per Cap. Inc. ­0.06 0.45 ­0.67 0.22 Coef. Partial correlation (controlled by 0.21 population density) / Per.capita house.inc ­0.26 0.43 ­0.65 Sources: MEF ­ National Office for Programming and Budgeting; UNDP 2006b; INEI. * Executed by executing units 300 (DRE Junin), 301 (UGEL Tarma) and 302 (UGEL Satipo). Total current expenditure on programs for preschool, primary and secondary education. ** Executed by the Regional Government. brooms and chalks available, etc. In light of this, it is not surprising that many schools in Junin operate in appalling conditions. In addition, the proportion of expenditure assigned to capital investments in Junin is ­despite a significant increase over the last few years­ very low in comparison with other regions. In 2006, Junin invested 3 soles per student on capital expenditure while Huancavelica invested twice this amount and Ayacucho ten times more. Current expenditure on education in Junin is not equitable We found that the provinces with higher average household income have higher public expenditure on education; the same trend is true at the district level. As is evident in Table 3-8, there is a partial correlation between per student expenditure for goods and services and the average socioeconomic status of each province (measured by household per capita income),12 although it is low. Results are stronger for salary expenditure. Figures consistently indicate that distribution is inequitable and favors provinces with higher family income. The statistical analysis of salary expenditure per student as well as class size and pupil-teacher ratio (which are measures of the use of human resources), show absolute and partial correlations of significant magnitude. These correlations suggest that the averages for salary expenditure per student are higher in provinces with higher income (correlations close to 0.5) and that these provinces assign fewer students to each classroom. At the district level, we find a positive relation between the average household per capita income and salary expenditure per student. All this indicates an inequitable distribution of salary expenditure, which is the main component of education 12Table 3-8 only incorporates expenditure by the three main executing units in the region of Junin for Preschool, Primary Education and Secondary Education programs. The data assigned in the previous tables also consider the expenditure executed by other programs, such as Higher Learning and Governmental Planning. 20 What can a Regional Government do to improve education? The Case of Junin expenditure in Junin. Likewise, capital expenditure is higher in provinces with higher family income, although the differences are not very pronounced. Beyond demonstrating very low per capita investment figures, there is a positive correlation between capital expenditure and the average economic status at the provincial level, which suggests that resource allocation is not particularly equitable. Nevertheless, the correlations found were small, which means that the results are not conclusive. This is also the case for the investment from local governments: the correlation is almost inexistent. Non-salary expenditure is insufficient and is not distributed based on homogenous criteria A significant limitation to obtaining better results in the education system is the inadequate distribution of non-salary resources to schools. First, the amount allocated to each school is too small, and second, the criteria for distributing these funds across schools are neither homogenous nor precise. As is evident in Table 3-8, non-salary current expenditure per student is around 13 soles, which represents much less that the 1 percent of annual per capita household income in the region. Although this figure is aggregated at the regional level, it constitutes an initial indicator that the amounts executed in this area are too low and are insufficient to cover the basic requirements necessary to offer a quality education. Figure 3-2. Current expenditure on education in Junin: fiscal flows Executing Units (UE) DRE Junín UGEL Tarma UGEL Satipo UGEL UGEL Huancayo UGEL Yauli UGEL (non-UE) Concepción Chanchamayo Junín Chupaca Pricing Units (UC) UC1 UC2 UC3 Schools (IE) IE1 IE2 IE3 IE4 With regard to budget allocation for schools, the executing units (UE) are the ones responsible for distributing resources across schools in their jurisdiction. The allocation of resources to each UE is determined by MEF according to the current budget. It is also important to note that each UE has its own criteria to decide how to distribute resources across schools. In the case of Junin, there are three UEs to which current expenditure budget entries are assigned: UE 300 (Education Junin), UE 301 (Education Tarma) and UE 302 (Education Satipo). The first operates in the Regional Education Office (DRE), which is the Regional Government's administrative body for education. The other two operate in the UGELs of Tarma and Satipo, which depend on the DRE. Each UE manages a set of pricing units (UC) that represent a number of schools from a specific education level or program, and may be from the same geographic jurisdiction (for example, a complete district) or various districts that are geographically proximate (Figure 3-2). It is illustrative to note, for example, that one UGEL in the region was assigned 1,480,000 soles for salaries. This same UGEL received 8,000 soles for electricity and 156 soles for other goods and services needed for schools to operate throughout the year. The figures for electricity and salaries show that it is very probable that this UC administers schools with large student populations; nevertheless, the budget allocation for classroom materials, cleaning items, and so on, are not very reasonable. There are similar cases at the preschool and secondary school levels. For example, while one UC for preschools has 58,000 soles for salaries , it only assigns 213 soles for goods and services, excluding payments for basic services 21 What can a Regional Government do to improve education? The Case of Junin (electricity or water bills). A UC for secondary education is allocated 293,000 soles for salaries and only 145 soles for goods and services. These examples show that the expenditure on goods and services is, proportionally speaking, very low. In order to make the aforementioned budget allocation, each UGEL applies different criteria and even worse, there are no mechanisms in place to compensate the schools that need more resources. One of the methods used consists of distributing equal amounts to each school regardless of the number of students or the fact that institutions with large student bodies require more inputs. Another method observed involves per student allocations that do not take into account fixed costs of small, rural schools. These schools end up receiving very small amounts despite the fact that basic infrastructure must be maintained regardless of the number of students enrolled. In both systems, the goods and services entry is distributed once the expenditure for goods and services (water and electiricity) has been taken off. No compensations are considered for characteristics associated with higher education costs such as the geographic location of the school or the population's poverty level. Budget allocation problems affect both ordinary resources and the funds that come from the canon and "sobrecanon." Only 300,000 soles (around 5 percent of the total of canon and sobrecanonresources in the region) were allocated to expenditure in goods and services in EBR, and were assigned in full to the Program for Maintenance and Repairs at Teaching Units. Nevertheless, the amount budgeted was not executed on this activity but was instead divided between two additional activities: maintenance of the school 30509 Leonor Ordóñez-Jauja (11,458 soles) and the Maintenance Plan for Emblematic Education Institutions in the region (245,984 soles). This means that all the canon and sobrecanon resources executed in the EBR were assigned to the "best" public schools in the region. Something similar occurred with the 133,000 soles executed for capital expenditures: 44,000 soles were used on classrooms, workshops and laboratories (computers, chemistry and biology) in just one school and 32,000 on replacing a classroom in another school. The rest was used to finish classrooms and workshops in four more institutions. Management problems are not evident when comparisons are made with other regions, except in preschool education There is a positive correlation between per student expenditure and net coverage, which indicates that the regions that spend the most per student have better results for coverage. Although this relation does not assume direct causality between the investment made and the result obtained, we can expect to find certain enrollment rate for each region based on its expenditure per student. If the anticipated result is achieved with the expenditure made, it is a sign of efficient spending. In this regard, the region of Junin shows different results for each education level:13 expenditure is more efficient than the regional average for secondary education but it is among the most inefficient in preschool education, given that the net enrollment rate is well below the estimated projection based on its expenditure level.14 According to estimates, with the current investment in preschool education in Junin enrollment rates should reach at least 56 percent (Figure 3-3). 13The enrollment rates analyzed correspond to preschool and secondary education levels given that they show greater variability than the primary education level. All of the regions show net enrollment rates in primary education that exceed 85 percent since the year 2003 (source: INEI, ENAHO). 14It could be argued that expenditure efficiency with regard to enrollment in Preschool Education depends in part on the proportion of students enrolled in non-formal programs (PRONOEI) given that these programs cost significantly less and as such they could increase efficiency in terms of coverage but not necessarily quality. Nevertheless, the percentage of preschool students in Junin that attend PRONOEIs is close to the national average (28 percent versus 26 percent). 22 What can a Regional Government do to improve education? The Case of Junin Figure 3-3. Results for education coverage in Peru by region 100% 90% Junín et 80% Ratn 70% mell ronEt 60% Ne 50% 40% Junín 30% 0 500 1000 1500 2000 2500 Expenditure Per Student (nuevos soles - 2006) Educación inicial Educación secundaria Log. (Educación inicial) Log. (Educación secundaria) Source: INEI, ENAHO 2006. Figure 3-4. Education achievements in Peru by region 600 Junín 500 n iot icanu 400 m moc 300 nislt Junín suert 200 esT 100 0 0 500 1000 1500 2000 2500 Expenditure Per Student (nuevos soles - 2006) 2do. de ed. primaria 3ro. de ed. secundaria Log. (2do. de ed. primaria) Log. (3ro. de ed. secundaria) Source: INEI, ENAHO 2006. 23 What can a Regional Government do to improve education? The Case of Junin Similarly, if we compare per student expenditure with learning outcomes in each region, we also find a positive relation. Both in primary and secondary education,15 the region of Junin demonstrated better than expected performance according to the national estimate (Figure 3-4). This means that the education expenditure in Junin is more efficient that in other regions, with the exception of preschool education. Expenditure efficiency varies between schools with similar characteristics Despite relatively positive results for Junin, we analyzed the efficiency of schools in the region. For this purpose, we used the sample of 81 rural schools considered in this study and evaluated the quality of student learning through tests for fluency and reading comprehension. The analysis used a Stochastic Frontier Model to estimate a frontier for education production.16 The efficiency or dependent indicator in the model was the number of words that a student could read correctly per minute. The control variables incorporated primarily socio-economic and geographic characteristics and took into account education input provision and the teaching staff's experience. (Figure 3-5). Figure 3-5. Efficiency-achievement matrix, Stochastic Frontier Model (Junin) 70 60 r to 50 caidni 40 acyciffE30 20 10 0 0 0.2 0.4 0.6 0.8 1 1.2 Efficiency Indicator Source: Authors. This exercise demonstrated that there was a wide spectrum of variability among the results given that some schools obtained education outcomes that were a 100 percent match with the estimated value in the prediction model while others reached less than 10 or 20 percent of the achievement level predicted based on their resources (efficiency indicator). Approximately half the schools in the sample show efficiency results that are lower than the region's average. The efficiency level of the schools in the sample is 75 percent on average, which means that if current conditions were to remain unchanged, these schools could improve their results by 25 percent if they were to take maximum advantage of their inputs. According to this analysis, there is an almost perfectly linear relation between the schools' level of efficiency and their level of efficacy or learning achievements. This means that the efficiency of school management is directly related to learning quality, even when other school characteristics are similar. 15We used the average of learning results obtained in the National Evaluation 2004 as a reference. 16This is a parametric method that assumes that the error term that captures inefficiency in the productive process is distributed according to a Half Normal and the random perturbations according to a normal distribution. 24 What can a Regional Government do to improve education? The Case of Junin 3.4. Conclusions According to the evidence found, we can conclude that in Junin ­as well as the rest of the country­ significant progress has been made in education. The existing school infrastructure made it possible to achieve the high enrollment rates seen both in primary and secondary education. Additionally, there have been important budget increases in the sector over the past few years (primarily in the area of salaries) and almost all teachers now have a teaching title. Despite this, there are still a number of pending challenges. The main problems regarding school infrastructure relate to quality and maintenance. No additional investment will be profitable if there are no mechanisms in place to guarantee the adequate maintenance of the infrastructure provided. Additionally, the supply of basic pedagogical materials is insufficient and inadequate. In terms of human resources, Junin has enough teachers to cover the demand but they are assigned inefficiently and in a way that adversely affects single teacher rural schools. It is also necessary to stipulate regular teacher evaluation mechanisms directly linked to the fulfillment of teaching responsibilities and student learning results. A mechanism like the Score Card, which systematizes this kind of information, could be more efficient than requiring teachers to take exams. Nevertheless, any effort to evaluate teachers must be accompanied by measures to improve working conditions, which are currently very precarious in rural schools. It is important to recognize that there has been a significant increase in education expenditure in the country and in the region of Junin; nevertheless, this increase is neither sufficient nor adequately distributed. Education expenditure per student in Junin is below the national average, it is disproportionately allotted for compensation and salaries, and benefits provinces with more resources. Expenditure is particularly low and inefficient for preschool education. In general, we can conclude that the difficulties identified are due greatly to management problems. 25 What can a Regional Government do to improve education? The Case of Junin Chapter 4 Education Results In the past decade, Junin had made significant progress in its efforts to expand enrollment rates (MED-UEE, 2005). Currently, Junin is one of the regions with the best education levels for young adults: 78 percent of people aged 18 to 25 have finished secondary school, well above the national average. Compared to the rest of the country, a child from Junin is more likely to complete primary school, move on to the next level and graduate from secondary school. In fact, the region of Junin has the sixth highest percentage of students that finish secondary school. Additionally, the incidence of adult illiteracy is below 10 percent. Table 4-1. Education results and learning outcomes Category Peru Junin Ranking Population's education level Illiterate population (% 15 + years)1 88.9 88.9 13 Young population with complete secondary education ( percent)2 68 78 7 System Coverage (% population) Net enrollment rate in preschool (3-5 years old) 59 42 24 3 year-old boys and girls 24 45 23 4 year-old boys and girls 46 69 24 5 year-old boys and girls 56 64 17 Net enrollment rate in primary (6-11 years old) 93 95 8 Net enrollment rate in secondary (12-16 years old) 73 76 12 Student Progress (% students) Completion rate in primary (12-14 years) 72.5 79.8 8 Transition rate to secondary 91 95.6 10 Completion rate in secondary (17-19 years old) 51.3 61.8 6 Learning achievements: sufficient performance (% students) 3 2.º Primary Communication 15.1 12.1 9 2.º Primary: Mathematics 9.6 12 6 6.º Primary: Communication 12.1 7 14 6.º Primary: Mathematics 7.9 7.1 10 5.º Secondary: Communication 9.8 8.9 8 5.º Secondary: Mathematics 2.9 4.6 2 Source: MED-UEE, School Census 2007. 1INEI, ENAHO 2005 (data from Lima includes data from Callao). 2Population aged 18 to 25 years old. 3Results of the evaluation in 2004. Note: For more details, see annex 3.1. For the definition of variables, see annex 1.4. 26 What can a Regional Government do to improve education? The Case of Junin Despite the good news, the results for preschool education are unacceptable. Junin ranks 24th on net enrollment rates for this education level. Only Huanuco ranks lower. Barely 42 percent of children aged 3-5 attend a Preschool Education Center, compared to 59 percent nationally. The largest gap is found among the youngest children (between the ages of 3 and 4). Even more troublesome are the learning outcomes, which have remained stagnant at a low level in Junin and in the rest of the country, despite progress and innovations in education policy.17 Only 12 percent of second grade students in Junin perform at the expected level for their grade ("sufficient") in Mathematics or Communication. Meanwhile the national average is 15 percent in Communications and 9.6 percent in Mathematics. At the end of primary, the children in Junin have learned less than children in the rest of Peru, given that only 7 percent perform at the expected level. At the end of secondary, only 8.9 percent of the students have achieved the learning goals for Communication. Only 4.6 percent of students in fifth grade of secondary (11th grade) demonstrate sufficient performance in Mathematics, but with this low performance Junin is still ranked second in the country ­an indication that the quality of education in Peru is low(MED-UEE, 2005). This situation has persisted for a number of years18 and it is evidence that the low quality of education, although more critical in rural areas, is generalized and relentless. 4.1. Measuring reading skills For the purposes of this work, we evaluated reading fluency and comprehension using an oral and a pencil-and- paper (silent) reading test (see annex 3.3). Both tests were administered to third grade students. The oral reading instrument was also administered to second and sixth graders to measure progress throughout the education cycle. Silent reading evaluation (Photo: Jorge Rivas) The pencil-and-paper test was the one prepared by the Ministry of Education for the census evaluation of 2006 (herein to be known as the census test). The final score is the sum of correct answers.19 The oral reading test consists of two narrative texts with appropriate difficulty level for second and fourth grades, followed by comprehension questions. The children who read the first text fluently and can answer the comprehension questions are asked to read the second text. The questions seek to evaluate the child's ability to express the passage's main ideas or events in his or her own words and give an opinion on the text. The Oral reading exam (Photo Jorge Rivas) 17Between the years 2000 and 2006, expenditure on Education and Culture has increased 43 percent in the country and 44 percent in Junin (MEF 2007) as a result of: new legislation for the sector (Congress of the Republic 2003), progress made concerning administrative reforms to decentralize education, the development of an articulated and flexible curriculum nationwide (MED 2005) and a National Education Project to 2021. (CNE 2006). Better indicators and mechanisms were developed to evaluate the education system and national tests to determine academic performance have been perfected. 18The first comparative study in Latin America was conducted by UNESCO in 1997. Students in third and fourth of primary obtained the lowest results in the region with the exception of Paraguay (UNESCO-OREALC, 2003). Three years later, Peru obtained the lowest scores among the countries evaluated in the international PISA tests in the year 2000 (OECD, 2002). Two years earlier, the national test of 2002 showed that barely a third of the students in the fourth grade could read basic texts and less than 1 percent achieved a sufficient performance level in comprehension and the use of grammar and spelling rules (MED-UMC, 2002). 19 Alpha de Cronbach = 0.65 internal reliability coefficient. 27 What can a Regional Government do to improve education? The Case of Junin children that were unable to read the first text were asked Box 1. "DOGO" to read a list of five short words, and those that could not Once upon a time, there was a fat and furry dog read at least four words correctly were asked to name the named Dogo. Dogo lived with a family that letters on a list that included five vowels and five loved him very much. Dogo was an obedient consonants. dog, he took care of the house and ate all his food. One day, he went for a walk with his owner The results show a positive, statistically significant Lucas and got lost. Lucas was sad but relation between the oral test and the written test. These fortunately, Dogo showed up shortly after. Then, results also show that children that read more fluently they both went home together. have a better chance of understanding what they read.20 For example, in third grade the children that read less than 30 words a minute obtain on average comprehension scores of 7.5 percent while their classmates that read 110 words or more per minute have an average score of more than 72.7 percent. This same relation is found in the other grades. 4.2. Fieldwork Results The aforementioned instruments were administered in a random, representative sample of primary schools in Junin. Additional information was gathered on a set of factors that may be influencing the education results in each school. The data was collected between May and June of 2007 in 83 public schools in the region's nine provinces. A total of 50 school principals, 118 teachers, 181 parents and 1,164 children were interviewed; 92 hours of classroom observation were conducted; and 1,687 students from the second, third and sixth grades were tested on reading performance (see annex 3.2).21 The findings were complemented with information obtained from UGEL directors, area heads and specialists from all of the UGELs in the region. Next, we present the results. Results for reading skills in Junin are generally low On the oral evaluation, we found that 35 percent of second grade students cannot even read one word of a simple text (25 percent are unable to even recognize the letters)22 and barely 43 percent reached the fluency standard for their grade.23 In third grade, only 27 percent read with the fluency expected and 18 percent of sixth grade students read less than sixty words per minute (Table 4-2). It is even more important for children to understand what they read, to be able to put it in their own words, to relate it to their experiences, and to express opinions regarding the material read. But the reading comprehension results are just as poor: in the second and third grades less than a third of the students 20 The census test has a partial correlation of r = 0.55 with fluency and r = 0.26 with oral cmprehension, both statistically significant (sig. < 0.01). We also observed a strong, positive and statistically significant correlation between the two components of the oral reading test: fluency and comprehension (r = 0.71, sig. < 0.01). 21 Not all the children evaluated were interviewed. Instead, interviews were conducted only with children who had taken the oral test. It is important to point out that during the administration of the census test, we experienced strong resistance from teachers. In the end, and after much insistence, we were only able to administer the tests in 69 of the 83 schools selected. Of these 69 schools, 33 had participated in the census evaluation of 2006. Despite the fact that these children were familiar with the readings and questions, we found no significant differences between the scores obtained by them and those obtained by students in schools that had not participated in the census evaluation (see annex 3.6). For more details on the difficulties encountered during test administration, see table annex 3.2-h. For an anecdotic recount of fieldwork experiences, see annex 3.8. 22 In the rural areas of the rainforest, we observed that the children that were unable to read the "Dogo" reading found it difficult to recognize and identify the words and letters that they were shown. Frequently, they confused the "a" with the "o", the "f" with the "t" and the "b" with the "d". When asked about this, teachers indicated that the reading deficiency relative to printed texts was due to the fact that children had little contact with this kind of material because the only printed texts that they had read were provided by MED and contained very few readings. 