Report No: ACS18935 . Hashemite Kingdom of Jordan Education Sector Public Expenditure Review . June 17, 2016 . GED05 MIDDLE EAST AND NORTH AFRICA . 1 Standard Disclaimer: . This volume is a product of the staff of the International Bank for Reconstruction and Development/ The World Bank. The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. . Copyright Statement: . The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/ The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, http://www.copyright.com/. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. 2 Contents Acronyms ...................................................................................................................................................... 6 Executive Summary....................................................................................................................................... 7 I. Introduction .......................................................................................................................................... 9 II. Overview of Education Outcomes ...................................................................................................... 10 III. The Public Education Learning Environment .................................................................................. 19 IV. Overview of Public Education Spending ......................................................................................... 30 A. Capital Spending ............................................................................................................................. 32 B. Recurrent Spending ........................................................................................................................ 34 C. Syrian Refugee Students ................................................................................................................. 35 V. Teacher Compensation ....................................................................................................................... 39 A. Teachers Are Paid According to Qualification and Seniority .......................................................... 42 B. Teachers Are Relatively Well-paid .................................................................................................. 44 VI. Recommendations .......................................................................................................................... 49 A. Motivating Teachers to Perform ..................................................................................................... 49 B. Realigning Capital Investments ....................................................................................................... 54 C. Addressing Regional Variation ........................................................................................................ 56 References .................................................................................................................................................. 61 Boxes Box 1: Early Childhood Education in Jordan …………………………………………………………………………………………12 Box 2: USAID Support to the Education Sector in Jordan………………………………………………………………………33 Box 3: Strengthening Accountability in Jordan’s Education System ……………………………………………………..51 Figures Figure 1: Average years of school, Jordan and comparators (1970-2010) ................................................. 11 Figure 2: Gross enrolment ratios in kindergarten, basic, and secondary education, by governorate (2010- 11) ............................................................................................................................................................... 12 3 Figure 3: Gross enrolment ratios in kindergarten, basic, and secondary education, by gender (2013-14) .................................................................................................................................................................... 13 Figure 4: Total enrolment in kindergarten