23The expected achievement for fluency was estimated by using the standards identified in literature for different grades of Primary school as a reference (véase el anexo 3.4). We consider that a child has reached a standard if he achieves the level anticipated at the end of the previous grade. 28 What can a Regional Government do to improve education? The Case of Junin Table 4-2. Level of fluency achieved per grade (% of students) Grade Level achieved 2º 3º 6º Cannot read1 35 10 0 Read 1-24 words 17 5 0 Read 25-44 words 20 19 3 Read 45-59 words 19 37 15 Read 60-89 words 9 21 37 Read 90 or more words 0 6 46 1Identifies six or fewer letters form a list of ten that includes five vowels (% of total sample). perform at the level indicated in the curricular plan (MED, 2005).24 Around half of these children could not retell the story they had just read ("Dogo") in their own words. When a third grade student was asked to tell the story in his own words, he replied frustrated: "I don't remember it. I need to go over it a lot." According to the fieldwork reports, many students tried to repeat the passage literally and if they couldn't recall the passage from memory, they opted not to answer or answered in a mediocre and imprecise way: "It's about a dog named Lucas; the kid got lost" (9 year-old, third grade). Students also found it difficult to identify the part of the text that they liked the most and explain why. The majority of the students gave concrete and literal answers such as: "I liked the dog because his name is Dogo and he takes care of the house" (13 year- old, sixth grade). This difficulty is more than likely the result of teaching methods that rely on memory and repetition instead of reasoning and creativity. In the classes observed, there was a clear tendency to reinforce the literal answer as the correct answer. Hence, many children confuse reading comprehension with literal repetition of a passage. Nevertheless, some children respond more elaborately and make affective connections between the story and their personal experience: "Sometimes when something is lost and you find it you are very happy" (7 year-old, second grade); "I like the part where he gets lost because I like adventure stories where people get experience and I liked the part when he was happy because he had found it"(10 year-old, sixth grade). 24 Unlike fluency, there are no precise standards for comprehension. As such, the scores are calculated according to the competences indicated for each grade in the current curricular design. This was applied to both the oral test and the census test and in both cases the percentage of correct answers was estimated based on the maximum possible score (annex 3.5). 29 What can a Regional Government do to improve education? The Case of Junin Figure 4-1. Fluency and comprehension: level reached in May in comparison with the level expected Fluency 150 140 125 120 110 minute per 90 90 86 (6to) correctly 60 60 read 49 (3ro) 30 Words 30 26 (2do) 0 Abr Jun Ago Oct Dic Abr Jun Ago Oct Dic Abr Jun Ago Oct Dic Abr Jun Ago Oct Dic Abr Jun Ago Oct Dic Abr Jun Ago Oct Dic 1st 2nd 3rd 4th 5to 6to Grade and month of the academic year Expected level Achieved level Comprehension 100% 92% 96% 80% 80% (%) 60% 55% level 52% (6to) 40% 32% 27% (3ro) Comprehension 20% 16% 17% (2do) 0% Abr Jun Ago Oct Dic Abr Jun Ago Oct Dic Abr Jun Ago Oct Dic Abr Jun Ago Oct Dic Abr Jun Ago Oct Dic Abr Jun Ago Oct Dic 1st 2nd 3rd 4th 5to 6to Grade and month of the academic year Expected level Achieved level 30 What can a Regional Government do to improve education? The Case of Junin Urban full grade school, second grade (Photo: Jorge Rivas) Rural school in Shabashipango (Photo: Jorge Rivas) It is understandable that performance on an oral test administered by an unfamiliar evaluator may be affected by the anxiety or inhibition generated by the conditions of the test administration, and particularly in the case of students whose native language is not Spanish. Thus, performance on the census test offers a valuable benchmark for the third grade: the average performance on the census test tends to be higher and more homogenous than on the oral test.25 Even so, only 17 percent reached reading level 3 (sufficiency). The bulk of students reached level 1 (40 percent) and 2 (38 percent), and 4 percent were situated in level 0 (cannot read).26 According to the first census evaluation conducted by the Ministry (MED-UMC, 2007), 18 percent are in level 0 and 24 percent reach level 3 while levels 1 and 2 group 29 percent and 30 percent respectively. But perhaps most startling is the fact that the delay produced in the early grades is dragged into subsequent grades and becomes increasingly larger. The study revealed that children in Junin read below the expected level and that the gap between the actual achievement and expected achievement widens throughout the school years both in terms of fluency and comprehension. Taking the expected values as reference, we can situate the performance of the students within a continuum and compare their progress not only in terms of prior grades but also with regard to the expected achievement based for each grade and month. Figure 4-1 uses these referential values to show the results as the average number of months of delay with regard to the expected performance level. For example, we can see that the students in the second grade are already three to four months behind in their reading performance. In the third grade, the delay is equivalent to six or seven months of studies; and by sixth grade, students show a two-year academic delay, as they perform at a level expected for students who are finishing the third grade. Unfortunately, it is not surprising to find significant disparities in learning results between schools and students with different socioeconomic characteristics. These inequalities have been documented in numerous national and international studies and are evident to anyone who comes in contact with Peruvian public education. It is surprising however to find that the learning delay for reading is a generalized problem. This means that even students with more social and educational advantages within the public system are unable to reach the expected performance levels. Furthermore, there is no evidence that any of the sixth grade students have reached curricular objectives with regard to reading comprehension and barely 10 percent 25For example, while 19 percent of the third grade students that took the oral test were unable to read even half the text, 21 percent were unable to respond any of the comprehension questions. On the written test, only 4 percent were unable to understand the narrative text (which was even longer than "Dogo." 26The achievement levels were calculated based on the number of correct answers (see table annex 3.5-d) and were meant to follow the criteria established by the MED (annex 5.1) due to the fact that we had no access at the time to the specific methodology used by the ministry. 31 What can a Regional Government do to improve education? The Case of Junin Table 4-3. Average number of months of delay in reading skills Fluency Comprehension Census test 2º 3º 6º 2º 3º 6º 3º Total (overall average) ­4 ­7 ­24 ­3 ­7 ­22 ­6 Location Urban 1 ­3 ­20 0 ­3 ­18 ­2 Rural ­7 ­9 ­26 ­5 ­11 ­25 ­13 Single teacher ­8 ­9 ­28 ­9 ­14 ­31 -5 School type Multi-grade ­7 ­10 ­27 ­5 ­13 ­27 ­11 Full grade ­1 ­4 ­21 ­1 ­4 ­18 ­17 Spanish ­3 ­5 ­23 ­2 ­7 ­20 ­14 Mother language Quechua ­8 ­8 ­29 ­11 ­13 ­23 ­14 Ashaninka ­13 ­17 ­31 ­13 ­14 ­38 ­3 Preschool education No ­6 ­9 ­27 ­5 ­9 ­26 ­14 Yes ­3 ­6 ­23 ­2 ­7 ­20 ­6 Students who are not lagging behind 44% 29% 10% 27% 32% 0% 41% read with the fluency expected. Even the sixth-graders in urban schools, who are the most advanced, suffer from a two-year academic delay in the development of reading competences. Evidently, the problem of low quality education is not only present in single teacher, rural and poor schools although these institutions are undoubtedly worse off. The poor performance in schools that concentrate more human resources and materials and the fact that these resources are unequally distributed to the disadvantage of the most vulnerable groups are indicators of inneficient education management. Who is at a disadvantage? "Here they begin to read in the third grade; when I get them in the first grade they can't even hold a pencil," said a teacher at a rural, single-teacher school in Junin. It is certain that students who are raised in more literate environments, such is the case of those who live in urban areas and come from middle and upper classes, generally are familiar with many aspects of reading and writing before they begin school. The environment provides them with multiple opportunities to explore the written language and build a bridge that facilitates early reading (Bus, 1994). On the contrary, boys and girls from poor, rural areas live in environments where written information is almost inexistent, which means that opportunities for contact with the written language are scarce prior to schooling. This gap, which is forged before beginning school, becomes evident when they face the literate world of the classroom. Let's look at the best case scenario: it is estimated that a second-grade 7 year-old student whose native language is not Ashaninka, who received preschool education and attends an urban, full grade school can read 57 words a minute. In the third grade, this student will read 67 words a minute, and in the sixth grade, at the age of 11, will read 151. This means that the student may even exceed the expected standards. This same student, when in the third grade, would be able to answer 78 percent of the questions on the census test correctly. In the sixth grade, he or she would answer correctly all the oral comprehension questions, but by then, attending an urban school or being male will not have a significant effect on comprehension levels (Table 4-4). 32 What can a Regional Government do to improve education? The Case of Junin Table 4-4. Regression results by grade Dependent variable Fluency Comprehension Census test Grade 2.º 3.º 6.º 2.º 3.º 6.º 3.º Independent variables Rural ­16.48** ­9.15 ** ­5.73 ** ­3.1% ** ­3.2% ** ­4.4% ** Full grade school1 4.58** 5.82 ** 7.90 ** 5.0% ** 11.1% ** 14.4% ** 3.1% ** Ashaninka2 ­15.04** ­24.22 ** ­9.76 ** ­11.5% ** ­5.6% ** ­21.1% ** ­23.0% ** Female ­3.11** 0.82 3.68 ** 2.0% ** 3.2% ** 1.0% ** Preschool education 5.32** 3.95 ** 8.20 ** 5.3% ** 0.7% 7.4% ** 1.8% ** Age ­1.85** ­0.92 ** ­4.81 ** 2.1% ** 0.0% ­3.5% ** -0.8% ** Constant 46.98** 57.58 ** 134.56 ** ­4.5% 20.2% ** 85.2% ** 72.9% ** Cases (N before applying FA) 329 530 285 215 479 285 475 Prob > F 0.000 0.000 0.000 0.000 0.000 0.000 0.000 R2 0.293 0.224 0.221 0.045 0.067 0.215 0.280 R2 Adjusted 0.293 0.224 0.221 0.044 0.066 0.215 0.280 MSE root 20.41 22.38 23.24 22.9% 26.9% 26.5% 14.3% Average VIF 1.51 1.49 1.54 1.65 1.42 1.14 1.45 1The differences between multi-grade schools and single teacher schools are not significant; for this reason, the final models only include the variable that distinguishes full grade schools from all others. 2The effect of other indigenous languages is not significant (controlling for the other variables, there are virtually no differences between Spanish speakers and Quecha speakers); as such, the final models only include the variable that distinguishes Ashaninka students from the rest. * sig < .1; ** sig < .01. Note: See the definition of variables and models in annex 3.7. In general, after controlling for the effects of the students' age, sex and access to preschool education, native Ashaninka speakers perform significantly lower than native Spanish or Quechua speakers, regardless of the type of school attended or the area in which they live. Likewise, students in single teacher or multi-grade schools obtain results that are significantly lower than those seen in full grade schools, even when they are located in an area similarly rural. The students from schools located in rural areas obtain lower results than those enrolled in urban schools, irrespective of school type. Finally, students that received preschool education obtain better results than those who did not. The location (rural or urban) and the school type (full grade or not), as well as the student's access to preschool education, age, sex and native language, explain 29 percent of the variance in the census test scores, 33 percent of the variance in fluency scores in second grade, and 22 percent of the variance in fluency scores in third and sixth grades. The comprehension level is so low in the early grades (e.g., half of all second grade students understand less than 5 percent) that socio-economic variables do not contribute to explaining the variance. In the sixth grade, reading performance is higher, and so it is the effect of social factors on scores, although the effects of rural location and gender are insignificant once other variables are controlled. Ashaninka students have the lowest results of all The student's native language is the variable with the highest explanatory power in terms of reading performance, even when other factors are kept constant (Table 4-4). The results of the census test show that in the third grade, after controlling for other socio-economic variables, there is a gap equivalent to 1.5 years of academic work between Ashaninka and non-Ashaninka students. In the sixth grade, Ashaninka students show more than a year delay in comprehension compared to their non-Ashaninka peers, regardless of the type of school they attend, their age, their previous access to preschool education or their gender. 33 What can a Regional Government do to improve education? The Case of Junin Figure 4-2. Reading performance of Ashaninka children and non-Ashaninka students in comparison with expected results 150 140 etu 125 130 120 110 min erp 90 90 88 correctly 69 60 60 64 reads 52 39 Word 30 30 28 19 0 1 Abr Jun Ago Oct Dic Abr Jun Ago Oct Dic Abr Jun Ago Oct Dic Abr Jun Ago Oct Dic Abr Jun Ago Oct Dic Abr Jun Ago Oct Dic 1st 2nd 3rd 4th 5to 6to Grade and month of the academic year Expected level No Ashaninka Ashaninka 100% 96% 92% 93% ) 80% 80% %(levelnoi 60% 58% 52% ns hee 40% pr 38% moC 32% 27% 24% 20% 16% 21% 17% 18% 0% 2% r n o t c r n o t c r n o t c r n o t c r n o t c r n o t c Ab Ju Ag Oc Di Ab Ju Ag Oc Di Ab Ju Ag Oc Di Ab Ju Ag Oc Di Ab Ju Ag Oc Di Ab Ju Ag Oc Di 1st 2nd 3rd 4th 5to 6to Grade andmonth of the academic year Expected level No Ashaninka Ashaninka 34 What can a Regional Government do to improve education? The Case of Junin Half of the Ashaninka children were unable to read a single word of text. These students identified, on average, barely one word of five on the list and only four out of ten letters. An Ashaninka student in the second grade reads an average of 1.4 words a minute, compared to the 16 words that Quecha students read, or to the 29 words that a native Spanish speaker can read. The comprehension scores reflect these differences as well, which are most notable in sixth grade when the average reading level of Ashaninka students denotes a pedagogical delay of almost four years in the area of comprehension. It is important to point out that in the sixth grade all of the children, not just the Ashaninkas, have fallen significantly behind the standard, though it is evident that the situation of the Ashaninkas requires special attention (2). It is only natural that parents are reluctant to send their children to school if they feel that they are not learning because they believe that the children will be more productive helping in field or household tasks. A poor quality education means that there are fewer opportunities for social mobility, which is another reason why parents think that it is "unprofitable" to send children to school. Multi-grade and single teacher schools that offer lower quality education The students in full grade schools achieve better results in reading comprehension and fluency than their counterparts in single teacher or multi-grade schools even if they share the same social background. In the second grade, students at single teacher and multi-grade schools read with a fluency level that indicates almost a year of delay while their counterparts in full grade schools read at the expected level. The effect of the school type on reading comprehension at the end of primary school is the second factor that best explains differences in performance (Table 4-4). In sixth grade, students in single teacher schools demonstrate a comprehension lag equivalent to over three academic years, more than a year behind students with similar backgrounds that attend full grade schools (Figure 4-3). Without a doubt, parents perceive this reality and repeatedly express the desire --and in many cases make the effort-- to send their children to full grade, urban schools because they believe that these schools offer higher quality education. The teachers have also indicated that teaching in the city is easier and that their efforts produce better results, and thus, are more rewarding. According to these teachers, urban students have more resources to buy the materials requested for class, have access to multiple information sources, and have parents who, on average, have a higher education level. Figure 4-3. Student reading performance by school type based on the anticipated result 150 100% 140 96% 92% 93% 125 130 etuni 120 ) 80% 80% 110 m erpyl 94 (%level 65% 90 90 noi 60% 52% ect 77 ns rr 69 co 44% d 60 60 57 heerp 40% 38% reas 32% 33% rdo 39 moC 40 W 30 30 37 20% 16% 20% 19% 16 12% 0 r n o t c r n o t c r n o t c r n o t c r n o t c r n o t c 0% t t Ab Ju t c r t c r c r t c r t c r c Ag Oc Di Ab Ju Ag Oc Di Ab Ju Ag Oc Di Ab Ju Ag Oc Di Ab Ju Ag Oc Di Ab Ju Ag Oc Di r n o n o n o n o n o n o Ab Ju Ag Oc DiAb Ju Ag Oc Di Ab Ju Ag Oc Di Ab Ju Ag Oc DiAb Ju Ag Oc Di Ab Ju Ag Oc Di 1st 2nd 3rd 4th 5to 6to Grade and month of the academic year 1st 2nd 3rd 4th 5to 6to Grade and month of the academic year Expected level Multigrade, single teacher Full grade Expected level Multigrade, single teacher Full grade 35 What can a Regional Government do to improve education? The Case of Junin The quality of education in urban schools but is still not sufficient In second grade, students from rural schools demonstrate a learning gap equivalent to seven months of academic work in reading fluency, as they read barely eighteen words per minute. In contrast, their peers from urban schools read, on average, above the expected level. In third grade, there are significant differences in scores on the census test: the average of the urban schools is 72 percent, which indicates a two- month delay in pedagogical work, while the average at rural schools is 61 percent, which is the equivalent to a delay of one year and two months. The gap persists until the end of primary, but when students in urban schools reach the sixth grade, they have fallen far behind the expected performance level by an equivalent to two academic years (compared with 2.6 years in the case of rural schools). At that poin, the disparities are overshadowed by a generalized low performance. 4.3. Conclusions Significant progress has been made in terms of increasing expenditure, producing legislation, and implementing administrative reforms in the Education sector. Nevertheless, the quality of public primary education in Junin is low and unequal. Children in rural, impoverished and remote areas receive lower quality education than those living in cities and attending full grade schools. Ashaninka students are at a greater disadvantage. In any case, it is surprising to see that the students in schools with more resources and more and better teachers also show poor performance that worsens as the years go by. This suggests that that poor management is a key, underlying problem. We believe that it is time to generate a change at the school level. This can be done in many ways, but the information gathered gives us a clue about where to start: it is necessary to draw up a clear, concrete objective that all primary schools can understand. 36 What can a Regional Government do to improve education? The Case of Junin Chapter 5 The greater challenge: education management New regional management initially focused its efforts on controlling and reversing crippling corruption in the education sector. Among the measures taken, a newly appointed Regional Director proceeded to replace UGEL directors, who have in turn changed key personnel within their UGELs. Although this task is critical, it is not enough to correct the system's inefficiency, the low quality of education, and the inequality of education opportunities in the region. The Regional Government should lead the process for substantive change in these areas, some of which will surely produce immediate and relevant results for the population. As is evident in annexes 4.1 and 4.2, the overlapping of education functions, as defined by current laws, allows the Regional Government of Junin to assume a leading role in strenghtening its autonomy, as long as it has the political will to do so and a clear plan to put in place. Currently, education service administration for primary, secondary and higher (non-university) education is a function of the regional and municipal governments.27 The Organic Law on Regional Governments (LOGR, article 47) and the Organic Law of Municipalities (LOM, article 82) lays down the functions of each level of government in more detail. Additionally, the General Law on Education (LGE) defines the functions of the school (article 68), UGEL (article 74), DRE (article 77) and MED (article 80). As stipulated by law, the objective of education management as a whole is to "[...] strengthen the decision making capacity of schools so that they can act with pedagogical and administrative autonomy." (LGE, article 64). According to the legal framework currently in place, several key functions such as curricular diversification, teacher training, pedagogical innovation, literacy efforts, infrastructure and equipment, and even strengthening decentralized systems for education administration, are shared by two or more levels of government. Moreover, the LOGR, modified by Law 27902, stipulated that the Regional Director must be selected through a public contest. In December 2006 a law was passed (Law 28296) which regulates the transitory regime for regional sector offices of regional governments, stipulates that these offices are dependent bodies of the corresponding regional governments and are run by regional directors, who are directly appointed or discharged by the Regional President with guidance from the regional sectoral manager (which in the case of education is the Social Development Manager). Both the LOGR and Law 28296 reconfigure the relations between regional offices and the Central Government but in practice, the regional education authority has two heads, one sectoral and the other administrative. This overlapping of 27Law of Decentralized Bases, articles 36 and 43; constitutional reform of Chapter XIV of Title IV on decentralization, articles192 and 195 37 What can a Regional Government do to improve education? The Case of Junin accountability relations --which are not clearly defined in legislation-- can cause confusion and facilitate an evasion of responsibilities. The LGE tends to give more administrative responsibility to the local sector unit (UGEL) while the organic laws for regional governments and the municipalities emphasize the Regional Government. In any case, the legal framework leaves enough margin for the Regional President to reconcentrate some key education functions, since the UGEL is a deconcentrated body mapped to the Regional Government. As such, some deconcentrated functions that are not adequately executed by the UGELs could be reconcentrated in the Regional Education Office and this decision can be made by the Regional Government. Paradoxically, although the law favors education decentralization, a significant number of pedagogical functions remain under the control of the Central Government, even when international practice shows that these can be decentralized.28 For example, MED is responsible for regulating the design, production, distribution and use of education materials, and it is also in charge of textbook selection and procurement, meaning that if a school wants to acquire books or materials that are different from those chosen by MED, it must do so with its own resources. Naturally, the schools in poor, rural areas have fewer possibilities than those located in wealthier areas. This issue came up repeatedly during the interviews, particularly in schools serving students whose native language is not Spanish. The transference of functions and resources from the education sector to the Regional and Local Governments has progressed slowly in Junin. In May 2004, the Regional Education Council of Junin initiated a process to reorganize and restructure the Regional Education Office in three phases: first, the administrative reorganization and restructuring of all agencies and UGELs in the region; second, the reorganization and restructuring of the education service delivery in order to overcome the regional education emergency; and third, definition of the Regional Policy Guidelines for Junin by formulating a Regional Education Plan. Nevertheless, the process was truncated and none of these steps were implemented until the new regional administration took office in 2007. 5.1. The UGEL: the weakest link According to LGE, UGEL's primary objective is "[...] to strengthen capacities for pedagogical and administrative management at schools so that they can become autonomous" (article 73, our emphasis). This entails, fundamentally, providing technical, pedagogical and institutional support to the schools in their jurisdiction as well as monitoring and supervising their operations. In practice, the UGELs' management has been characterized primarily by its focus on completing an infinite number of administrative tasks. This causes delays in and creates obstacles to efforts to address more substantive issues. In Junin, the situation is particularly severe. Based on testimony from UGEL and DRE employees, and interviews to school principals, we found that the UGELs are unable to meet the schools' basic requirements in a timely, efficient manner because they are either unaware of the needs or are short-handed to address them. 28International experience is diverse when it comes to deciding which level of government should be in charge of education functions. South Africa devolved basic education functions to provincial governments; Indonesia devolved responsibilities to district governments; and Pakistan used a combined model. Armenia and New Zealand delegated the majority of administrative responsibilities to school councils. El Salvador did the same with rural schools in remote areas. Nicaragua introduced a policy that allowed schools to self-administer. Brazil devolved responsibilities for primary education, previously managed by the states, to municipal governments. Many other countries have decentralized specific functions such as the purchase of textbooks and materials, renovating school infrastructure, school maintenance, etc. (Winkler, 2006). 38 What can a Regional Government do to improve education? The Case of Junin It is important to point out that there are UGEL specialists29 who are deeply committed to their work. We have witnessed their dedication during our fieldwork. To fulfill their functions, and in spite of a number of precarious conditions, these specialists invest their own resources, travel distances that no other employee in the sector would be willing to cross, and do so for a lower salary than what they would get if they had remained in the classroom (since most of them are teachers). These specialists have a great deal of knowledge about the communities and schools they work with and have a genuine interest in education in their jurisdiction. Unfortunately, individuals such as these are a minority, as the conditions and incentives they receive for that work do not encourage good performance. This is fundamentally due to three managerial factors that can be corrected: (i) the disproportion between the number of schools to attend and the UGELS' capacity, which indicates a lack of correspondence between functions and resources (ii) problems in the selection of specialists and, (iii) a lack of incentives for specialists to perform well. The field of action exceeds the UGEL's capacities In terms supervision, monitoring and pedagogical support provided to schools, we found that one of the biggest problems faced by the UGELs comes from the large number of students they serve, who are generally distributed in a large number of schools, many of which take several days to reach. In Chanchamayo, four primary school specialists report to be in charge of 378 schools. These specialists indicate that it takes them 5 days roundtrip to reach the furthest school and that only 30 percent of the schools are visited each year. The UGEL Huancayo aims to visit each school at least three times a year to follow up on the specialists' recommendations to teachers. However, its specialists indicate that because each of them is in charge of around 350 schools means that they are unable to supervise even 3 percent of the schools in their list. The schools in Santo Domingo de Acobamba and Pariahuanca districts are the farthest away and, consequently, are visited the least: the trip takes seven hours by car and two or three days on foot. The distance is further exacerbated by the insecurity generated by the presence of terrorist groups in the area. Likewise, the specialists of the UGEL Satipo are in charge of 89 preschools , 455 primary schools and 73 secondary schools. They indicate that they used to visit schools twice a month over the course of a week but due to problems with terrorism, they now only visit the nearest schools. Consequently, the relation between the UGEL and rural schools is very weak and in some cases inexistent. "The UGEL does not help", is one of the most frequent complaints from teachers and principals given that the specialists mostly visit the schools easiest to access. Teachers in rural areas permanently feel abandoned and forgotten and contend that they do not receive enough support from the UGEL and the Ministry of Education. Only 36 percent of single teacher schools and 26 percent of teachers at multi-grade schools indicated that they had received support from the UGEL to improve their pedagogical work. Our research shows that the education authorities' pedagogical support for teachers is insufficient or in many cases inexistent. Teachers claim that the limited training available is applicable to very different contexts from those found in the schools where they work, which means that the training offered is not applicable to their students. Perhaps the most surprising situation faced by the UGELs is their peculiar "autonomy without resources." The design for a decentralized system should clearly specify how this will be financed; for example, by intergovernmental transfers, by increasing the power of subnational governments to collect their own revenues, or by directly financing schools. According to Winkler (2006), financing should be consistent with the new assignment of expenditure responsibilities but, as we have already seen, this also lacks clarity. In other words, the budgetingnorms are still hazy. 29Employees (typically teachers) in charge of providing technical, pedagogical and institutional support to schools. 39 What can a Regional Government do to improve education? The Case of Junin Currently, there is no coherence between the functions and resources allocated to monitoring and pedagogical/technical support. The principal obstacle, in the opinion of specialists at all the UGELs, is the limited budget for visits, which contemplates only forty gallons of gasoline per month for evaluation, monitoring and technical, pedagogical and administrative visits. The UGEL specialists in Chanchamayo claim that when the money arrives to the UGEL, the accountant neglects to assign funds to specialists because "he thinks that school monitoring is not important." UGEL specialists in Huancayo also attribute their poor performance to the lack of funds for trips and travel expenses, stating that they share one truck with enough gasoline to cover only 50 kilometers. The river route to Atalaya has 48 schools mapped to the UGEL Satipo; but the trip along the Tambo river takes a week and a half and requires three gasoline cylinders (180 gallons). Problems in the selection of specialists The selection of UGEL specialists is not carried on according to quality or efficiency criteria. On the contrary (and despite the fact that it should be done through public contest), some of the specialists hired are teachers who were expelled from their schools due to poor performance; yet, these are the very people who are asked to "advise" and "monitor" their colleagues in the classroom. This is a serious distortion of the nature of this job that favors an administrative limbo in which expelled teachers avoid prosecution or punishment, and in some cases, can aspire to an administrative career in the sector. Some of the specialists and other UGEL officials are teachers or principals that have been removed from their positions in schools due to "Non-compliance with functions" or "rupture of relations" between the teacher or principal and parents or the community. Among the cases reported, we find situations that range from teachers with high absentee rates (some do not show up for months) to cases of sexual abuse that are difficult to prove in court. The Director of Legal Counseling at the UGEL in Huancayo --whose testimony concurs with reports gathered at other UGELs-- argues that it is very difficult and costly for UGEL specialists and principals to initiate administrative processes against teachers because the majority of those accused file complaints against the accusing officer, alleging abuse of authority and usurpation of functions. The few administrative processes that are opened linger indefinitely, and from the limited number of cases that finally get a ruling, only very few end with sanction because there are no mechanisms in place to enforce it. The teachers that are sanctioned continue to work in the area unnoticed, or move to other locations, where there are no records of their wrongdoing. The system does not facilitate communication and access to consolidated information to prevent this from happening. This is compounded by a lack of continuity of supervision and support to schools, because of the high turnover of specialists and weak dissemination of information. In general, specialists take with them all the information they have gathered during their service, along with any materials (including office supplies) and the only copies of instructions and norms. The relations they established with schools and communities leave with them as well, and the new specialists have to start from scratch. Few incentives for good performance As we have already mentioned, there are specialists who are committed to doing a good job. Unfortunately, their efforts are rarely, if ever, rewarded. In general, they go unnoticed by the administration, whose mechanisms for personnel evaluation are limited, at best, to delivering documents and reports (which are not systematized) and training. 40 What can a Regional Government do to improve education? The Case of Junin Box 2. Education Municipalization in Junin In October of 2006, the Central Government of Junin decided to implement the Pilot Plan for the Municipalization of Education Administration at the Preschool and Primary School level in fifty municipalities throughout the country (two per region). The criteria used to select the districts to implement this experience were not clear. The country had no typology of municipalities and without this instrument it is very difficult to successfully conduct a process that should assign functions and competences according to capacities and possibilities. Basic data from the province of Concepcion and the districts where education municipalization was conducted (Santa Rosa de Ocopa and Comas) show a varied picture. Among the 1,831 districts in Peru, Santa Rosa de Ocopa ranks 461 in the Human Development Index (HDI), while Comas ranks 1,220. Moreover, Comas ranks 1,713 in schooling (percentage of school-aged children enrolled in school) and 1,497 in education attainment (population's average years of education). Furthermore, Comas has 34 primary schools in its jurisdiction and is the second largest school district in the province after Andamarca, while Santa Rosa de Ocopa, near the provincial capital, has only two primary schools. The budgets for both municipalities in the year 2006 were very limited, as was the case in the rest of the country (1.47 million soles in Comas and 470,865 soles in Santa Rosa which have, respectively, 7,952 and 2,079 inhabitants). The municipality of Comas has nine employees (four permanent workers, four contract employees and a temp). In our visit to the municipality of Santa Rosa de Ocopa, among other things, we found that it was equipped with several computers but had no internet access. Given the lack of resources, it is not surprising, for example, that neither municipality has a project management office for pubic works and infrastructure. Their activities are completely focused on small projects and activities linked with food programs and compliance with basic registry and license processes. In Junin, the process does not progress smoothly and is sometimes thrown into reverse gear. When we visited the UGEL in May, we were informed that they had already established the municipal education councils (CEM) in both districts. These are the instances that are in charge of education administration at a municipal level. Nevertheless, in our visit to Santa Rosa de Ocopa in June, we were informed that the Technical Secretary had resigned and another was to be appointed. The CEM had not been conveened in any of the districts. This is exacerbated by the communication strategy used by the teacher's organization to stir up hostility in the population regarding "threats of privatization." These individuals go door-to-door to "inform" parents that if municipalization goes through, they will have to pay 100 soles in enrollment fees. The local mayors were taken by surprise by ferocious resistance from misinformed, concerned parents. And there is more: UGEL officials in Concepcion have been reluctant to transfer responsibilities to the municipalities, arguing that they lack capacities and had not complied with the process's formal requirements. On the other hand, the mayors claim that the UGEL "frightens" municipal officials: "The come back [from training at the UGEL] saying that they don't want municipalization because it is very complicated," said the mayor of Comas, who was visibly annoyed and added that the UGEL imposes a series of bureaucratic obstacles. The system is plagued by distorted incentives that reward people who know how to beat it, and fail to acknowledge those who follow the rules. In general, teachers assigned to administrative jobs at the UGEL or DRE take their teaching posts with them from the school, because neither the Personnel Assignment Chart (CAP) nor the UGEL budget allows them to be hired as payroll employees. These new officials continue to receive their salaries as teachers, but have no right to bonuses for classroom work (around 500 soles, almost 50 percent of a teacher's income). Nevertheless, many find a way to keep receiving this money and are often backed by management at the UGEL. During our visit to the DRE, we found documents issued by an UGEL director contending that his administrators continue to teach and requesting the corresponding bonus for them. The head of personnel at the DRE in Junin, who received these documents, worked previously in this UGEL and had first-hand knowledge of the tasks performed by the specialists listed in the letter, who, according to them, had never performed classroom duties. This makes us wonder whether the UGEL specialists, who supposedly lack the time to monitor and supervise, can find time to teach in a classroom; or the new UGEL officials continue to cheat the system and collect benefits illegally. In any case, the situation reveals serious problems with accountability and management of human and financial resources at the local level. Unlawful collections are not limited to bonuses for work that was never done. In fact, specialists 41 What can a Regional Government do to improve education? The Case of Junin request illegal payments from schools in each jurisdiction. According to testimony gathered by recently designated officials at eight of the nine UGEL, it is common to find specialists that ask principals for flat fees to conduct bureaucratic procedures that should be free, including authorizations and favorable reports. The UGEL and its specialists have very little power to promote change in the schools. In general, after a school visit, the specialist makes recommendations to correct the problems detected. Ideally, the specialist should return to the school to verify that the recommendations have been followed. Nevertheless, as we have already indicated, there are not even enough funds to visit all schools once a year. Specialists at the Satipo UGEL, which has more schools and dispersion that any other UGEL in the region, have opted to visit only schools with contracted teachers, because appointed teachers have no incentives to follow the UGEL's recommendations and the specialists cannot take any disciplinary measures against them. Therefore, the Regional Government has two alternatives: it can reassign monitoring, supervision and support tasks to another management level that has enough resources and capacity to carry them out properly (outsourcing is also a possibility); or it can assign more funds to the UGELs and put in place control mechanisms to improve the selection and performance of specialists, including evaluations done by the schools. In principle, this last option appears to be the best one, but it is also the most difficult for the Regional Government to implement under the current circumstances. Regarding the first option, there are two issues to consider: capacities and transaction costs. Typically, the larger the jurisdiction for a monitoring uning, the easier it is to staff it with adequately trained monitors or specialists, but it is more costly to send them to the schools because of the longer distances. On the other hand, smaller monitoring jurisdictions place monitors closer to the schools, which means lower transaction costs and more fluid relations with schools; yet, it is harder to find trained monitors that would want to live in remote areas. We found that the UGELs are at a disadvantage in both areas: they lack the capacity and face very high transaction costs. Two alternatives remain: concentrate monitoring and supervision tasks at the regional level, where there is more capacity and resources; or deconcentrate these functions to the district level, where transaction costs are lower. Perhaps the best alternative at this time would be to combine both: meaning that design and financing should be the responsibility of the Regional Government and execution should be conducted by local teams in each district. This does not mean transferring functions to municipal governments (or transferring UGELs functions to provincial governments). 5.2. School management and participation One of the most innovative aspects of Peru's current legislation on education is, without a doubt, the emphasis on the school autonomy. According to the LGE, the objective of education administration is to consolidate the pedagogical and administrative autonomy of schools by strengthening their decision- making capacity. The LOM indicates that the municipalities must: "Monitor pedagogical and administrative management of the schools in their jurisdiction in coordination with the Regional Education Office or the Local Education Management Units to strengthen institutional autonomy" (article 82, our emphasis). Additionally, the LOGR stipulates that the function of the Regional Governments is to: "Work with the Local Governments to strengthen schools by promoting their autonomy, innovation capacity and democratic management, as well as intersectorial coordination and organization in school networks with social participation" (article 47, our emphasis). This emphasis has resulted in the evolution of the Institutional Education Council, CONEI, formerly known as the School Education Council. The CONEI is defined by law as: "[...] a body for citizen participation, consensus and oversight" (LGE, article 69). It is presided over by the principal and includes vice-principals, and representatives of the teachers, students, alumni and parents. It is important to mention that, according to law, participation by parents can be foregone "[...] when the institution's 42 What can a Regional Government do to improve education? The Case of Junin characteristics justify it" (LGE, article 69). Other community institutions can be part of the CONEI when invited by its members. Single teacher and multi-grade schools in neighbgoring communities are organized in school networks for mutual support, and in that case one CONEI is formed for each network instead of separate CONEIs for each school. Despite the legal determination to strengthen pedagogical and administrative school autonomy, we found that a common element in the schools visited is the lack of decision-making power regarding administrative aspects that are fundamental to improving education quality. There is a gap between the legal text and practice. The schools' demand for more autonomy If certain key administrative and pedagogical decisions were to be actually made by schools, instead of having to sort through endless bureaucratic hurdles, the number of unattended needs at the school level would likely be reduced, along with the currently high dissatisfaction expressed by principals, teachers and parents. Among the points mentioned in the interviews, there are three central issues that need to be addressed to strenghten school autonomy: personnel, resource management and textbooks. According to LGE, the role of schools is to: "[...] participate, through the Institutional Education Council, in the evaluation of teaching and administrative personnel for entry, promotion and permanence purposes. These actions are conducted jointly with intermediate management instances and according to specific normativity" (article 68, section k). In practice, schools have a very small margin to influence the hiring, promotion or firing of their personnel. Of all principals interviewed, 34 percent mentioned their desire to have more autonomy to manage their teaching personnel, in particular with regard to hiring, training and sanctions. It is also important to point out that testimony gathered at more than one UGEL indicated that principals illegally charged teachers for the right to work at their schools (around 1,800 soles a year), which shows that the problems associated with weak accountability also effect education management at this level. The ability to define and manage an Box 3. The journey of a broom operating budget will allow schools to The long journey through the education bureaucracy can be cover emergency expenses, purchase illustrated by describing the procedures for a school to obtain a materials and improve infrastructure and broom. According to the fieldwork interviews, it goes more or equipment. According to the LGE, schools less like this: the school principal asks the pricing unit for a must also "[...] formulate, execute and broom. The pricing unit asks for an estimate from the supplier evaluate the institution's annual budget" and then takes this estimate to the UGEL, which in turn takes the (article 68, section g). One in four request to the DRE. The DRE issues a check to the supplier, who principals aspire to administrative sends the broom to the UGEL, where it is stored until the director of the pricing unit can pick it up. When this person has the time, autonomy and direct resource he or she will take the broom to the school that requested it. management. Meanwhile, another school that does not need a broom gets one because it is easier to make bulk purchases, but is left without the Giving schools effective autonomy to chalk that had requested. formulate and execute their budgets as indicated by law, would contribute to increasing efficiency and allowing faster responses to operating needs at schools. But the problem lays both with autonomy to manage resources and access to these resources, for autonomy is of little use to schools if they must administer their poverty. Another of the obstacles indicated by the principals and teachers in the schools visited is that they are unable to make decisions regarding the procurement of textbooks that are relevant to their students' reality, particularly in rural rainforest schools. In this matter, the LGA stipulates that it is the school's competence to "[...] diversify and complement the basic curriculum, conduct tutorial actions, and select textbooks and education materials" (article. 68). In practice, this has yet to happen. Rural teachers are of 43 What can a Regional Government do to improve education? The Case of Junin the opinion that the books elaborated and distributed by MED are very unsuitable to their students' reality, which hinders rather than enable the learning process. Empirical international evidence on the impact of decentralization suggests that simply changing education's organization --for example, creating school boards-- has little if any impact on the provision of education services. It is the practice of new responsibilities what generates real impact, and the efficient exercise of these responsibilities depends on the training, capacity and incentives for school personnel, and the resources assigned to the schools (Winkler, 2006). A shared vision and clear objectives The first set of school efficiency factors is the existence of a shared vision, objectives and goals. According to international standards, efficient schools make an effort to establish clear and concrete goals and objectives. The institutional plans of these schools are characterized by their realism and by the presence of objectives that can be understood by everyone. In addition to concrete goals, efficient schools establish a hierarchy of objectives that clearly distinguishes between principal and secondary objectives, which facilitates pedagogical decision making (UNICEF, 2004). Nevertheless, four years after the General Education Law mandated that administration at schools must be guided by the Institutional Education Project (PEI), only a third of the schools visited had prepared their PEI. What is worse, almost half the schools have no written institutional objectives. With regard to learning objectives, 40 percent of these schools have yet to elaborate a curricular plan (and the visit took place in May, almost three months after the start of the school year). Preparing a PEI takes time, effort, resources and capacities that many principals or teachers do not have or are unwilling to invest. This is exacerbated by the fact that they receive limited support from education authorities to implement good planning. Nevertheless, the PEI is mandatory. Moreover, submitting institutional documents to the UGEL is the primary (if not only) criteria for evaluating the performance of school principals, regardless of the quality of the plans proposed or the progress made in their implementation. All of this encourages the delivery of mediocre and poorly prepared products. In practice, the process to prepare the PEI rarely involved the whole education community as stipulated by law. We have gathered numerous testimonies that indicate that even when the PEI is signed by the members of the CONEI and parents' representatives, it is not safe to assume that they have participated in its elaboration or even read the document. According to reports from the UGEL, it is common to find the schools that copied or purchased a PEI from another institution to comply with the norm and have all of papers in order. In all of the UGELs we were told that they received the same documents from different schools. Although the schools have institutional plans, pedagogical objectives and long-term visions, these do not necessarily translate into clear guidelines for education practices and an ability to evaluate progress. In fact, one of the main obstacles to translating these objectives into practice is their own phrasing. As we have already mentioned, the key element to school efficacy is the formulation of clear, concrete, and realistic goals that are understandable to every stakeholder. On the contrary, the objectives that we found in the PEI that we reviewed tend to be very complex ("Promote active investigation in the context of the values that form a conception of life for the student and society, and help provide a coherent interpretation of faith"), abstract ("Propulgate [sic] a love for truth in everything we do"), imprecise ("Promote and practice values, attitudes and abilities") and, above all, ambitious ("Generalize equality of opportunities for all Peruvians to access a quality education" or "Provide an instruction that meets education challenges, committing fully to change our society"). Consequently, teachers and principals have no concrete tools to evaluate the institution's achievements or their own work. 44 What can a Regional Government do to improve education? The Case of Junin In their mission statements, visions and objectives, many schools pledge a willingness to form leaders and autonomous citizens, with high self-esteem, who are capable of getting ahead, and are equipped with the right tools to do so. This pre-fabricated discourse is repeated by teachers with is no evidence that they ever applied it in the classroom. Time and again, we find deep incongruence between what people say they wish to achieve and the predominant pedagogical and disciplinary methods. In the classroom observations, we saw many instances where teachers who had spoken of teaching children to be "leaders," "creative" and "competitive" just minutes before, refused to allow students to give their opinions in class, resorted to threats to maintain order, limited to present class contents, or requested that students repeat or copy what was written on the board. Futhermore, it is more difficult to put an objective into practice if you are unfamiliar with it in the first place, and according to our experience, even principals are often unaware of the school's objectives. If we compare the objectives given by the principals in the interviews and those extracted from the PEI, we find no relation between the two in most cases. 5.3. Conclusions Although Peru has made important progress in decentralization over the past few years, schools have yet to exercise the functions assigned to them by law four years ago. A significant portion of the problems schools and teachers face in producing good education outcomes can be traced back to management deficiencies. The main causes identified in this chapter are: (i) the lack of school autonomy to make decisions about important aspects of institutional life; and, (ii) the lack of effective pedagogical and institutional support from the UGEL and other education management instances. The current legal framework (see annex 4.2) allows the Regional Government of Junin to assume leadership in efforts to deepen school autonomy, if it has the political will and a clear plan. 45 What can a Regional Government do to improve education? The Case of Junin Chapter 6 Recommendations to a Regional Government Committed to Education As noted in previous chapters, the region of Junin has made significant progress in building schools across its territory; hiring teachers with teaching titles (93 percent); reaching high completion rates for secondary education (78 percent); and increasing the education budget (44 percent since the year 2000). Nevertheless, serious problems persist such as the poor maintenance of school infrastructure; the inadequate distribution of teaching personnel within the region; the inadequate distribution of non-salary expenditure across provinces and districts; and the low quality of primary education, as evidenced by the vast majority of students not meeting minimum learning standards. Within this situation of generalized poor quality education, we also find persistent inequality. All of these problems are related to deficient management and could be solved by doing the following: (i) establishing a clear commitment to measurable goals; (ii) offering pedagogical and financial support to schools to reach these goals; and, (iii) regularly measuring the schools' progress and informing the community. These recommendations entail granting more power to the schools, which requires a commitment from the Regional Government to assign more resources to them, give them more autonomy to manage them, and provide them with the necessary technical and pedagogical support to reach short-term learning goals and ensure minimum, dignified operating conditions. For this purpose, the Regional Government should monitor progress and establish an information system that reports on and evaluates progress made toward the goals set.30 A key element here involves incorporating criteria for equity that ensures that the schools that face the greatest disadvantages are not left behind. This implies, for example, more and more specialized support for single teacher schools or those serving native populations. 6.1. Declare measurable goals In this study, we have identified three key areas that need immediate attention so that concrete, feasible goals can be set in the short to medium term. First, students' reading performance is very poor and uneven, which carries out problems for future learning. Second, the net enrollment rate reached in preschool education falls below the projection based on expenditure, which reveals a problem of inneficient resource 30We have chosen not to recommend measures related to human resource management because there are political economy factors that make this an extremely sensitive issue, one of national relevance and long-term solution. As such, we prefer to focus our attention on management issues that can be addressed by the Regional Government in the short to medium term. 46 What can a Regional Government do to improve education? The Case of Junin management. Third, the poor conditions of school infrastructure, as well as the precariousness of equipment and the scarcity of pedagogical materials, hinder education efforts; the conditions of the sanitary services, roofs and walls are often unacceptable and occasionally put the students' and teachers' integrity at risk. We recommend that the Regional Government set one realistic, attainable goal to address each of these issues, stated in a way that all members of the education community can understand (including parents with low levels of education), and therefore monitor. The Regional Government could stipulate, for example, that: (i) By the end of the 2008 academic year, there will be a 5 percent increase --relative to 2007-- in the proportion of students that reach a sufficient level of proficiency in the Ministry of Education's census test for both Communications and Logic-Mathematics, and none of the students will score zero. (ii) In three years, there will be 15 percent increase in the proportion of 3 to 5 year olds that has access to preschool education without exceeding the maximum number of students allowed per class at this level. (iii)By the inauguration of the 2010 academic year, all schools will have adequate and sufficient sanitary services and sinks; the classrooms will have their roofs, windows, chairs and desks in good condition and these inputs will be properly maintained throughout the school year. Reaching short-term goals like these will allow for an immediate mobilization for education and to strengthen the accountability processes to set goals in the medium term. Learning goals In 2006, MED initiated an important effort to evaluate second grade reading skills with the objective of consolidating a national evaluation system with universal coverage. The first year, the census covered 55 of schools, with the commitment to continue this evaluation every year until 90 percent of the schools and 75 percent of the students participate. These efforts include disseminating results in the school level by using an easy-to-understand report sheet for parents, teaching personnel and local authorities. The high coverage of this kind of evaluation will provide performance benchmarks for each school in comparison with others in the district, province and region. Annual assessment will make it possible to measure the progress of each school relative to the previous year. MED experts established the reading performance levels31 and the cut-off points to determine the reading level of each student. In the 2006 census evaluation, 18 percent of the students in Junin were situated in level 0, 29 percent in level 1, 30 percent in level 2 and 24 percent in level 3, which is the equivalent of "sufficient performance" (MED-UMC, 2007).32 Taking into account the census evaluation's structure and the achievement levels that have already been defined by MED, we propose using three measurement levels. The way that these criteria are observed varies according to the progress that the students are expected to make. The goal is to reach optimum levels in reading performance in half or more second grade students in 2010. To do this, we suggest that the school: (i) establishes a specific period of time for classroom reading, emphasizing the order of the words in a sentence, the order of a sequence of sentences, fluency, and answering questions about the reading; (ii) 31Items in the test were grouped according to their degree of difficulty in four achievement levels (level 3, level 2, level 1 and level 0), meaning that level 3 groups together the most difficult tasks. 32Annex 5.1 shows the reading tasks that students should be able to handle at each achievement level. 47 What can a Regional Government do to improve education? The Case of Junin identifies students that have reading comprehension difficulties; (iii) motivates students to find different solutions to problems that come up during reading (experiences to explore and research to find possible solutions using materials and objects that allow them to group, order, arrange in a series, count, make comparisons, experiment and estimate); (iv) implement a pair-reading system in which the students from higher grades read to younger children; (v) ask parents to read and test reading at home; and, (vi) organize reading workshops and venues for sharing with parents. Finally, given that students lag further behing as they advance in their schooling, it is also important to set standards for the fourth and sixth grades, and design appropriate instruments for their evaluation in the school census. This point is extremely important but must be addressed, in coordination with MED, in a strategy for the medium term. If the efforts to measure education quality are focused exclusively on the second grade, there is the risk of disregarding learning achievements in subsequent grades, which may remain below expected levels, or even worse, decline because all the attention was paid to second grade. Goals for enrollment and expenditure efficiency In the second chapter, we indicated that less than 42 percent of the children from 3 to 5 years of age attend a preschool education center or program in Junin. This figure is well below the national average and the results that could be obtained with a more efficient use of available resources. According to our estimates, the current investment in preschool education in Junin should lead to a net enrollment rate of at least 56 percent. The main challenge is ensuring that the investment in this education level renders the results expected in terms of enrollment withouth increasing class size to more than 20 students per classroom. The challenge is focused on ensuring that the preschool education expenditure is efficient given that a 15 percent increase over three years would achieve a coverage level of 57 percent. This supposes conducting a financing study of preschool education service delivery in Junin that identifies the most efficient way to allocate current resources, before increasing spending at this level. This objective will guide policy planning and implementation to correct current inefficiencies without incurring in additional expenses and in a way that would allow the objective to be reached. Once these inefficiencies have been corrected, it will be pertinent to evaluate the possibility of increasing spending on preschool education. Finally, this goal will be hard to reach without an awareness campaign to encourage parents to send their children to a CEI or PRONOEI. Goals for operating conditions at schools We recommend that the Regional Government define and adopt minimum operating standards for schools as a medium and long-term strategy. This consists of a predetermined set of basic inputs and school resources that have been agreed upon with the community and which are considered minimum requirements for the school to be able to offer students adequate learning conditions. For this purpose, the following can be considered basic pedagogical inputs: textbooks and materials for students; equipment for schools, including desks, chairs and shelves; the physical characteristics of the school such as basic infrastructure (for example, the conditions of roofs and windows, square meters of classroom space per student); sanitary conditions (for example, potable water, restrooms, sinks) and school maintenance (for example classroom and school clean up). As a long-term design, we recommend: (i) defining minimum operating standards for the region's schools through a joint effort with the region and according to current standards in the country; (ii) using these standards to measure the reality of the schools in the region; (iii) preparing pluriannual plans for investment and technical assistance, giving priority to schools that are far from meeting standards; and, (iv) implementing investment plans with the school community once the resources are available in the 48 What can a Regional Government do to improve education? The Case of Junin regional level. To define minimum standards, MED's standards should be taken into account and efforts should be based on the region's reality and financial capacity to contribute additional resources to achieve the same. Given that it will take time and political commitment from the regional and central authorities to design and implement the aforementioned plan, it is important to produce a significant change sooner than later. In this sense, we suggest starting with what has been identified in this study as the more critical points: the sanitary infrastructure, the conditions of roofs and windows in classrooms and --to a lesser extent-- the state of students' chairs and desks. For this purpose, we propose setting a goal to ensure that at the beginning of the 2010 academic year, all schools have enough operative restrooms and sinks; all classrooms have roofs, windows, chairs and desks in good condition; and these inputs are maintained in working order throughout the school year. This goal is two-fold. First, it aims to ensure that students and teachers learn and teach in dignified, adequate conditions. In this sense, we believe that the Regional Government of Junin should declare unacceptable that some schools in the region lack drinking water, sanitary services and sinks, and adopt a goal that all schools in the region must comply with this norm by the year 2010.33 Insufficiencies in these minimum elements have negative consequences for student hygiene, health and learning; as such, we suggest that the Regional Government adopt this goal as soon as possible. The Regional Government can focus its school rehabilitation efforts on building better sanitary and sink structures by 2010, while each school can concentrate its efforts on: repairing its roofs and windows, fixing desks and chairs, and providing adequate maintenance throughout 2008-2010. This can be done with funds that the Central Government will tranfer to schools for classroom maintenance and repairs according to stipulations in the 2008 Budget Law. A goal such as the one suggested here can contribute to guiding the expenditure of the new resources received from the central level by schools. The second purpose of this goal is to put in place a monitoring system throughout the school year, given that, as seen in this study, the poor operating conditions in schools in Junin are due in great part to inadequate maintenance efforts. This monitoring system --which can consist in combining self-reports, perception surveys and random verification visits-- should be linked to a more extensive monitoring system for the use of funds delivered to schools for classroom maintenance. 6.2. Provide necessary pedagogical and financial support It is not enough to set goals for schools to meet. Instead, the Regional Government needs to commit to providing necessary pedagogical and financial support to schools to achieve these objectives.34 Pedagogical support We have seen that schools need significant support and accompaniment that is not being provided due to a lack of resources, capacities and in some cases willingness. For this reason, it is necessary to design a support system that is not affected by the bureaucratic and economic obstacles faced by the UGEL. Support must be directed to the schools that face the greatest obstacles: the most rural, the poorest, and the most remote, or in other words, the most abandoned. First, in order to provide trained technical assistance to improve reading comprehension results, we propose calling on a group of experts --qualified amautas-- that provide schools with techniques to teach reading. 33This also supposes adopting actions to rehabilitate sanitary services and sinks in schools that need it to ensure that these services are operative. 34The CNE is aware of this very need and has elaborated a proposal to implement a "national system for technical support and pedagogical accompaniment for teachers, principals and education institutions" in rural areas (Barletti, 2007). 49 What can a Regional Government do to improve education? The Case of Junin This call for experts should be extended to a wide range of professionals (institutes, NGOs, universities) and criteria for selection should guarantee the quality of the assistance requested. To increase efficiency, schools could evaluate the amautas' work. This proposal coincides with CNE's call (Barletti, 2007) for a pedagogical accompaniment scheme that is conducted by a qualified teacher or trainer (who has a good reputation and is equipped with the necessary professional competences). This scheme should be applied frequently and extensively (weekly visits and long work days). Second, pedagogical support should include a design for teacher guidelines and materials for reading instruction, and should focus specifically on teaching reading to multi-grade classrooms, bilingual populations and students with special needs, which are the areas widely lacking from pre-service teacher training. Third, in addition to direct pedagogical support, another type of collaboration, equally as important, consists on reducing the workload of teachers working in rural areas, eliminating overpopulated single teacher schools (for example, making them into multi-teacher multi-grade schools). As we have seen in previous chapters, there are problems with the distribution of teachers in the region. By using the legal framework and current rationalization system for teaching posts, the Regional Government can solve the most serious problems that the sector faces by ensuring that overpopulated single teacher schools become multi-teacher full grade schools as soon as possible. 50 What can a Regional Government do to improve education? The Case of Junin Box 4. Solaris: improving reading with simple strategies and everyone's participation Experience has shown that there are inexpensive, quick and motivating ways to generate change. An inspiring example is the recent experience of the NGO Asociación Solaris Perú (Solaris), in charge of implementing the project for Intervida Education Centers 2004-2006 in the country. Solaris decided, in agreement with the principals of the Intervida schools, to evaluate the reading performance of their students. Six schools in Trujillo, Chiclayo, Andahuaylas, Ilave and Arequipa participated in this initiative. In the first evaluation, conducted in August of 2006, students showed considerable difficulties in terms of both fluency and comprehension. In the school in Ilave, second grade students read an average of 27 words per minute and in Arequipa, 53. Reading comprehension oscillated between 29 percent (in Ilave) and 65 percent (in Andahuaylas). Once these schools realized that their students' reading levels were low, they designed a set of strategies to improve fluency and reading comprehension. The strategies implemented included: (i) establishing a specific period of time for classroom reading; (ii) reinforcing knowledge and management of reading comprehension strategies in self-training teacher meetings that can subsequently be applied in the classroom; (iii) working in teams by school cycles (grades) to coordinate the implementation of reading strategies; (iv) implementing reading in pairs, which consists of students from higher grades reading to younger children: (v) providing guidelines and venues in the school to familiarize parents with reading techniques and involving them in efforts read with their children at home; and, (vi) encouraging parents to measure reading at home to verify their children's progress. In the school in Andahuaylas, a campaign was held to collect books and expand the library. Additionally, each of the schools informed parents about the results of the first measurement and included parents in efforts to support reading by: (i) more frequent use of the library; (ii) allocating time for reading at home; (iii) measuring reading at home; (iv) visits to the classroom to see reading work; and, (v) reading workshops and venues for sharing with parents the progress and concerns. Three months later, a second measurement showed a significant increase in reading fluency: children read 25 percent more fluently in the second grade and 135 in the fourth. Comprehension of the passsage read improved by 38 percent on average in second grade and 21 percent in the fourth. Additionally, students were visibly more confident in their performance. Among the lessons learned, the Asociación Solaris's report indicated that it is key to have a friendly, clear standard for reading measurement that is understandable by children, parents, teachers and principals. Second, parents become involved in their children's learning when objectives are clear and concrete, which is the case of reading measurement. Third, the fact that principals and students have a source of comparative verification of the effect that pedagogical work has on student progress in reading comprehension facilitates accountability processes. Fourth, coordinated and organized work between teachers and parents promotes a positive reinforcement of the schools' efforts. Fifth, the self-training and continuous training that the teachers had to implement strategies and resources contributed favorably to improving reading skills. Sixth, the students begin to acquire the habit of reading when the strategies implemented become part of the daily life. Last, reading measurement is a good way to initiate an overall intervention in the school, as it mobilizes all stakeholders and provides a concrete, tangible point of reference for anticipated achievements. Financial Support We suggest giving each school a budgeted amount per student. These resources can be spent on goods, services and maintenance, and should be distributed based on transparent criteria with built-in compensation mechanisms to favor the smallest, most disadvantaged schools. Additionally, it is necessary to strenghten the resource management capacity of administrative and monitoring personnel.35 School principals, as well as members of CONEI and the APAFA should be trained in administration, financial management and accountability. Ensure more efficient administration The per student budgeted amount allocated to schools will be used to finance the purchase of expendable material36 and other operating expenditures,37 including school maintenance expenditures. We suggest the 35 International evidence provides many examples of resource transfer to schools. In all cases, these transfers are clearly regulated, as are accountability processes, expenditure verification and audits. 36 This refers to the materials that enable learning and teaching in the classroom: (i) expendables, such as chalk, paper of different colors and textures, etc.; (ii) visual, such as slides, maps, signs and others; (iii) toys and educational games for preschool and first grade of primary; (iv) teaching aids such as spheres, compasses, geometry games, abacuses, story books and others; and, (v) sports equipment, such as balls for different sports, nets and others. 51 What can a Regional Government do to improve education? The Case of Junin Regional Government makes a single deposit for expendable material for the nine-month period that these establishments operate. These funds should be executed in a timely fashion throughout the school year while the monies destined to operating costs should be deposited throughout the twelve months of the year. In order to make this transfer operative, a decision must be made regarding who will participate in the school's administration. The Regional Government should be responsible for making monthly deposits to the school board to cover operating costs, excluding items for personnel services. The relation with the bank will begin when a current account is opened by the president and treasurer of the school board. We suggest that this account be used exclusively for transfering deposits and making payments for the aforementioned expenditures. The school board would have the following duties: (i) managing money according to the regulations that the Regional Government determines for this purpose; (ii) issuing monthly checks38 for expendable material and other expenditures; (iii) performing a monthly check of income and expenditures; and, (v) being held accountable to the Regional Government and the education community on a regular basis. In order to do this, parents needed to be equipped with the knowledge and abilities necessary to participate in school management in a way that enables them to identify and prioritize needs, look for support, and jointly solve problems that arise in the school or affect the school. Evaluate the approximate cost, destine the necessary resources and distribute them equitably As noted before, the current average expenditure on goods and services is 13 soles per student per year. We propose increasing the average annual non-salary expenditure in the region of Junin and transfering these amounts directly to each school. Peru has already made progress in this direction. According to article 10, section 1 of the General Budget Law of the Republic for the 2008 fiscal year, 270 million dollars will be directly transferred to schools for classroom maintenance. The allocation of these funds will be based on the number of classrooms in each school on a scale that grants a larger amount per classroom to schools with fewer classrooms to ensure that the allocation process does not put smaller schools (typically rural) at a disadvantage. The amounts allocated to each school are shown in annex E of the law. The law also mandates the creation of an oversight committee at each school comprised by the mayor of the district, the school principal, and the president of the APAFA (Parents'Association). This committee is responsible for issuing a report to the UGEL on the results of the use of the money tranferred. 37This is understood as operating expenditures other than those that the board must make to pay its checks (including bank fees and check fees) and expenditures for infrastructure, furniture, and equipment at the education center. 38The suggestion to set up sole bank account in which all the income and expenditures received by the schools through this transfer can be recorded is meant to facilitate auditing given that the current account will contain the complete income and expenditure history. In rural areas, the Banco de la Nacion (National Bank) can be used. 52 What can a Regional Government do to improve education? The Case of Junin In this context, the Regional Government has to redefine its role to ensure that the resources assigned to schools are used in an efficient, pertinent way, for example: (i) establishing the minimum operating standards to guide schools for the use of maintenance funds; (ii) strengthening management and maintenance capacities at the school level; (iii) evaluating the efficiency of the expenditure of these funds and measuring the results obtained; and, (iv) informing the community about the resources delivered to each school and how they were used. 6.3. Regular progress evaluation and dissemination of results The political and technical experts responsible for education in Junin and the authors of this report are convinced that it is possible to improve the quality of education in the region. To reach this objective, we need instruments to measure progress in all the aforementioned areas. These instruments must be used within a framework that promotes transparency, dialogue and cooperation. We suggest that schools be given more power to strengthen or empower the community, to achieve more social control over the school, and to promote more cooperation with and support to teachers. This recommendation is based on several factors. First, the community needs to feel that it has rights over its Box 5. How can we ensure good financial management in schools? International experience The United States, the United Kingdom, Australia and Canada were the first countries to decentralize education administrative responsibilities to the school level. Over time, these models have also been replicated in other countries in the region. The main objectives of this kind of reform are to elevate learning achievements by improving the quality and efficiency of school administration; strengthening local capacities; and increasing the participation of the actors involved in the process such as the users (parents/students) and service providers (principals and teachers). These experiences have been developed in countries such as Mexico, El Salvador and Chile. Mexico: Support to School Management (AGE) This program consists of providing financial support and training on school administration to parents' associations (APF). Its implementation is focused on rural primary education schools and was gradually implemented during the 1990s. Financial support consisted in a quarterly transfer to the APF account for an amount that varied between 500 an 700 dollars per year, depending on the school size. Expenditure execution was subject to annual financial audits of a random sample of schools. Additionally, the use of funds was limited to specific areas that were stipulated in an operating manual so that the money would not be spent on salaries for teachers. In this context, the majority of funds should be destined to infrastructure and maintenance work. In 2003, AGE's intervention was complemented by improvement in the APF's capacity to manage funds; increased parent participation in the schools; and provision of information on student learning and guidelines on how parents can help their children improve their learning. In 2005, more than 45 percent of primary schools in Mexico had an AGE (Banco Mundial, 2006a). Central America: School-centered administration El Salvador (EDUCO), Guatemala (PRONADE) and Honduras (PROHECO) have implemented decentralization reforms that include school-centered administration. In these models, the community uses entities that have been legally set up by members to select and pay teachers, control attendance and performance of both students and teachers, manage the funds assigned for school materials, and participate in activities to improve schools. The respective ministries contract parent associations to provide education services to the community, transfer funds to them with clear instructions regarding adequate use, and offer teaching materials, supervision and training. These associations select, hire and supervise teachers; build and maintain schools; and mobilize the support that the community provides its schools. District supervisors encourage the community to participate by providing parent associations with administrative and technical support, offering technical assistance to teachers, lending support for opening new schools, and conveying information regarding existing programs. These models contemplate periodic evaluation of administrative and education matters by ministry personnel, and the parent associations are obliged to allow and collaborate with all internal and external audits that the ministries choose to conduct. Ministry personnel review all expenditures made by the parent associations prior to authorizing new transfers (Di Gropello, 2006). Chile: Education subsidies In the early 1980s, the government decentralized school administration by transferring administrative responsibilities for public schools from the Ministry of Education to the municipalities. It also changed the financing scheme for education. Public schools are still financed by the central government but municipalities have begun to receive a per student subsidy for each child enrolled in their schools. Consequently, student absenteeism has a direct effect on education budgets. More importantly, private schools that charge tuition started to receive the same subsidy per student than public schools. The subsidy per student is delivered to schools every month based on average attendance, which is validated through random audits to schools. Approximately 92 percent of the students in primary and secondary school in Chile attend subsidized public or private schools (Consejo Asesor, 2006). 53 What can a Regional Government do to improve education? The Case of Junin own school if participation is to be strengthened. Second, if the board is made up of community members, it is physically more available to make daily visits to schools in order to monitor teacher attendance, verify the establishment's logistics and maintenance needs, and discuss aspects that need to be solved either on a weekly or more frequent basis. Third, community members have more interest in developing ties between the school and the community given that social control would be more direct. Finally, joint efforts between parents and teachers can drive improvements in quality, particularly in rural areas, where teachers work in isolation and without resources.39 The most effective way to take advantage of the population's active participation is by establishing an information system to ensure compliance with the minimum operating standards stipulated for each school. These reports should also be presented on a periodic basis. Also, to ensure accountability relative to the economic administration of resources, these reports should include the most information on expenditures. In order for this to work properly, it is necessary to create and publicize a system that provides periodic reports on the school's performance relative to the goals set for learning and minimum operating conditions. Both reports should be based on data collection mechanisms that ensure updated information is available for the most relevant time periods. Reports on reading skills We suggest setting up public bulletin boards in schools to show the school's progress with regard to first graders' reading comprehension. It would also be a good idea to publicize at local radio stations the performance of innovative, exemplary teachers. The public bulletin board could contain the information included in table 5-1, which shows the results obtained by each student in each evaluation, and indicates the percent distribution of students in the optimum, medium and low categories. For comparison purposes, the table also includes the percentage of students at the district and regional levels that reach the optimum level. We recommend diversifying the mechanisms to present results to parents, students, communities and society in general to ensure that this is a gratifying, stimulating contribution to pedagogical work, and it is not perceived as a purely administrative function of control. For this purpose, it is important to organize cultural events where students can show off their reading abilities (poetry readings, letters, original recipes, biographies, etc). Parents or community members can also participate as readers in order to make the activity more amenable while bringing the school closer to the parents and the community. It is important to mention that MEF's results-based budgeting approach sets learning goals for second grade in Language and Mathematics. The Regional Government of Junin has developed a set of indicators that includes the percentage of students achieving a sufficient reading comprehension level at the end of the second grade of primary education. The aforementioned measurement and the emphasis on reading comprehension are consistent with this metric. 39 We might consider a more aggregated participation scheme, for example through a board with community members at the UGEL level or in the rural networks, which can lower operating and maintenance costs. Nevertheless, this increase in efficiency may mean that the community participates less and the quality of the services is lower. With a greater aggregation level, each school will have one community member on the board (and without a doubt some schools may have no board members). The participation of the entire community will depend on whether or not this individual channels information, reducing as such the probability that the members of the community will be actively involved in school life. Worse yet, school personnel will not receive enough stimulation form the community to provide better quality services, improve the way students and parents are treated, and increase community involvement among others. 54 What can a Regional Government do to improve education? The Case of Junin Table 6-1. Report on second grade reading comprehension School District Region of Junin Measurement Students Optimum Medium Optimum Students Optimum Students evaluated level level Low level level evaluated level evaluated First measurement 33 15% 27% 58% 30% 323 25% 5.000 Second measurement 33 50% 30% 20% 60% 315 55% 4.900 Third measurement 30 80% 12% 8% 90% 310 87% 4.787 Table 6-2. Measurement scale for the physical environment Scale Physical environment Optimum Fulfills 7 Medium Fulfills 5 o 6 Low Fulfills 4 or less Table 6-3. Report on the school's physical environment School District Region of Junin Percentage of Percentage of Measurement School qualification schools at an Schools schools at the Schools optimum level evaluated optimum level evaluated First measurement (April) Optimum 60% 10 50% 100 Second measurement (June) Low 66% 10 55% 100 Tercera medición (September) Medium 65% 10 60% 100 Reports on minimum operating conditions One of the aspects of the process to monitor minimum operating conditions, for example, should relate to the quality of the physical environment where children learn. To obtain this information, it would be recommendable to take a tour of the installations and observe the surroundings with previously established criteria in mind. Table 5-2 shows a possible evaluation scale if seven criteria are considered. In this case, the individuals in charge of the schools, along with a group of parents, will review the indicators in the months of April, June and September and the follow-up teams from the Regional Government will do the same on each visit to the school. Once the mechanisms for information registry are defined, they should be displayed in a public place (on a bulletin board, for example). Each report should show the results for reading comprehension of the students who took the test (as shown in Table 5-1 for example) and the percent distribution of students in the categories defined to determine the school's operating conditions (for example, among optimum, medium and low levels as shown in Table 5-2). For comparison purposes, the results for each district and the region of Junin can be included. Reports on financial resources It is necessary to create information systems to report regularly each school's finances in order to ensure the school is held accountable to the Regional Government and the community for money received. The individuals in charge of each school should present a half-year report for each transfer received, along with the respective liquidation, to the Regional Government and the Community. The funds administered by the schools will be subjected to internal or external audits by the Regional Government. For this purpose, the school administratives must provide, at the request of auditors, all documentation related to fund transfers in order to verify adequate use. The school must, within the specified time period, return any funds that the audit has listed as missing. The individuals responsible should hold at least two meetings a year with the entire education community in order to present the budget plan for these items 55 What can a Regional Government do to improve education? The Case of Junin and the results obtained through their execution. Financial and operating reports should also be provided at the request of the Regional Government. In terms of the school budget, the individuals in charge of the school should keep a monthly log of budget execution that is in accordance with the activities prioritized in the school plan. For this purpose, it is necessary to evaluate the following: (i) the transparent administration of resources; (ii) filing income and expenditure receipts, separating fiscal records from others that, due to their size, are not subject to fiscal scrutiny; (iii) a registry of income and expenditures according to norms that the Regional Government has issued for this purpose, (iv) the financial report for the education community and the Regional Government; and (v) any documentation on the accountability process to the education community and Regional Government regarding the goals obtained by the center's management. Dissemination of information aggregated by districts, provinces and the region We recommend creating periodic financial and education results systems by district, UGEL and DRE Junin that include a report on access to preschool education. Given the priorities that the Regional Government has set in terms of education administration in the region, it is important that the authorities at the intermediate bodies (local governments, UGELs and DRE) generate and disseminate aggregated information regarding their areas of intervention. 56 What can a Regional Government do to improve education? The Case of Junin Anexos 1. Annexes in Chapter 1 Annex 1.1. Accessibility of Public Services in Junin Annex Figure 1.1-a. Population according to the time they take to get to the nearest public school from their homes by their usual means of transportation (% accumulated). Population (%) 100% 99.1% 91.2% 75% 75% 54.4% 50% 25% Primary school Secondary school 0% 1 5 10 15 20 25 30 35 40 45 50 1 3 5 7 9 11 13 15 17 19 21 24 Minutes Hours Time to get to school from home Annex Figure 1.1-b. Annex 1.1-b. Population according to the time it takes them to reach other public facilities from their homes by their usual means of transportation (% accumulated) 56 57 What can a Regional Government do to improve education? The Case of Junin Population 100% 92.2% 90% 86.5% 80% 82.2% 70% 60% 50% 50.4% 40% 39.8% 33.0% 30% Health post 20% Police station 10% Municipality 0% 1 5 10 15 20 25 30 35 40 45 50 55 1 4 8 12 16 20 24 48 Minutes Hours Time to get to the establishment from home Annex 1.2. The Ashaninka war and the spiral of violence in the rainforest In the central rainforest, the war period began with a major setback for the revolutionary movement Tupac Amaru (MRTA), which had been shaken by the capture of its leader Victor Polay in Huancayo on February 03, 1989. After an attack on the police post at Pichanaqui, a column of thirty heavily armed subversives operating in the area of Satipo and Chanchamayo began to displace towards the sierra to join up with another column that was descending from Tulumayo in the highland section of Junin. The objective was to take joint action to seize the city of Tarma. Nevertheless, on April 28th of 1989, both columns met up with Peruvian Army forces in the district of Molinos (Jauja) and were practically annihilated in combat. This stopped MRTA's advances in this part of the central rainforest and left the door open for the expansion of the Communist Party of Peru - Sendero Luminoso (PCP-SL). In the sector of Oxapampa, MRTA suffered another setback. At the end of 1989, PCP-SL began to fight with MRTA for territory. In the months of October and November, residents discovered the presence of PCP-SL in an outlying area of the district of Puerto Bermudez y Palcazu. On November 19th, an armed unit of PCP-SL attacked the installations of the Special Pichis Palcazu Project in Puerto Bermudez (PEPP). The unit looted some area stores and robbed the Agrarian Bank. However, in the end MRTA's actions expelled the Shining Path from the area. In December of 1989, in the district of Puerto Bermudez, province of Oxapampa, MRTA soldiers assassinated Alejandro Calderon, a prestigious native leader and President of the ANAP (organization that represented 52 native communities from Rio Pichis and its tributaries). Before Calderon's death, his son, Alcides Calderon, called for an uprising of the Ashaninka army. MRTA left the area prior to combat, but the Ashaninka army continued to mobilize and expand its field of action by forming alliances with the Sinchis40 of Mazamari that extended to Pichanaqui in Satipo. 40The Sinchis were a supposed counter-terrorist unit of the Civil Guard created during Belaunde's first administration (1963­68), which played an important role in combating the 1965 insurgency. In his second administration (1980-85) Belaunde expected the Sinchis to play a 57 58 What can a Regional Government do to improve education? The Case of Junin The Ashaninkas of Oxapampa occupied the basins of Pichis and Palcazu, setting up checkpoints and searching travelers. To them, the colonos--who had previously taken over territory-- were generically also considered terrorists. To settle old scores, the Ashaninka rose up to kidnap, torture, rape, detain and disappear or assassinate colonos, causing many families to flea. The Ashaninka Army's progress was slowed down by protests and mobilizations by coffee growers' associations in the district of Pichanaqui on the right-hand bank of the Perene River. The Ashaninkas suffered a different fate at the Rivers Ene and Tambo, an area held by PCP-SL. Throughout 1989, PCP-SL exercised control in this area and generated intense violence that destroyed native communities and colonist holdings. This turn of events was primarily due to the presence of an anti-subversive base set up at kilometer 4 of the Marginal Highway in Satipo at the end of 1989, propitiating the reaction of natives and colonists against PCP-SL. Sixty-three percent of all killings and disappearings reported to the Truth and Reconciliation Commission (CVR) in the central jungle occurred between 1989 and 1992. Some communities that opposed PCP-SL moved to safer areas. This was the case of the communities in the Valley of Tsiriari that sought safe haven in Mazamari after PCP- SL had assassinated fifteen of its ronderos in Dos de Mayo. Three indigenous leaders were also killed --decapitated after being tortured for opposing PCP-SL methods to gain community support-- which unleashed the reaction of the Ashaninka sector known as the ronda de Poyeni, made up of natives from Tambo River. This group was smaller than the Ashaninka army from Oxapampa. From 1990 on, the ronda de Poyeni battled the main PCP-SL army unit, which was also under pressure from the Navy. In 1991, the Army urged the population to organize into counter-terrorist groups and set up some bases in strategic areas. Patrols were intensified in emergency zones. Additionally, the native rondas, which had taken on the PCP-SL at their own initiative, received support from the Army. From 1991 on, the Ashaninkas from the Ene and Tambo Rivers organized into self-defense committees to fight PCP-SL. In 1991, another counter-terrorist base was set up in Cutivireni, district of Rio Tambo. Next, the Army did the same in the Esmeralda and Natalio Sanchez valley, also in the district of Rio Tambo, all of which were located on the right-hand bank of the Ene River. But not all of the native communities rejected the Shining Path. Many also became support bases for PCP-SL. PCP-SL joined with these communities to fight the rondas and the armed forces, forcing as they had done in the valley of the Apurimac River massive displacement towards the mountains. This was the beginning of one chapter in a confusing and particularly prolonged war. In the mountains, subversive columns were protected and fed by native populations. The army had a hard time locating these units and was limited by the fact that the mountains could not maintain numerous populations without great effort and enormous sacrifices and difficulties. The Ashaninkas' permanence in subversive efforts became increasingly less voluntary. MRTA, after several defeats, reinitiated its actions on the Perene River. Its columns were reinforced by a division brought from the northeast. MRTA's leadership probably chose to focus on the central jungle due to the difficulties counter-terrorist forces were experiencing in their attempts to control the mountains. In any case, MRTA regrouped and decided to mount large-scale actions: on December 1, 1991, MRTA attempted to take the community of Pichanaqui but failed. In January of 1992, the Army destroyed one of the group's camps in the Mazamari, killing nine subversives. similar role in combating Sendero Luminoso in Ayacucho. The Sinchis were trained to fight classical insurgents, something the members of Sendero were not. Frustration and impotence led to abuse of the population. 58 59 What can a Regional Government do to improve education? The Case of Junin After these setbacks, MRTA revaluated its geographic strategy. In mid April of 1992, it decided to establish to relatively stable secondary fronts to conduct political indoctrination activities with population groups, particularly colonists. In this context, they planned this time successfully-- a new attack against Pichanaqui on the anniversary of their defeat in Molinos. Success prompted MRTA to undertake larger scale actions. On May 1st of 1992, around 100 subversive attacked and destroyed the anti-subversive base in Villa Rica, Oxapampa. The Army responded, and in the weeks that followed a series of battles took place. But MRTA also began to feel the pressure of the rondas. At the end of 1992 the war raged on in the region. Source: (CVR, 2003). 59 60 What can a Regional Government do to improve education? The Case of Junin Annex 1.3. Socio-economic and Ordering Characteristics Socio Economic Characteristics Region Population Regional GDP Regional GDP per Household per Total Extreme 2005 2005* capita (2005) Ranking capita income Ranking Poverty* Ranking Poverty* Ranking Amazon 389,700 699 1,794 24 215.25 20 59.1 16 20.0 13 Ancash 1,039,415 5,037 4,846 7 283.33 12 42.0 9 20.8 14 Apurímac 418,882 839 2,002 23 193.48 22 74.8 21 39.7 20 Arequipa 1,140,810 7,562 6,628 4 466.12 4 26.2 6 3.4 5 Ayacucho 619,338 1,428 2,305 22 188.33 23 78.4 23 41.3 21 Cajamarca 1,359,023 5,362 3,946 10 216.66 18 63.8 17 29.0 17 Callao 810,568 - - - 692.42 1 - - - - Cuzco 1,171,503 3,839 3,277 15 227.54 17 49.9 11 22.8 15 Huancavelica 447,054 1,385 3,097 19 145.41 25 88.7 24 72.3 24 Huanuco 730,871 1,856 2,540 21 157.84 24 74.6 20 48.6 23 Ica 665,592 3,341 5,020 6 438.72 6 23.8 4 1.3 2 Junin 1,147,324 4,529 3,947 9 306.62 10 49.9 12 16.5 10 La Libertad 1,539,774 6,627 4,304 8 329.19 9 46.5 10 18.2 12 Lambayeque 1,091,535 4,064 3,723 12 436.98 7 41.1 8 9.5 8 Lima 7,819,436 79,314 10,143 2 677.37 2 25.1 5 1.4 3 Loreto 884,144 3,353 3,792 11 215.72 19 66.3 18 33.9 19 Madre de Dios 92,024 316 3,434 14 299.06 11 21.8 3 4.3 7 Moquegua 159,306 2,191 13,756 1 440.77 5 27.3 7 3.9 6 Pasco 266,764 1,560 5,846 5 239.84 15 71.2 19 31.0 18 Piura 1,630,772 5,280 3,238 17 263.31 13 54.0 13 13.1 9 Puno 1,245,508 3,348 2,688 20 244.14 14 76.3 22 41.6 22 San Martin 669,973 2,192 3,271 16 211.11 21 54.3 15 17.2 11 Tacna 274,496 2,351 8,563 3 560.46 3 19.8 2 3.3 4 Tumbes 191,713 692 3,607 13 353.80 8 15.8 1 0.4 1 Ucayali 402,445 1,295 3,217 18 232.93 16 54.0 14 22.8 16 PERU 26,207,970 148,458 5,665 285.68 44.5 16.1 Sources: MEF; UNDP 2006a; INEI, ENAHO 2006. * Millions of new soles in 1994. Lima included the Constitutional Province of El Callao. 60 61 What can a Regional Government do to improve education? The Case of Junin Annex 1.4. Dictionary of Statistical Indicators A. Socio-economic characteristics Economic Concepts GDP per capita 2005 (new soles): Gross domestic product per person. Household per capita income (Nuevos Soles month): Monthly household income per person. Total Poverty (%): Percentage of the population living in households whose household expenditure per person is below the poverty line. Extreme Poverty (%): Percentage of the population living in households whose household expenditure per person is below the extreme poverty line. Geographic distribution R. coast (% pop.): Percentage of the population living in the natural region of the coast. R. sierra (% pop.):. Percentage of the population living in the natural region of the sierra R. jungle (% pop.): Percentage of the population living in the natural region of the jungle "Rurality" of the population and basic education: o Students enrolled (Preschool/Primary/Secondary): Percentage of the students enrolled in preprimary, primary or secondary schools located in rural areas. o Teachers (Preschool/Primary/Secondary): Percentage of teachers working at preprimary, primary or secondary schools located in rural areas. o Schools (Preschool/Primary/Secondary): Percentage of preprimary, primary and secondary schools located in rural areas. Location Relative to the Population Population near an EI Primary (%): Percentage of the population that normally needs one hour or less to reach the closest state primary schools. Population near an EI Secondary (%): Percentage of the population that generally takes an hour or less to reach the closest state school for secondary studies. Population far from a Secondary EI (%): Percentage of schools for Secondary education that are more than four hours from the district capital. Basic services Schools with all services (%): Percentage of schools that have electricity, potable water connections and a public sewage system connection. Schools with electricity (%): Percentage of schools that have electricity. Schools with potable water (%): Percentage of schools that have a connection to a public network for potable water. Schools with sewage system (%): Percentage of schools that have a connection to a public sewage network. Schools with no services (%): Percentage of schools with no basic services (no electricity, no connection to a potable public water system or public sewage). Sanitary services Schools with sanitary services (2004, %): Percentage of schools that have some type of sanitary service or restroom (2004 was chosen because is the last year for which different types of sanitary services were recorded). Schools with toilets (%): Percentage of schools that have some type of toilet. Schools with latrines (%): Percentage of schools that have some type of latrine. Schools with septic tanks or other sanitary services (%): Percentage of schools that have some time of septic tanks or some other type sanitary service other than the abovementioned. Schools with sanitary service in use (2004, %): Percentage of schools that have some type of sanitary service in use. Toilets in use (%): Percentage of schools that have some toilets in use. Latrines in use (%): Percentage of schools that have some latrines in use. Septic tanks in use (%): Percentage of schools that have some type of septic tank or other type of sanitary service other than the abovementioned and is in use. 61 62 What can a Regional Government do to improve education? The Case of Junin Schools with toilets (2006, %): Percentage of schools that have some type of toilets (most recent information available). Maintenance Classrooms that do not require major repairs (%): Average percentage of classrooms that do not require major repairs per school. B. Status of school infrastructure Furniture Students with desks (%): Average percentage of students with a desk per school. Reading Materials School library in preschool (% IIEE): Percentage of preschool education institutions that have a school library. School Library in Primary (% IIEE): Percentage of primary schools that have a school library. School Library in Secondary (% IIEE): Percentage of secondary schools that have a school library. Information technology and communication Student ratio per computer: Average number of students per computer per school (excludes centers for preschool education). Connection to Internet: Percentage of schools with a connection to Internet (excludes preschool education centers). Status of Human Resources Quantity Students per teacher (students): Average number of students per teacher across schools. o Preschool: Average number of students per teacher in preschool institutions. o Primary: Average number of students per teacher in primary schools. o Secondary: Average number of students per teacher in secondary schools. Average class size (students): Average number of students per classroom teacher in schools. o Preschool: Average number of students per classroom teacher in preschool institutions. o Primary: Average number of students per classroom teacher in primary schools. o Secondary: Average number of students per classroom teacher in secondary schools. o Single teacher: Average number of students per classroom teacher in single teacher schools. o Multi-grade: Average number of students per classroom teacher in multi-grade schools. o Full grade: Average number of students per classroom teacher in full grade schools. Distribution Teaching personnel in the classroom (% of teachers): Average percentage of teachers that work in the classroom per school. Overpopulated classes in rural areas (%): Percentage of schools located in rural areas whose average class size class exceeds 40% of the maximum class size established by norm. Overpopulated classes in multi-grade schools (%): Percentage of multi-grade schools whose average class size class exceeds 40% of the maximum class size established by norm. Overpopulated classes in single teacher schools (%): Percentage of single teachers schools whose average class size class exceeds 40% of the maximum class size established by norm. Overpopulated classes in urban schools (%): Percentage of schools located in the urban areas whose class size is 50% lower than the maximum size stipulated by norm. Primary school teachers in multi-grade classrooms (%): Percentage of teachers that teach in single teacher or multi-grade primary schools. Primary students in multi-grade classrooms (%): Percentage of students enrolled in single teacher or multi- grade primary schools. 62 63 What can a Regional Government do to improve education? The Case of Junin Quality Teachers with titles (%): Percentage of teacher that hold a teaching title (information is only available for 11,200 schools in the country). Teachers evaluated in the last census (%): Percentage of teachers that participated in the 2007 census evaluation. C. Financial situation of education Quantity Per student expenditure relative to per capita regional GDP (%): Expenditure executed per student as a proportion of regional per capital GDP. Expenditure on education and culture relative to regional budget (%): Expenditure executed for Education and Culture as a proportion of the budget executed by regional governments. Expenditure by student (new soles): Annual expenditure executed for Basic Regular Education (preschool, primary and secondary) per student. o Preschool: Annual expenditure executed per student in preschool. o Primary: Annual expenditure executed per student in primary. o Secondary: Annual expenditure executed per student in secondary. Distribution Expenditure executed by items o Per capita non-salary expenditures (new soles): Annual per student expenditure executed for non- salary items (goods and services, investments, other capital expenditures, others). o Per capita capital expenditure (new soles): Annual per student expenditure executed for investment items. Expenditure on Regular Basic Education (%). o Preschool: Proportion of expenditure on Regular Basic Education (preschool, primary and secondary) that is destined for preschool. o Primary: Proportion of expenditure on Regular Basic Education (preschool, primary and secondary) that is destined for primary. o Secondary: Proportion of Expenditure on Regular Basic Education (preschool, primary and secondary) that is destined for secondary. D. Education results and learning achievements The population's education level Young population with complete secondary studies (%): Percentage of young adults aged 18 to 25 that have completed at least secondary education. System coverage (% pop.) Net enrollment rate in preschool (3-5 years): Percentage of children from 3 to 5 enrolled in any regular school. Net enrollment rate in primary (6-11 years): Percentage of children from 6 to 11 enrolled in any regular school. Net enrollment rate in secondary (12-17 years): Percentage of children from 12 a 17 enrolled any regular school. Student Progress (% students) Completion rate for primary (12-14 years): Proportion of the population aged 12 to 14 that completed primary school. Transition rate to secondary: Proportion of students that graduated from 6th grade of primary and attended 1st of secondary the following year. Completion rate for secondary (17-19 years): Proportion of the population aged 12 to 14 that completed secondary school. 63 64 What can a Regional Government do to improve education? The Case of Junin Learning achievements: Sufficient performance (% students) 2º Primary Communication: Percentage of students in the second grade of Primary that reached a sufficient level in the area of Reading Comprehension on the National Assessment of Student Performance 2004. 2º Primary Mathematics s: Percentage of students in the second grade of Primary that achieved a sufficient level in the area of Logic-Mathematics on the National Assessment of Student Performance 2004 6º Primary: Communication: Percentage of students of the sixth grade of Primary that achieved a sufficient level in the area of Reading Comprehension on the National Assessment of Student Performance 2004. 6º Primary Mathematics: Percentage of students in the sixth grade of Primary that achieved a sufficient level in the area of Logic-Mathematics on the National Assessment of Student Performance 2004 5º Secondary Communication: Percentage of students of the fifth year of Secondary that achieved a sufficient level on the Reading Comprehension Section of the National Assessment of Student performance 2004. 5º Secondary Mathematics: Percentage of students in the fifth year of Secondary that achieved a sufficient level on the Mathematics section of the National Assessment for Student Performance 2004 Annex 1.5. Summary of Statistical Indicators Category / Indicator Source Junin Ranking in Peru School Infrastructure Schools with all services (%) MED 37 11 Schools with no services (%) MED 16 7 Schools with sanitary services (2004, %) MED 90 12 Schools with sanitary services in use (2004,%) MED 89 12 Schools with toilets (2006, %) MED 45 13 Classrooms that so not require major repairs (%) MED 78 10 School Equipment Students with a desk MED 93 9 Student library in Preschool (% IIEE) MED 26 14 Student library in Primary (% IIEE) MED 35 20 Student library in Secondary (% IIEE) MED 53 21 Student ratio per computer MED 24 8 Connection to internet (%) MED 10 13 Human resources Students per teacher (students) MED 19 14 Average class (students) MED 21 14 Overpopulated classes in rural schools (%) MED 14 18 Overpopulated classes in urban schools (%) MED 49 23 Teachers evaluated in the last census (%) MED 48 19 Financing for education Per student expenditure relative to the regional per capita GDP (%) MEF 22 13 Per student expenditure (new soles) MEF 885 12 Non-salary per capita Expenditure (new soles) MEF 19 22 Expenditure on per capita investment (new soles) MEF 3 16 Education results and learning achievements Young adults with complete secondary (%) INEI 78 7 Enrollment rate in preschool (3-5 years) INEI 42 24 Net enrollment rate in primary (6-11 years) INEI 95 8 Net enrollment rate in secondary (12-16 years) INEI 76 12 Completion rate in primary (12-14 years) MED 80 8 Completion rate in secondary (17-19 years) MED 62 6 Sufficient performance: 2º Primary Communication (% students) MED 12 9 Sufficient performance: 2º Primary Mathematics (% students) MED 12 6 64 65 What can a Regional Government do to improve education? The Case of Junin 2. Annexes in chapter 2 Annex 2.1. Status of school infrastructure and ordering Status of school infrastructure Basic Services Region Schools with Schools with Schools with all services Ranking Schools with Ranking potable water Ranking sewage Ranking Schools with Ranking (%) electricity (%) (%) systems (%) no service (%) Amazon 20 18 35 21 55 15 26 14 39 18 Ancash 38 10 62 12 75 8 42 10 18 8 Apurimac 20 17 55 16 60 11 22 20 26 12 Arequipa 69 4 83 4 82 5 74 4 10 4 Ayacucho 21 16 47 18 55 16 24 16 34 16 Cajamarca 14 22 26 24 50 17 18 21 44 20 Callao 86 1 98 1 87 2 86 1 2 1 Cuzco 35 12 63 10 56 14 40 11 27 13 Huancavelica 13 23 63 11 48 18 15 23 26 11 Huanuco 18 20 30 22 45 21 23 18 48 22 Ica 62 5 80 5 84 3 67 5 10 3 Junin 37 11 80 6 57 12 39 12 16 7 La Libertad 46 9 59 14 75 7 50 9 21 10 Lambayeque 54 6 65 9 66 10 57 6 28 14 Lima 84 2 96 2 90 1 86 2 3 2 Loreto 9 24 28 23 12 25 11 24 72 24 Madre de Dios 22 15 37 20 35 23 24 17 57 23 Moquegua 48 8 67 8 78 6 54 8 16 6 Pasco 20 19 59 15 35 22 22 19 40 19 Piura 33 13 52 17 56 13 35 13 34 17 Puno 24 14 59 13 48 19 25 15 32 15 San Martin 15 21 39 19 46 20 16 22 46 21 Tacna 71 3 85 3 82 4 75 3 10 5 Tumbes 52 7 71 7 73 9 56 7 19 9 Ucayali 7 25 23 25 13 24 11 25 75 25 PERU 42 63 63 45 26 Source: MED, School Census 2006. 65 66 What can a Regional Government do to improve education? The Case of Junin Status of school infrastructure Sanitary services Region Schools with Schools with sanitary services Ranking Schools with Ranking Schools with Ranking septic tanks or Ranking (2004, %) toilets (%) latrines (%) other HS (%) Amazon 89 13 23 19 36 12 30 1 Ancash 83 17 40 11 32 14 12 8 Apurimac 68 24 25 18 39 11 4 22 Arequipa 96 5 80 5 11 21 6 21 Ayacucho 74 22 22 20 41 8 10 12 Cajamarca 92 10 19 23 62 1 12 9 Callao 99 2 94 2 3 24 3 23 Cuzco 78 19 36 13 28 17 14 5 Huancavelica 56 25 15 25 31 15 11 11 Huanuco 89 15 20 21 57 2 12 7 Ica 95 8 81 4 7 23 7 20 Junin 90 12 39 12 39 10 12 6 La Libertad 91 11 49 9 27 18 15 4 Lambayeque 89 14 62 7 16 19 11 10 Lima 98 4 94 1 2 25 2 25 Loreto 74 21 25 17 41 9 9 15 Madre de Dios 98 3 26 15 51 5 22 2 Moquegua 100 1 59 8 33 13 8 17 Pasco 77 20 25 16 42 7 10 13 Piura 81 18 41 10 30 16 10 14 Puno 84 16 20 22 57 3 8 18 San Martin 94 9 26 14 51 4 17 3 Tacna 96 6 86 3 8 22 2 24 Tumbes 95 7 74 6 14 20 8 16 Ucayali 69 23 17 24 45 6 7 19 PERU 87 49 29 9 Source: MED, Basic statistics 2004. 66 67 What can a Regional Government do to improve education? The Case of Junin Status of school infrastructure Sanitary services Maintenance Schools with Classrooms that Region sanitary Toilets in Latrines in Septic Schools do not require services in use Ranking use (%) Ranking use (%) Ranking tanks in Ranking with toilets Ranking major repairs Ranking (2004. %) use (%) (2006, %) (%) Amazon 89 13 97 14 98 7 99 5 25 21 72 17 Ancash 83 17 99 7 96 17 96 16 50 11 68 22 Apurímac 67 24 97 12 98 4 94 20 28 19 68 21 Arequipa 96 5 99 3 95 20 96 15 86 5 89 3 Ayacucho 73 22 97 11 98 8 99 3 28 18 66 23 Cajamarca 92 10 92 24 99 2 98 9 24 24 75 15 Callao 99 1 99 1 94 22 98 8 98 1 94 1 Cuzco 78 19 99 6 97 11 97 12 50 10 72 19 Huancavelica 56 25 96 21 97 15 93 23 23 25 61 25 Huanuco 88 15 90 25 97 16 98 7 27 20 72 18 Ica 95 7 97 13 94 21 95 19 88 4 86 4 Junin 89 12 99 4 98 5 98 11 45 13 78 10 La Libertad 90 11 98 10 95 19 96 14 56 9 80 8 Lambayeque 89 14 99 2 98 6 94 22 66 8 81 6 Lima 98 4 99 5 97 10 96 17 97 2 92 2 Loreto 74 21 96 20 97 14 95 18 30 17 79 9 Madre de Dios 98 3 93 22 98 3 99 4 33 15 81 7 Moquegua 99 2 93 23 86 25 94 21 73 7 75 16 Pasco 77 20 96 19 97 13 100 1 32 16 82 5 Piura 80 18 98 9 97 12 93 24 49 12 75 14 Puno 84 16 98 8 98 9 98 6 25 22 65 24 San Martin 94 8 97 17 99 1 98 10 39 14 76 13 Tacna 96 6 96 18 93 23 100 2 89 3 77 12 Tumbes 93 9 97 15 89 24 86 25 84 6 78 11 Ucayali 68 23 97 16 96 18 97 13 25 23 71 20 PERÚ 87 97 96 96 55 79 Sources: MED, Basic statistics 2004 and School Census 2006. 67 68 What can a Regional Government do to improve education? The Case of Junin Annex 2.2. Status of school equipment and ordering Status of school equipment Furniture Reading material Information technology and communication Students School library School library School library Ratio of Region without Ranking in Preschool Ranking in Primary (% Ranking in Secondary Ranking students per Ranking Connection Ranking desks (%) (%) schools) (%schools) computer to internet Amazon 89 19 11 22 36 19 70 9 172 24 5 22 Ancash 85 23 33 10 56 8 72 4 46 11 11 12 Apurimac 85 25 35 8 52 9 62 15 130 19 8 15 Arequipa 97 3 30 12 46 14 65 13 13 6 27 5 Ayacucho 87 22 5 24 26 24 43 23 59 15 8 14 Cajamarca 89 17 37 6 61 4 66 11 230 25 3 24 Callao 99 1 38 4 57 7 72 5 9 1 48 1 Cuzco 87 20 46 3 57 6 70 7 47 12 11 11 Huancavelica 87 21 6 23 40 18 58 18 162 21 3 25 Huanuco 85 24 38 5 62 3 70 8 95 18 6 20 Ica 96 4 17 20 33 22 70 6 12 4 21 8 Junin 93 9 26 14 35 20 53 21 24 8 10 13 La Libertad 91 14 32 11 47 13 61 17 38 10 15 9 Lambayeque 92 10 33 9 58 5 65 12 23 7 21 7 Lima 98 2 35 7 52 10 64 14 10 2 41 2 Loreto 90 16 26 16 35 21 54 20 134 20 4 23 Madre de Dios 96 5 0 25 5 25 16 25 88 17 7 17 Moquegua 94 7 21 19 44 15 76 2 12 5 22 6 Pasco 91 15 78 1 75 1 77 1 57 14 7 16 Piura 89 18 22 18 40 17 57 19 67 16 11 10 Puno 92 11 29 13 41 16 53 22 47 13 7 19 San Martin 92 13 50 2 68 2 75 3 168 22 7 18 Tacna 93 8 26 15 51 11 62 16 11 3 31 4 Tumbes 92 12 22 17 50 12 69 10 27 9 31 3 Ucayali 95 6 13 21 30 23 31 24 170 23 5 21 PERÚ 92 31 48 62 27 16 Source: MED, School Census 2006. 68 69 What can a Regional Government do to improve education? The Case of Junin Annex 2.3. Status of human resources and ordering Status of human resources Quantity Region Students per Students per Students per Students per teacher (students) Ranking teacher (Preschool) Ranking teacher (primary) Ranking teacher Ranking (Secondary) Amazon 24 24 21 15 25 15 23 25 Ancash 18 9 20 12 18 12 12 9 Apurimac 21 16 24 22 21 22 17 22 Arequipa 14 6 15 5 15 5 12 10 Ayacucho 19 13 21 14 19 14 13 14 Cajamarca 21 17 20 13 23 13 17 20 Callao 13 4 14 2 14 2 11 6 Cuzco 22 19 22 18 24 18 16 18 Huancavelica 23 20 24 23 24 23 18 24 Huanuco 25 25 24 24 27 24 17 19 Ica 15 7 17 8 15 8 11 5 Junin 19 14 18 10 21 10 12 8 La Libertad 19 15 17 9 22 9 13 13 Lambayeque 18 10 17 6 20 6 14 16 Lima 13 3 14 1 14 1 11 4 Loreto 24 23 23 20 26 20 15 17 Madre de Dios 17 8 21 16 17 16 12 11 Moquegua 11 1 14 3 10 3 8 1 Pasco 18 12 20 11 20 11 12 7 Piura 22 18 22 17 23 17 17 23 Puno 18 11 23 21 18 21 13 15 San Martin 23 21 22 19 25 19 17 21 Tacna 12 2 15 4 12 4 9 2 Tumbes 14 5 17 7 13 7 11 3 Ucayali 24 22 26 25 26 25 12 12 PERU 19 18 21 14 Source: MED, Basic statistics 2006. 69 70 What can a Regional Government do to improve education? The Case of Junin Status of human resources Quantity (Class Size) Region Average Preschool Primary Secondary Single teacher Multi-grade Full grade class size Ranking (students Ranking (students Ranking (students per Ranking (students per Ranking (students per Ranking (students per Ranking (students) per class) per class) class) class) class) class) Amazon 24 20 22 15 25 21 24 20 24 21 25 21 23 20 Ancash 20 9 20 12 19 9 22 14 18 9 20 10 21 12 Apurimac 23 17 24 23 21 13 26 23 21 12 22 17 26 24 Arequipa 17 6 16 5 16 4 22 13 15 6 16 5 20 8 Ayacucho 20 11 21 14 20 11 20 6 19 10 21 12 21 13 Cajamarca 22 16 20 13 23 17 22 12 24 19 22 15 22 16 Callao 17 5 16 3 17 7 21 7 11 4 16 4 19 4 Cuzco 25 21 23 19 25 20 26 24 24 20 24 20 26 25 Huancavelica 24 18 24 22 24 18 25 21 22 17 25 23 24 21 Huanuco 26 25 24 24 28 25 23 19 27 25 27 25 23 19 Ica 18 7 18 8 17 6 23 15 16 7 18 7 20 7 Junin 21 14 19 10 22 15 21 11 22 16 21 13 20 6 LaLibertad 22 15 19 9 23 16 23 16 22 15 22 16 21 11 Lambayeque 21 13 18 7 22 14 23 17 22 14 20 11 21 10 Lima 17 3 15 2 17 5 21 8 13 5 15 3 19 3 Loreto 25 23 23 20 27 23 21 10 26 23 25 22 23 18 Madre de Dios 19 8 22 16 18 8 20 5 17 8 21 14 22 15 Moquegua 13 1 15 1 11 1 15 1 9 1 12 1 17 1 Pasco 20 10 20 11 20 12 19 4 20 11 19 9 20 5 Piura 24 19 22 17 24 19 26 25 23 18 24 18 25 22 Puno 20 12 24 21 19 10 23 18 21 13 19 8 23 17 San Martin 25 22 23 18 26 22 25 22 26 22 24 19 25 23 Tacna 16 2 16 4 14 2 18 3 11 3 14 2 18 2 Tumbes 17 4 18 6 15 3 21 9 11 2 16 6 20 9 Ucayali 25 24 26 25 27 24 18 2 27 24 26 24 22 14 PERU 21 19 22 22 22 20 21 Source: MED, Basic Statistics 2006. 70 71 What can a Regional Government do to improve education? The Case of Junin Status of human resources Distribution Teaching Overcrowded Region personnel in Overcrowded Overpopulated classes in single classes Ranking classes in rural Ranking classes in multi- Ranking teacher schools Ranking (% teachers) schools (%) grade schools (%) (%) Amazon 96 3 17 21 5 21 30 21 Ancash 93 12 6 8 1 8 12 9 Apurimac 96 2 10 11 2 13 21 13 Arequipa 89 17 3 4 1 3 4 5 Ayacucho 94 7 8 10 4 19 14 10 Cajamarca 96 4 11 14 1 6 29 20 Callao 85 25 - 3 14 2 3 Cuzco 93 10 12 16 3 15 27 19 Huancavelica 97 1 13 17 3 17 25 16 Huanuco 95 6 21 23 5 22 38 24 Ica 87 21 6 9 5 23 5 7 Junin 91 16 14 18 2 12 23 15 La Libertad 89 18 10 13 4 18 21 14 Lambayeque 88 19 12 15 3 16 25 17 Lima 85 24 5 7 1 7 5 6 Loreto 92 14 23 24 5 24 36 23 Madre de Dios 92 13 5 6 0 1 8 8 Moquegua 87 22 1 3 1 5 1 2 Pasco 96 5 10 12 1 4 18 12 Piura 92 15 14 19 4 20 26 18 Puno 94 8 4 5 2 9 17 11 San Martin 93 11 16 20 2 11 32 22 Tacna 88 20 1 1 1 2 3 4 Tumbes 86 23 1 2 2 10 0 1 Ucayali 93 9 21 22 6 25 39 25 PERU 90 12 3 23 Source: MED, Basic statistics 2006. 71 72 What can a Regional Government do to improve education? The Case of Junin Status of human resources Distribution Quality Under- Primary Primary school Teachers Region populated school classes in Ranking teachers in Ranking students in Ranking Teachers with Ranking evaluated in Ranking urban schools multi-grade multi-grade titles (%) the last (%) classes (%) classrooms Census (%) (%) Amazon 17 6 73 23 74 24 89 17 35 23 Ancash 32 17 54 17 48 14 91 15 40 22 Apurímac 15 4 66 21 61 21 99 1 60 13 Arequipa 45 21 35 7 27 7 95 10 88 4 Ayacucho 32 16 61 20 55 20 97 4 19 25 Cajamarca 22 8 75 24 73 23 96 6 27 24 Callao 51 24 17 1 12 3 68 25 90 2 Cuzco 26 12 54 16 51 17 90 16 57 14 Huancavelica 16 5 77 25 76 25 81 22 47 20 Huanuco 24 11 66 22 67 22 89 18 61 12 Ica 44 20 26 5 18 6 95 9 65 9 Junin 49 23 50 13 47 13 93 12 48 19 La Libertad 36 18 41 9 40 11 95 8 53 17 Lambayeque 45 22 33 6 29 8 92 13 46 21 Lima 51 25 21 4 15 4 79 24 83 5 Loreto 9 2 52 15 51 18 80 23 68 7 Madre de Dios 18 7 48 12 38 9 94 11 88 3 Moquegua 30 15 36 8 18 5 99 2 62 11 Pasco 22 9 56 18 53 19 87 19 48 18 Piura 28 14 47 11 44 12 85 20 66 8 Puno 23 10 58 19 51 16 98 3 62 10 San Martin 8 1 51 14 49 15 92 14 71 6 Tacna 37 19 20 2 11 1 96 5 54 16 Tumbes 27 13 21 3 12 2 96 7 55 15 Ucayali 13 3 43 10 39 10 81 21 93 1 PERU 41 44 40 84 50 Sources: MED, Basic statistics 2006, School census 2006; UMC - Census evaluation on teachers 2007. 72 73 What can a Regional Government do to improve education? The Case of Junin Annex 2.4. Status of education financing and ordering Financial status of education Quantity (executed 2006) Per student Expenditure on Per student Per student Per student Region expenditure Per student relative to regional Ranking education and Ranking expenditure Ranking expenditure Ranking expenditure Ranking expenditure Ranking per capita GDP culture relative to (new soles) in Preschool in Primary in Secondary (%) regional budgets (%) (new soles) (new soles) (new soles) Amazon 45 1 61 17 815 20 556 21 797 14 1,008 20 Ancash 21 17 74 3 995 8 675 13 913 9 1,288 9 Apurímac 44 2 61 16 880 13 631 17 868 11 1,031 18 Arequipa 18 21 58 22 1,166 5 839 4 1,033 3 1,469 5 Ayacucho 42 3 64 12 962 10 624 19 929 7 1,209 11 Cajamarca 21 15 69 5 821 19 631 18 790 15 998 21 Callao - 68 8 857 16 802 6 663 24 1,150 13 Cuzco 22 14 69 7 724 24 464 25 706 20 895 25 Huancavelica 27 8 66 10 839 17 648 16 780 16 1,073 15 Huanuco 28 7 74 2 708 25 610 20 630 25 941 24 Ica 20 18 63 13 987 9 691 12 867 12 1,301 8 Junin 22 13 59 19 885 12 659 14 775 17 1,147 14 La Libertad 19 19 59 20 835 18 729 9 730 19 1,062 17 Lambayeque 21 16 60 18 767 22 509 23 670 23 1,029 19 Lima 12 23 721* 1 1,194 4 893 3 1,029 4 1,531 4 Loreto 23 12 65 11 872 15 726 10 739 18 1,278 10 Madre de Dios 30 6 59 21 1,029 6 805 5 873 10 1,387 6 Moquegua 11 24 43 25 1,494 1 1,243 1 1,308 1 1,883 2 Pasco 18 20 69 6 1,028 7 653 15 919 8 1,380 7 Piura 23 11 57 23 758 23 512 22 700 21 973 23 Puno 35 5 71 4 954 11 491 24 954 5 1,177 12 San Martin 27 9 62 15 874 14 712 11 809 13 1,073 16 Tacna 14 22 53 24 1,200 3 736 7 953 6 1,806 3 Tumbes 40 4 68 9 1,454 2 1,231 2 1,110 2 2,119 1 Ucayali 24 10 63 14 777 21 732 8 677 22 982 22 PERU 17 86 943 705 836 1,236 Source: MEF - SIAF * Includes budget executed by MED. 73 74 What can a Regional Government do to improve education? The Case of Junin Financial status of education Distribution Per capita Per capita Initial cost (% Expenditure Expenditure Region non-salary on Primary on Secondary current Ranking Capital Ranking total Basic Ranking (% total Basic Ranking (% total Basic Ranking expenditure expenditure Regular Regular Regular (new soles) (new soles) Education) Education) Education) Amazon 20.1 21 0.3 22 11% 14 57% 2 32% 24 Ancash 36.2 12 10.1 8 10% 16 50% 12 40% 13 Apurímac 40.8 10 19.4 5 11% 11 53% 8 36% 19 Arequipa 42.3 9 5.4 14 11% 12 42% 19 47% 4 Ayacucho 60.7 5 30.4 2 9% 22 57% 3 34% 23 Cajamarca 18.1 23 0.1 25 11% 13 58% 1 31% 25 Callao 91.0 2 8.5 10 16% 3 37% 23 47% 3 Cuzco 28.0 17 8.6 9 10% 18 55% 5 35% 20 Huancavelica 38.3 11 5.7 13 12% 10 54% 6 35% 21 Huanuco 23.0 18 6.8 12 9% 21 56% 4 34% 22 Ica 44.3 7 21.4 4 13% 7 41% 20 46% 5 Junin 18.8 22 3.4 16 8% 23 49% 14 43% 11 La Libertad 21.7 20 0.1 23 10% 20 50% 11 40% 14 Lambayeque 34.2 14 7.9 11 8% 25 48% 16 44% 7 Lima* 303.9 1 128.2 1 12% 9 40% 21 48% 2 Loreto 16.6 24 1.4 19 14% 5 50% 13 37% 17 Madre de Dios 68.6 3 1.3 20 13% 8 44% 18 44% 9 Moquegua 49.4 6 14.8 6 17% 2 39% 22 44% 10 Pasco 29.8 15 2.8 17 10% 17 46% 17 44% 8 Piura 28.4 16 13.1 7 10% 19 51% 9 40% 15 Puno 34.8 13 0.1 24 8% 24 50% 10 41% 12 San Martin 22.9 19 4.1 15 10% 15 54% 7 36% 18 Tacna 63.9 4 25.2 3 13% 6 35% 24 52% 1 Tumbes 43.3 8 0.7 21 21% 1 33% 25 45% 6 Ucayali 14.7 25 2.2 18 14% 4 48% 15 38% 16 PERÚ 88.4 32.3 11% 47% 42% Source: MEF-SIAF. *Expenditure in n Lima also includes investment that the MED executed and which benefited other departments. 74 75 What can a Regional Government do to improve education? The Case of Junin Annex 2.5. Expenditure executed on education and the number of students in the system (2000-2006) Annex Figure 2.5-a. Evolution of total expenditure on education and culture and enrollment in basic regular education, 2000-2006 350 350 9,000 6,600 Junin Peru 6,550 300 8,000 340 Students 6,500Students NS) NS) 7,000 250 6,450 330 6,000 6,400 200 5,000 (thousands) (Millions (thousands) 6,350 320 (Millions 150 4,000 6,300 6,250 310 3,000 100 6,200 2,000 6,150 Expenditure50 300 Expenditure1,000 6,100 0 290 0 6,050 2000 2002 2004 2006 2000 2002 2004 2006 Expenditure Students Expenditure Students Fuentes: MEF-SIAF; MED. Annex Table 2.5-a. Expenditure executed on education and culture (million of soles at prices in the year 2000) 2000 2002 2004 2006 Junin Region Preschool 14 16 17 21 Primary 88 96 103 118 Preschool 72 82 91 104 Others 44 47 60 71 Total 218 242 272 313 Peru Preschool 448 500 532 584 Primary 1,955 1,983 2,315 2,484 Preschool 1,400 1,621 1,981 2,183 Preschool 1,555 1,760 2,081 2,405 Total 5,358 5,864 6,909 7,657 Source: SIAF. Annex Table 2.5-b. Number of students in public institutions 2000 2002 2004 2006 Junin Region Preschool education 38,506 37,332 36,806 35,057 Primary education 196,110 185,583 179,785 170,932 Secondary education 105,156 107,166 105,027 101,603 Total number of students in basic regular education* 339,772 330,081 321,618 307,592 Peru Preschool education 931,761 941,392 922,379 930,540 Primary education 3,731,182 3,673,626 3,534,197 3,337,868 Secondary education 1,816,100 1,952,507 1,991,286 1,985,113 Total number of students in basic regular education * 6,479,043 6,567,525 6,447,862 6,253,521 * Includes Preschool education, Primary education (minor) and Secondary education (minor). Source: SCALE (MED). 75 76 What can a Regional Government do to improve education? The Case of Junin Annex Table 2.5-c. Expenditure executed on education and culture per item (current soles) 2000 2002 2004 2006 Junin region Salaries and compensations 194,268,152 217,689,896 241,647,403 277,795,054 Goods and services 14,365,678 14,295,557 16,152,329 17,353,360 Other current expenses 3,619,891 3,186,577 3,785,937 4,443,969 Investments 4,198,614 4,878,070 7,004,878 10,901,025 Financial investments 0 13,405 46,872 57,290 Other capital expenditures 1,416,700 1,520,221 3,179,091 2,941,643 Total 217,869,035 241,583,726 271,816,509 313,492,342 Peru Salaries and compensations 3,875,095,524 4,519,135,204 5,106,253,777 5,792,140,018 Goods and services 769,223,789 812,729,883 915,715,031 941,652,117 Other current expenses 110,369,944 133,836,946 394,273,982 181,757,022 Investments 527,186,228 326,166,986 359,639,823 562,667,886 Financial investments 13,015,401 11,983,407 19,366,505 13,970,234 Other capital expenditures 61,573,458 57,466,636 112,079,217 164,184,938 Debt amortization 277,804 1,881,739 912,063 141,234 Interests and charges for debt 768,942 871,143 327,957 3,304 Total 5,357,511,091 5,864,071,943 6,908,568,355 7,656,516,754 Source: SIAF. Annex 2.6. Per student expenditure by level of education (% of per capita GDP) Annex Table 2.6-a. Per student expenditure and level as a percentage of per capita GDP* 2000 2002 2004 2006 Junin Regional Preschool education 5 6 6 6 Primary education 6 7 7 7 Secondary education 9 10 11 10 Peru Preschool education 7 7 7 6 Primary education 7 7 8 8 Secondary education 10 11 12 11 *Regional in the case of Junin and national PBI in the case of Peru. Sources: SIAF, BCRP, MEF, INEI. 76 77 What can a Regional Government do to improve education? The Case of Junin Annex 2.7. Expenditure executed on Education and Culture by program and budget allocation (current soles 2006) Expenditure executed on Education and Culture per program and budget allotment (current soles 2006) Programs Budget Assignment Total Junin MED* Regional Gov. INFES** Administration 4,643,202 6,105,388 10,748,590 Government Planning Assistance for educandonos 1,467,341 719,264 2,186,605 Training on fine tuning Culture Telecommunications, science and technology Special Education 2,297,478 2,297,478 Preschool Education 106,691 22,994,667 7,425 23,108,783 Primary Education 654,292 131,741,234 13,563 132,409,089 Secondary Education 116,193,826 387,363 116,581,189 Higher Education 37,974,069 20,346,129 58,320,198 Education Infrastructure 6,525,100 6,525,100 Physical Education and sports Disaster preparation Total 44,845,596 306,923,086 408,351 352,177,033 Percentage 13% 87% 0% 100% * MED. ** National Institute of Education and Health Infrastructure. Annex 2.8. Distribution of Education Expenditure (2000-2006) Annex Figure 2.8-a. Expenditure distribution, 2000 y 2006 Capital expenditure Non-salary current expenditure Salary expenditure 100% 80% 60% 89% 89% 40% 72% 76% 20% 0% Junin Peru Junin Peru 2000 2006 Source: SIAF. 77 78 What can a Regional Government do to improve education? The Case of Junin Annex Table 2.8-a. Key reasons 2000 2002 2004 2006 Junin Expenditure on salaries and compensation as a percentage of total expenditure 89.17 90.11 88.90 88.61 Expenditure on goods and services as a percentage of total expenditure 6.59 5.92 5.94 5.54 Current expenditure as a percentage of total expenditure 97.42 97.35 96.24 95.57 Total expenditure on education and culture 0.12 0.12 0.12 0.12 Peru Expenditure on salaries and compensation as a percentage of total expenditure 72.33 77.06 73.91 75.65 Expenditure on goods and services as a percentage of total expenditure 14.36 13.86 13.25 12.30 Current expenditure as a percentage of total expenditure 88.75 93.21 92.87 90.32 Total expenditure on education and culture as a percentage of national GDP 2.9 3.0 3.1 2.8 Source: SIAF, BCRP. Annex Figure 2.8-b. Per student expenditure on basic regular education according to item ( current soles 2006) 1600 1400 Non-Salary per student expenditure Per Student Salary Expenditure 1200 1000 Junín = 885 soles Promedio nacional = 943 soles 800 600 400 200 0 AUGEU BES ANC I PAIU SO O A O NIT O EU DI ASH IC CHO NIN ACI ADT URA SCO UCO TA PUN JU CAMI TUM DE PASC ARM OTER ANC LO LLAAC PI AREQ APUR CAVEL BERIL CARAMAJ ASNO AZ AYALCU CU BAYEQ HUAN MOQ REDAM AYACU SAN ANUH AL AM CA AML * The items for expenses in Lima contain expenditure executed for national level investment projects. Sources: SIAF, MED. 78 79 What can a Regional Government do to improve education? The Case of Junin Annex Figure 2.8-c. Non-salary per student expenditure in EBR according to item (current soles 2006) 100 90 Other capital expenditure Investments 80 Other current expenditure 70 Goods and services 60 50 40 Promedio nacional = 32 soles 30 Junín = 19 soles 20 10 0 S A O O A CA DI M ACI EU O N I CO NI LLAAC DE CNAAT GUE IC BES PAIU ASHC UNOP NUCO RTA ADT ASN UNIJ RCAA PASC URAIP CUS O OTER E DRA CUCHOAYA MUT OQU EQRA URIPA AVELC AN AYEQB UAH M BERIL AZ MAJ LO AYALCU M M ANUH AML NAS AL AM CA Sources: SIAF, MED. Annex Table 2.8-d. Per student expenditure according to region (2006) Salaries Goods and Other current Other per capita Services expenditures Investments Expenses Total Amazon 795 18 2 0 0 815 Ancash 959 24 3 10 0 995 Apurimac 839 19 3 19 0 880 Arequipa 1,123 35 2 5 0 1,166 Ayacucho 902 17 13 30 0 962 Cajamarca 803 15 3 0 0 821 Callao 766 78 1 9 4 857 Cuzco 696 17 2 9 0 724 Huancavelica 801 22 10 6 0 839 Huanuco 685 14 2 7 0 708 Ica 943 22 1 21 0 987 Junin 866 14 1 3 0 885 La Libertad 813 18 3 0 0 835 Lambayeque 733 24 2 8 0 767 Lima* 890 159 10 128 7 1,194 Loreto 855 14 1 1 0 872 Madre de Dios 961 66 1 1 0 1,029 Moquegua 1,445 33 1 15 0 1,494 Pasco 998 26 1 3 0 1,028 Piura 729 14 2 13 0 758 Puno 919 32 2 0 0 954 San Martin 851 17 1 4 0 874 Tacna 1,136 38 1 25 0 1,200 Tumbes 1,410 41 2 1 0 1,454 Ucayali 762 12 1 2 0 777 * The investment item in Lima contains the expenditure executed for national investment projects Sources: SIAF, MED. 79 80 What can a Regional Government do to improve education? The Case of Junin 3. Annexes in chapter 3 Annex 3.1. Education results, learning achievements and ordering Education results Education level of the Population Coverage of the education system Student progress Net Net Completion Region Young adults enrollment Completion with complete enrollment Net enrollment rate in rate in Transition rate in secondary Ranking rate in pre- Ranking rate in primary (% Ranking secondary Ranking primary (% Ranking rate to Ranking secondary Ranking education (%) school (% 3- 6-11 years) secondary (% 17-19 5 years) (% 12-16 12-14 years) years) years) Amazon 40 25 58 11 92 20 63 19 65 16 79.2 25 32.6 21 Ancash 63 13 52 16 93 17 66 18 64 17 88.1 17 38.8 16 Apurimac 65 12 64 8 96 4 77 9 61 21 97.1 7 37.5 17 Arequipa 81 5 58 12 95 7 84 3 84 4 98.1 4 66.6 3 Ayacucho 47 21 46 21 95 10 63 20 56 22 86.2 18 29.6 23 Cajamarca 41 24 50 18 95 6 59 23 62 19 79.4 24 31.3 22 Callao 79 6 67 7 90 23 80 7 96 1 100 1 61.5 7 Cuzco 57 17 57 13 94 12 72 14 61 20 88.5 16 45.7 13 Huancavelica 43 23 48 19 94 15 67 16 54 23 92.9 11 25.9 25 Huanuco 50 20 39 25 93 16 54 25 51 25 80.4 23 26.7 24 Ica 82 3 72 4 97 1 85 2 74 13 99.2 3 62.9 5 Junin 78 7 42 24 95 8 76 12 80 8 95.6 10 61.8 6 La Libertad 59 16 61 9 92 19 63 21 74 14 89 15 41.8 15 Lambayeque 66 11 68 6 95 11 74 13 76 10 92.3 12 50.1 11 Lima 81 4 76 3 93 18 84 5 84 5 97.4 6 69.1 1 Loreto 54 19 46 20 85 25 60 22 52 24 81.1 21 35.4 19 Madre de Dios 71 9 56 14 94 13 80 8 79 9 89.6 14 54.2 9 Moquegua 85 1 69 5 97 2 83 6 86 3 100 2 65.9 4 Pasco 61 14 56 15 94 14 76 10 80 7 84.5 19 50.7 10 Piura 59 15 60 10 91 22 66 17 75 12 80.4 22 49.6 12 Puno 70 10 51 17 91 21 76 11 75 11 96.2 8 43.5 14 San Martin 44 22 43 23 95 9 55 24 73 15 90.4 13 36.3 18 Tacna 84 2 81 2 95 5 88 1 89 2 97.9 5 67.3 2 Tumbes 74 8 89 1 96 3 84 4 82 6 96.1 9 61.1 8 Ucayali 57 18 45 22 87 24 72 15 64 18 83.9 20 33.1 20 PERÚ 68 59 93 73 73 91 51 Source: INEI, ENAHO 2006. 80 81 What can a Regional Government do to improve education? The Case of Junin Learning achievements Learning achievements: sufficient performance (% students) Region 2º 2º 6º 6º 5º 5º Primary: Ranking Primary: Ranking Primary: Ranking Primary Ranking Secondary: Ranking Secondary: Ranking Comm. Math Comm. : Math Comm. Math Amazon 2.7 22 9.4 11 3.7 22 1.1 22 6.8 14 0.3 23 Ancash 10.9 12 6.4 12 11.7 9 6.9 11 6.6 15 3.2 9 Apurimac 2.3 24 3.2 21 1.2 25 0.7 24 3.2 24 1.4 16 Arequipa 31.7 3 18.8 4 20.0 4 15.0 3 13.7 4 3.4 7 Ayacucho 4.6 19 3.5 19 5.3 17 2.6 19 8.2 9 3.0 10 Cajamarca 5.1 18 4.5 16 4.8 19 3.8 15 4.5 21 1.2 18 Callao 26.9 5 16.7 5 20.9 3 9.0 8 14.8 3 3.3 8 Cuzco 3.3 20 2.0 23 3.0 23 1.8 20 6.6 16 2.4 12 Huancavelica 2.8 21 1.7 24 4.1 21 4.1 14 5.9 18 2.3 14 Huanuco 6.5 16 4.6 15 6.8 15 1.6 21 4.2 22 1.9 15 Ica 20.3 7 11.4 7 14.2 7 10.3 5 6.4 17 0.8 20 Junin 12.1 9 12.0 6 7.0 14 7.1 10 8.9 8 4.6 2 La Libertad 20.5 6 11.0 9 15.6 6 9.9 6 12.4 6 2.3 13 Lambayeque 18.1 8 11.4 8 13.2 8 8.9 9 11.6 7 3.8 6 Lima 31.5 4 19.1 3 21.6 2 15.0 2 13.2 5 4.2 3 Loreto 2.0 25 2.4 22 2.0 24 0.4 25 5.5 19 0.7 21 Madre de Dios 10.7 13 5.8 13 11.2 10 3.0 17 5.1 20 0.0 25 Moquegua 34.0 1 25.7 2 19.1 5 13.1 4 17.7 2 3.8 5 Pasco 7.7 14 3.4 20 8.3 12 9.1 7 6.8 13 3.9 4 Piura 11.3 11 4.1 17 8.9 11 4.9 12 8.1 10 2.6 11 Puno 6.2 17 3.9 18 6.8 16 4.4 13 1.6 25 0.4 22 San Martin 7.6 15 4.9 14 8.1 13 3.5 16 7.7 11 0.0 24 Tacna 33.2 2 29.3 1 23.9 1 15.1 1 21.3 1 6.5 1 Tumbes 11.5 10 9.5 10 5.0 18 2.9 18 6.9 12 1.2 17 Ucayali 2.7 23 1.4 25 4.6 20 0.8 23 3.7 23 0.9 19 PERÚ 15.1 9.6 12.1 7.9 9.8 2.9 Sources: MED-UMC ­ National Assessment of Learning Achievements 2004. 81 82 What can a Regional Government do to improve education? The Case of Junin Annex 3.2. Fieldwork research To complete existing statistical information, we conducted an in-depth assessment through a random sample of 86 schools in all the region provinces. Sample design We chose a stratified design based on the number of students enrolled in each type of school (single teacher, multi-grade, full grade). As such, we divided the schools in Junin into three strata and within each stratum we made a random selection of schools using the number of enrolled schools as a weighting factor. Nevertheless, aware of the considerable variability in learning results and teaching conditions among the different school types, we opted to ensure sufficient representativity within each stratum in order to conduct separate statistical analyses with adequate precision. As such, the final sample design disproportionately favors single teacher schools (41% of the total of schools), which have less probability of being selected due to the lower number of students (9% of students total). Schools were selected randomly in each stratum by using a list ordered by UGEL, district and number of students. The pre-selection contemplated the possibility of sample loss for different reasons (accessibility, resistance from school authorities, class cancellation) and assigned substitute schools following the list, (choosing the schools before or after the lost school on the list), so that the substitute would have the same characteristics as the school lost in terms of size and location. The selected sample included ninety schools, which allowed us to obtain a sample of 5% of the total of primary public schools in Junin, even if there was sample loss in the process. Indeed, difficulties encountered during fieldwork led to a final sample of n = 83 schools, which corresponds to 4.93% of the total number of schools in the sample framework (N = 1,685). In each school, we conducted a random selection of one section of second grade, one section of third grade and one section of the last grade available at the school (in general sixth grade). This random selection was conducted in the presence of the principal and grade teachers (due to resistance to assessment). In second and sixth grade, we made a random selection of five students from each selected section to administer an interview or oral test. In third grade, we made a random selection of ten students to administer an interview and an oral test and a written test was administered to all (except in schools that would not allow it). Adjustment factor Although the sample design chosen allows us run analyses of each stratum, when running statistical analyses on the sample set it is necessary to correct the bias introduced by the sampling method and adjust the sample based on selection probability. This is done by using an adjustment factor (AF), based on the calculation of the representation factor (RF). The RF for each student in the sample is equivalent to the total number of students in his or her stratum within the sample framework divided by the total member students in his or her stratum in the sample. For example, in the sample framework there are 15,892 students enrolled in single teacher schools. In the sample, we have evaluated 211 students from single teacher schools. As such, each of these students 15,892/211 = 75. The RF from students in multi-grade schools is much higher (202) because they are under-represented with regard to the population The AF for a given stratum is calculated from its RF as a percentage of the sum of all the three AFs. Thus, by increasing the representation of under-represented strata we ensure that the statistical analysis of the sample as a whole maintains proportions that are similar to those of the population in terms of school type. 82 83 What can a Regional Government do to improve education? The Case of Junin Annex Table 3.2-a. Representation factor per stratum Single Teacher Multi Grade Full grade Total Sample 211 331 624 1.166 Number of students Sample framework 15,892 66,954 88,086 170,932 Students (%) Sample 18 28 54 100 Sample framework 9 39 52 100 Sample 36 23 24 83 Number of schools Sample framework 688 813 184 1,685 Schools (%) Sample 43 28 29 100 Sample framework 41 48 11 100 Representation Factor 75 202 141 419 Adjustment factor 18 48 34 100 Description of the sample framework Annex Table 3.2-b. Distribution of public primary schools according to UGEL, type and area Single teacher Multi-grade Full grade Human Human Resident Total Rural Urban Rural settlemeMarginal Urban Rural settleme Marginal Urban ial UGEL nts urban nts urban urban. Chanchamayo 217 139 3 3 4 5 3 4 6 384 Chupaca 9 55 2 9 6 81 Concepcion 45 67 1 8 3 5 1 130 Huancayo 62 129 1 1 7 32 2 11 11 10 266 Jauja 16 77 8 17 6 1 13 138 Junin 39 19 5 1 1 3 4 72 Satipo 210 177 3 1 4 3 4 402 Tarma 76 69 2 1 2 3 17 170 Yauli 12 1 12 1 4 3 8 41 Total 686 1 744 4 23 42 66 5 28 74 11 1.684 Source: School census 2006. Sample description Annex Table 3.2-c. Schools in the sample according to UGEL and school type UGEL Single teacher Multi-grade Full grade Total % Chanchamayo 9 7 4 20 24 Chupaca 0 1 2 3 4 Concepcion 2 0 2 4 5 Huancayo 3 0 13 16 19 Jauja 0 1 3 4 5 Junin 2 0 1 3 4 Satipo 10 10 4 24 29 Tarma 3 2 2 7 8 Yauli 0 0 2 2 2 Total 29 21 33 83 100 Annex Table 3.2-d. Schools in the simple according to UGEL and school location UGEL Rural Urban Urban Marginal Total % Chanchamayo 17 0 3 20 1.2 Chupaca 2 1 0 3 0.2 Concepcion 3 1 0 4 0.2 Huancayo 9 5 2 16 0.9 Jauja 3 1 0 4 0.2 Junin 2 1 0 3 0.2 83 84 What can a Regional Government do to improve education? The Case of Junin UGEL Rural Urban Urban Marginal Total % Satipo 22 2 0 24 1.4 Tarma 5 2 0 7 0.4 Yauli 1 1 0 2 0.1 Total 64 14 5 83 4.9 % 3.8 0.8 0.3 5 Annex Figure 3.2-a. Percentage of the schools in the sample with students whose mother language is not Spanish, according to mother language Quechua 43.0 Aimara 0.0 Ashaninka 20.5 Nomatsiguenga 0.0 Machiguenga 1.3 Aguaruna 0.0 Shipibo-Conibo 1.3 Huitoto 0.0 Mastanahua 0.0 Another language 3.8 0 5 10 15 20 25 30 35 40 45 50 Annex Table 3.2-e. Students evaluated by UGEL and grade Type of evaluation 2º 3º 4º 5º 6º Total Oral test 334 57 3 4 288 686 Written test 0 526 0 0 0 526 Both test 0 475 0 0 0 475 Total 334 1.058 3 4 288 1,687 Annex Table 3.2-f. Students evaluated by UGEL and type of evaluation in which they participated UGEL Only interview Oral test Written test Both tests Total Chanchamayo 0 239 155 85 309 Chupaca 0 60 38 20 78 Conception 2 57 55 29 85 Huancayo 0 257 314 129 442 Jauja 0 74 71 34 111 Junin 0 32 20 11 41 Satipo 0 326 235 126 435 Tarma 1 80 64 25 120 Yauli 0 36 49 16 69 Total 3 1,161 1,001 475 1,690 % 0.2 68.7 59.2 28.1 100 84 85 What can a Regional Government do to improve education? The Case of Junin Annex Table 3.2-g. Students evaluated with the oral test, according to UGEL and mother language UGEL Spanish Quechua Aimara Ashaninka Nomatsiguenga Shipibo Total % Chanchamayo 190 17 0 30 0 1 238 20.5 Chupaca 56 4 0 0 0 0 60 5.2 Concepcion 60 0 0 0 0 0 60 5.2 Huancayo 236 19 1 1 0 0 257 22.1 Jauja 70 3 0 0 0 0 73 6.3 Junin 31 1 0 0 0 0 32 2.8 Satipo 242 14 0 68 1 1 326 28.0 Tarma 81 0 0 0 0 0 81 7.0 Yauli 35 1 0 0 0 0 36 3.1 Total 1,001 59 1 99 1 2 1,163 100.0 % 86.1 5.1 0.1 8.5 0.1 0.2 100 Instruments The Information Sheet on Schools gathers general school information such as geographic location; number of students, teachers and classrooms; existence of an APAFA or CONEI; infrastructure conditions, among others. This sheet contains information gathered from the school's administration and field observation. The individual student interviews follow a closed interview protocol to collect information on students and some opinions regarding the schools and its teachers. It is administered individually, orally and right before initiating the reading test. (Time estimated for administration: three minutes). The fluency and reading comprehension test measures reading fluency and comprehension in primary students. Second and third grade students were selected as well as students in the last year available in each school. This allowed us to see if the students who had completed second grade (that is those who are in third grade) could read and write at the expected level as well as determine the reading level of students who are about to finish school. The first level consists of a basic test of sixty words whose complexity is equivalent to second grade of Primary. Each student has to read out loud and the surveyor records the number of words read correctly and the time it took to read them. Next, the students answer some comprehension questions. If performance is satisfactory, this procedure is repeated with a second text whose complexity is equivalent to fourth grade. If the student cannot read the first text, he is asked to read separate words, and if he is unable to read the words he is asked to identify some letters. Students will be classified according to reading level: none, letters, words, simple text, complex text. (Estimated administration time: seven to ten minutes per student). The Reading comprehension test prepared by MED and used in the 2006 census evaluation was administered to all third grade students in the selected classrooms. (Estimated time of evaluation: fifty minutes). Classroom observation is designed to gather information on teaching and learning practices in the classroom that have been identified in international literature and research as factors associated with school effectiveness in poor sectors. This instrument has various sections: description of the environment; equity in gender and ethnicity; a record of the time allocation to teaching and learning activities; teaching practices that facilitate learning, and additional observation notes. (Estimated administration time: fifty minutes). Interviews to teachers, principals and parents gather information on institutional factors associated with school effectiveness, such as leadership quality, the existence of a shared vision and objectives, participation of teachers and parents in decision making, positive work environment, supervision, institutional follow up and assessment. In order to avoid distortion due to social desirability, we chose not 85 86 What can a Regional Government do to improve education? The Case of Junin to ask directly about expected good practices, because the subject can easily identify the "desired" response in such type of questions. The open nature of the responses enables us to obtain a rich, detailed description of the aforementioned factors even when subjects are not explicitly asked about them. The interviews are structured to select among a given set of answers (selection made by the surveryor based on the oral answer provided by the subject), but at the same time allow us to record the subjects' textual responses. The objective of this format is to facilitate subsequent codification without losing the richness of the testimonies gathered, because the real meaning of the data is often found in details and anecdotes. Interviews to teachers and principals cover the same areas although the questions differ. The areas explored are: the strengths and problems that each school faces. The institutional climate at the school and teacher satisfaction, relationships with parents, assessment and accountability at the school, decision- making processes at the school and the existence of a shared education project. Interviews to principals also included the relationship with and perceptions about the APAFA. (Estimated administration time: forty to sixty minutes) Interview to parents were shorter and explored the strengths and problems that the school faces. The interview asks about their relations with teachers and school, the relation with and perceptions about the APAFA, and participation in accountability and decision-making processes. (Estimated administration time: ten minutes). Each fieldwork researcher also kept a registry with observations about the schools visited, individuals interviewed, difficulties in the process, the limitations or scope of the instruments, etc. Annex Table 3.2-h offers a summary of the principal problems found in the schools where it was not possible to complete the assessment. Annex Table 3.2-h. Problems found in the evaluation Modular date Code (day/month) Result Problem 375667 May 21 Incomplete evaluation We were unable to interview two parents because there was not enough time. Teacher uprising in the middle of the evaluation: they called the media to 364513 May 21 Incomplete evaluation harass interviewers. We were unable to finish classroom observation or complete testing in the third grade. Incomplete Teacher uprising in the middle of the evaluation: they called the media to 364513 May 21 evaluation harass interviewers. We were unable to finish classroom observation or complete testing in third grade. May 23 Not evaluated SUTEP strike. 375683 May 22 Not evaluated The principal came late from a meeting and we were unable to coordinate. May 24 Not evaluated Planned visit but the school was closed due to a principal's meeting. 552265 May 28 Incomplete evaluation There were no parents to interview. We visited the school on two occasions. The first time, group 4 made the visit, but due to rain, the teacher had not attended school. Group 9 visited the school 925453 May 22, 25 Incomplete for the second time and was only able to administer some surveys to students and 28 evaluation and interview one parent. The third visit was conducted on May 28, and we found one teacher to whom we could administer all the tests. We were unable to interview a parent. We were unable to interview the principal because he had traveled to Satipo. 374520 May 23 Incomplete evaluation Due to a lack of time, we did not conduct classroom evaluation of the second grade and did not interview parents. We were unable to interview parents due to a lack of time. Parents were 1036961 May 23 Incomplete evaluation generally on their fields far away and performing agricultural activities (it was a native Ashaninka community). 795278 May 23 Incomplete We were unable to interview parents. In many areas, the majority of parents evaluation had left for their fields early in the morning and did not return until nightfall. 1101856 May 24 Incomplete A trip had been planned for students on this date. Nevertheless, due to heavy 86 87 What can a Regional Government do to improve education? The Case of Junin Modular date Code (day/month) Result Problem evaluation rains and student absenteeism, the trip was cancelled. As such, we did not administer the test to second grade students (none had come to class) and no observation took place. May 24 Incomplete It was the school's anniversary and celebrations were being held. We were evaluation only able to work with the principal. 1037043 We were able to administer almost all the instruments, but only one sixth grade May 25 Incomplete evaluation student took the test. Due to the lack of time, it was not possible to conduct classroom observation in the second grade. There were no classes because students and teachers were attending a school anniversary celebration in a neighboring town (school mentioned in the 1101419 May 24 and Incomplete previous item). We were only able to work with the principal. On the second 25 evaluation visit, we were able to administer almost all the instruments. Due to the lack of time, we were unable to administer the test to a sixth grade student, conduct classroom observation on the second grade or interview a parent. This visit was not planned (after returning to Puerto Ocopa) but failed because 1215755 May 25 Not evaluated the school was not in session. This was the last Friday of the month and teachers generally gather in the Education Red (Red Educativa) to pick up their checks. Later, the task was completed by group 7. School closed; we found only ten students and parents in the street and 374736 May 25 Incomplete evaluation administered the tests. We opted to visit the school the second time. The visit was conducted by another team. 377655 May 25 Incomplete evaluation The parents did not want to respond to the interview. The first time, the visit planned was cancelled because there was no gasoline in Puerto Ocopa and no boat was available to leave the port. We decided to return to Satipo to visit another school but due to difficulties with access, some members of the team didn't arrive. 624585 May 25 and Complete The second visit was conducted on the 29th and 30th of May: This time we 29 administration were able to return to Puerto Ocopa and reach the community by boat. The main difficulty was the path, which was very steep and narrow. The walk to the community took about three hours. We only administered the MED survey because the other students in the third grade could not read and couldn't even distinguish the letters on the test. Return transport did not arrive on time, which meant that the evaluator had to 659300 May 28 Complete administration walk approximately three hours to reach a town from which she could take a boat to Satipo. The tests were administered to the only three students in the third grade. Due 374736 May 28 Incomplete evaluation to a lack of time, we were unable to interview parents. The majority of parents were on their fields tending to crops. The only two students enrolled in the third grade did not attend school. Only 1215755 May 29 Incomplete two second grade students attended and another two students from the six evaluation grade, which meant that only four tests could be administered. We could not interview two parents because the majority was on their fields. 379602 May 30 Incomplete evaluation There were no parents to interview. The principal/teacher and the second teacher explicitly expressed opposition to 558668 May 30 Incomplete assessment and observation despite the fact that they knew about our visit. evaluation They would not let us interview parents. We were only able to interview some students. There were no parents to interview. We did not finish with the sixth grade 379123 May 31 Incomplete because there was no place to take the evaluation and it was too cold to do so evaluation on the patio. Additionally, there were many other activities going on and there was a lot of movement. There were no problems reaching the school, but we had to wait for a few 382390 May 31 Incomplete evaluation minutes until the cargo raft had enough cars to cross the river. We were unable to administer interviews to two parents. The teacher had not planned a class due to low student attendance (50%). 369355 June 1 Incomplete evaluation Nevertheless, he improvised a class that he called "reinforcement" (went over work from past classes). While he was conducting the class, he constantly 87 88 What can a Regional Government do to improve education? The Case of Junin Modular date Code (day/month) Result Problem spoke to the evaluators and repeated this phrase "we can't work this way." The children that had attended did not match the sample specifications for test administration (only two students from the first grade and three from fourth). With regard to parents, the majority had been in their fields since early, which made it impossible to conduct interviews. We were only able to talk to two adults, one of which was the mother of three students at the school. 610121 June 1 Not evaluated Planned visit, but cancelled because classes were not held due to sporting events. 378869 June 1 Not evaluated Planned visit, but the school did not have classes due to sporting events. School closed due to principal-teacher meeting with the Red. Parents 558775 June 1 Not evaluated mentioned that "the teachers went to pick up their checks" and could not be interviewed due to a "lack of time". It was not possible to interview the principal because the researcher's visit 31494 June 4 Incomplete evaluation coincided with a visit from a UGEL specialist. At noon, the principal decided to accompany the specialist to the highway and left school. The teacher informed the team that he had planned a Logic-Mathematics exam 31819 June 5 Incomplete evaluation and afterwards the students should spend the rest of the morning doing art work for the school's anniversary. There were no parents in the town because they had left for their fields to 30899 June 5 Incomplete harvest coffee. During the last segment of the observation, the students evaluation verbally expressed that they were tired. The teacher ended the class. The observation lasted forty minutes. There were no parents in town because they were in their fields harvesting 31831 June 6 Incomplete evaluation coffee. The children told interviewers that teachers abused them physically and verbally. 30912 June 6 Incomplete We were only able to interview three parents that were near the school prior to evaluation departing for their fields. The principal called teachers to a coordination and evaluation meeting before the evaluators could finish the administration of all instruments. According to the principal, it was impossible to talk to parents because all of them had left for their fields very early and did not come to the school. Despite this, we were able to do three interviews. At 10:30 in the morning teachers met in the principal's office to participate in a drawing and choose the classes with which the evaluators would work. The teachers were very worried and asked the researchers for their credentials and 30410 June 7 Incomplete documents. They ignored the principal's call to begin the drawing and asked evaluation the team to explain once again the visit's objective. The principal had to practically force the teachers to participate in the drawing. Work began at 10:45 in the morning. Third grade classroom observation lasted barely forty minutes because the teacher (who had an authoritarian, defiant attitude) had to go to another school with members of the students brigade who were going to participate in a flag-raising ceremony. This teacher and the second grade teacher were part of the committee and as such they could not be interviewed. Due to initial delays, we were only able to interview eight students from the third grade. We were only able to interview one parent because the majority of parents had 30446 June 8 Incomplete evaluation left early for their fields and most spent the week there. During the interviews, the children said that the teachers abuse them physically. Planned visits but we were unable to find the school because the UGEL had provided poor instructions. We decided to go to the Bellavista (El Tambo). School. The third attempt to visit was cancelled due to the SUTEP strike called 785634 June 19, 20 and 22 Not evaluated for Monday the 18th. Although the Presidents of APAFA and the CONEI approached the team, the comuneros (community members) openly forbade them from talking with researchers, arguing that all the comuneros needed to be present so that opinions could be seconded. 382374 June 19, 20 and 21 Not evaluated Visits cancelled successively due to SUTEP strike called for Monday the 18th. 365114 June 20 and 21 Not evaluated Visits cancelled due to the SUTEP strike called for Monday the 18th. 377127 June 21 Incomplete The school was closed due to a teacher's strike. We were only able to work 88 89 What can a Regional Government do to improve education? The Case of Junin Modular date Code (day/month) Result Problem evaluation with six children that were wandering around the town. The parents were on their fields and it was impossible to talk with any of them. Because the school was far from the city of Huancayo and access was very complicated, the head researcher was unable to conduct another visit (the return trip took four hours). Annex 3.3. Evaluation of reading competences Reading and writing constitute the foundation for all school learning. Students that cannot read fluently are unable to benefit from school and will have difficulty in achieving autonomy to access the knowledge that the school provides. Peruvian children begin reading and writing in the first grade of primary, as do most children around the world. For many, this may be a simple, almost natural learning experience. Nevertheless, becoming a good reader and expressing ideas in writing is a tough, complicated task. In order to learn to read and write, children must have achieved a set of abilities: language development, symbolic function, the ability to focus and follow instructions, sequence and auditory and visual memory, special perception skills and laterality. Decodification and comprehension are the two main reading skills. Decodification is the ability to pronounce a written word and is generally measured through fluency, which is understood as the ability to read with agility, precision and intonation. Comprehension is the process by which the meaning of the written word is extracted through the reader's previous knowledge. National and international research has shown time and again that fluency and comprehension are related, for in order to understand the meaning of a text, the reader must be able to decodify the text quickly and easily (Abadzi, 2006; Crouch, 2006; Samuels, 2002). Reading abilities can be evaluated through oral or silent reading. With regard to fluency, as a child progresses in his or her knowledge, the number of words that he or she can read per minute increases significantly. A recent study of the World Bank in Peru (Crouch, Abadzi, Echegaray, Pasco, & Sampe, 2005), determined that a child can read an average of 30 words per minute orally in first grade, 60 in second grade, 90 in third grade, 110 in fourth grade, and continue to make progress throughout the following grades. This coincides with previous studies which show that second grade students can read between 30 and 54 words orally in second grade, between 66 and 104 in third grade, between 86 and 124 in fourth, between 95 and 130 and in fifth and between 108 and 140 in sixth. Silent reading tends to be quicker (Manzo & Manzo, 1993). According to the National Curricular Design 2005, by the end of the second grade students should be able to recognize "characters, important facts, materials, ingredients and instructions" (MED, 2005, p. 119) in narrative and descriptive text; should be capable of inferring facts and ideas through clues and previous experience; and are able to provide an opinion on the text's facts and ideas. In the third grade, students will strengthen their competences, become familiar with informative texts and capable of recognizing data, topics and places and inferring "important information on the topic by means of explicit data and previous knowledge" (MED, 2005, p. 119). In the fourth grade, students can understand expositive text, recognize the type of text they are reading, identify the main and secondary ideas, topics and procedures, and infer the purpose for which the text was written. In the fifth grade, students are able to read argumentative texts, to use the details of the text to make inferences, and express a critical opinion regarding the text's contents and form. Finally, at the end of primary, students are expected to be able to make inferences through data that is implicit in the text. 89 90 What can a Regional Government do to improve education? The Case of Junin Annex 3.4. Calculating continuous values in reading performance The fluency score in the oral test is calculated using the maximum number of words read correctly in one minute. For those who are able to read both texts, the score is based on the reading for which they read the most words (in general, this corresponds to the second reading because it has more words). The final score for comprehension is calculated by adding the scores in section in A1, A2 and B2, all of which have an internal reliability coefficient of = 0.88. We decided to exclude section BI for various reasons: (i) tis item reduces internal reliability of the test ( = 0.83); (ii) it is too easy and hence has low discriminative power; (iii) it is the item that correlates the lowest with fluency, the census test and global comprehension scores and there are too many missing cases due to the fact that the instructions for administration indicated that A1 section should only be taken (for time reasons) if the students were unable to respond satisfactorily to section one; and (iv) despite the fact that the same instructions were given for section B2, this section is seen as contributing to test's reliability because it discriminates skills and measures a comprehension level that falls between A1 and A2. According to the standards identified by the World Bank (Abadzi, 2006; Crouch, 2006; Samuels, 2002), students in the third grade increase reading fluency every year by approximately 30 read words per minute (from 30 in first grade to 60 in second grade and 90 in third grade). Given that the school year should have no less than forty weeks of education work (ministerial resolution 0712-2006-ED, 14/11/2007), in an average week the student's reading fluency should increase by 0.75 words a minute. After the third grade, the reading curve is expected to be less steep, with an increase of one word per minute every two weeks until reaching 110 at the end of fourth grade and at least 140 words at the end of sixth grade. Our evaluation took place between May 20 and June 20, so we expected that the students from second and third grade had reached the standards for the prior grade (that they had already finished), plus the equivalent of eleven weeks of class assuming that classes had begun on March 1st, as stipulated by MED. At the time of the evaluation, second grade students should have been able to read at least 38 words per minute, and third grade students, 68 words. In sixth grade, they should be able to read at least 130 words. These referential values allow us to situate the performance of students evaluated within a continuum and compare progress not only with regard to prior grades but also relatived to the standard for their grade. 90 91 What can a Regional Government do to improve education? The Case of Junin Annex 3.5. Comparison of Items with Curricular Objectives Annex Table 3.5-a. Correspondence between the section two of the tests administered and reading competences according to the current Curricular Design Competences acquired for reading comprehension by grade according Sections of the tests administered Reading Level to the Current Curricular Design (MED, 2005) that correspond to these competences 1º 2º 3º 4º 5º 6º Oral reading test MED test Reads letters Names letters 24 Reads words Identifies words 25 1, 2, 3, 4, 5 Reads sentences Understands sentences 6, 7, 8, 9 Reads texts Narrative Text 1 and Text 2 Text 1 + descriptive and instructive Text 3 + informative Text 2 + expositive + argumentative Recognizes characters, characteristics and actions A1: 1a, 1b Narrative: 10, 11 + important facts, materials, ingredients and instructions A1: 1a, 1b, 1c Descriptive: 19, 20 Informative: 16, + data, topics, location A1: 1a, 1b, 1c, 1d B2: 3, 5, 6, 7 17 Descriptive: 21 + type of text, main and secondary A1: 1a, 1b, 1c, 1d ideas, topics, procedures A2: 1a, 1b, 1c, 1d B2: 3-7 Infers Characters' actions through illustrations, title, words, phrases and previous Narrative: 12, experiences 13, 14 + facts and ideas through clues and previous experiences + important information and the topic through Narrative: 15 explicit data and previous knowledge Descriptive: 22 + Infers the objective Informative: 18 Descriptive: 23 + through the text's details + through implicit data Expresses Characters and actions through previous experiences A1: 2a opinions about + facts, ideas A1: 2a, 2b + data, topics, A1: 2a, 2b + important ideas, objective and A1: 2a, 2b topic discussed A2: 2a, 2b Expresses critical + contents and form opinions about 1º 2º 3º 4º 5º 6º Percentage of achievement estimated for the oral test based on the curriculum In the oral test, we expect first grade students to remember at least two facts from the story in chronological order, and identify what they liked the most (4 points = 16%). At the end of the second grade, we expect students to narrate the main events of the story in chronological order, describe the characters, identify what they liked the most and explain why (8 points = 32%). In third grade, in addition to prior achievements, we expect students to provide more details on the first story and answer at least four out of five literal questions on the second test (13 points = 52%). In fourth grade, in addition to the previously mentioned skills, students should be able to name the facts of the second story in their own words and in chronological order as well as identify what they liked the most (20 points = 80%). In fifth 91 92 What can a Regional Government do to improve education? The Case of Junin grade, they should be able to provide additional details on the second story and explain why they chose a particular section as their favorite part and provide more complete answers to literal questions (23 point = 92%). In sixth grade, students should be able to obtain a perfect or near perfect score (24 a 25 points = 96- 100%). Annex Table 3.5-b. Expected score per item and grade for the oral reading test Reading Level Section 1º 2º 3º 4º 5º 6º 1. Dogo Level A 1.a. recalls Went on a walk with his owner 1 1 1 1 1 facts Got lost 1 1 1 1 1 1 They found him 1 1 1 1 1 1 They took him home 1 1 1 1 1 1.b. Narrates the story in order 1 1 1 1 1 1 1.c. Describes the characters 1 1 1 1 1 1.d. Provides additional details 1 1 1 1 2.a. Identifies what he or she likes most about the story 1 1 1 1 1 1 2.b. Describes what they liked most 1 1 1 1 1 Subtotal Text 1 4 8 9 9 9 9 2. Level A 1.a. recalls Carmen finds a cat [1 pt.] 1 1 1 Carmen facts They become good friends [1 pt.] 1 1 1 and The cat had kittens [1 pt.] 1 1 1 Ceniza They gave away the kittens [1 pt.] 1 1 1 1.b. Narrates the story in order 1 1 1 1.c. Provides additional details 1 1 2.a. Identifies what he or she liked most about the story 1 1 1 2.b. Describes why he liked it 1 1 Level B 3. What was the cat's name? [Ceniza = 1 pt.] 1 1 1 1 4. What did the cat like to do? [play with wool, go up the roof, 1 2 3 sleep on Carmen's legs = 1 a 3 pts.] 5. What did Carmen say to her mother [that they were going to give 1 1 1 1 the kittens away = 1 pt.] 6. Who took the kittens away? [A man = 1 pt. ; a good man = 2 pts.] 1 1 1 2 7. Were the kittens happy? [Yes = 1 pt.] 1 1 1 1 Subtotal Text 2 0 0 4 11 14 16 Total score anticipated per grade 4 8 13 20 23 25 Percentage of achievement anticipated (Score anticipated / total score) 16% 32% 52% 80% 92% 100% Percentage of achievement estimated for the census test based on the curriculum To complete these results, we conducted a similar exercise with the census test (administered only to the third grade students). According to the competences stipulated in the curricular design, students are expected, at the end of the first grade, to be able to identify words (questions 1 to 5), understand sentences (questions 6 to 9), recognize characters, characteristics and actions in a narrative text (questions 10 and 11), and infer the character's actions based on illustrations, title, words, sentences and previous experiences (questions 12 to 14). At the end of the second grade, students should be also able to recognize important facts, materials, ingredients and instructions in descriptive texts (questions 19 and 20). At the end of third grade, students should be able to recognize data, topics and locations (questions 16, 17 y 21), and infer important information on the topic based on explicit data in narrative, descriptive and informative texts and previous knowledge (questions 15 and 22). After the fourth grade, students should be able to infer the text's purpose (questions 18 and 23). 92 93 What can a Regional Government do to improve education? The Case of Junin Annex Table 3.5-c. Expected score by section and grade on the silent reading test 1º 2º 3º 4º 5º 6º 1-14 1-14, 1-14, 1-14, All All 19-20 15-17, 15-17, Questions that must be answered correctly 19-20, 19-20, 21-22 21-22, 18, 23 Anticipated score 14 16 21 23 23 23 Anticipated achievement percentage (score anticipated / Total score) 61 70 91 100 100 100 Annex Table 3.5-d. Definition of performance level Census test Oral test Percentage of Percentage of Level Score obtained achievement Fluency* comprehension Level 0 0-9 0-39 0-29 0 Level 1 10-15 40-69 30-59 1-19 Level 2 16-18 70-80 60-109 20-59 Level 3 19-23 80-100 110+ 60-100 *Words per minute. Annex 3.6. Differences in the percentage of achievement on the census test (May 2007) between schools that participated in the 2006 census and those that did not Annex Table 3.6-a. Median differences in scores for the census test administered in 2007 for students from schools included in the 2006 census and students from schools that were not included in the census Standard Standard Group Obs Median error Deviation [95% trust interval] Not included in the census 732 6831789 0053776 1454932 6726215 6937362 Census school 269 6731857 009798 1606997 6538948 6924766 Combined 1001 6804934 004732 1497151 6712075 6897793 Diff 0099932 0106752 0109553 0309417 diff = mean(No Census) - mean(Censed) t = 0.9361 Ho: diff = 0 degrees of freedom = 999 Ha: diff < 0 Ha: diff != 0 Ha: diff > 0 Pr(T < t) = 0.8253 Pr(|T| > |t|) = 0.3494 Pr(T > t) = 0.1747 Annex Table 3.6-b. Correlation between census results in 2006 and 2007 Percentage of students with satisfactory achievement (2006) Results 2007 Level 1 Level 2 Level 3 Total 0.2038* 0.2475* 0.2420* % of achievement 0.2045* 0.2481* 0.2424* Note: Only shows significant correlation to 10% (sig<.1) * sig<.01 Annex 3.7. Socio-demographic and institutional factors associated with reading performance In order to isolate the effect of each characteristic associated with the school and the students in terms of fluency, comprehension and performance on the census test, we conducted multivariate linear regressions with an adjustment factor to correct the disproportions caused by sampling. Dependent linear regression variables Fluency = Number of words read per minute Comprehension = Percentage of correct answers for both texts* 93 94 What can a Regional Government do to improve education? The Case of Junin MEDachievement = Percentage of correct answers on the census test Group variable: (second, third and sixth primary) * Regressions for Dogo corrected with the compre2 variable (percentage of correct answers on the first test: Dogo) for second and third grades, but these results were inconclusive and less robust than the comprehension variable in general. Annex Table 3.7-a. Results of fluency, comprehension and the census test per grade Grade Status Fluency Comprehension Census test 2º Median 25.9 16.7 . DS 24.4 23.7 . 3º Median 49.4 27.1 67.3 DS 25.4 28.0 15.6 6º median 86.1 54.6 . DS 26.3 29.8 . Dependent logit variables lflu = 1 if fluency>=standard and 0 if fluency=standard and 0 if comprehension =standard and 0 if MEDachievement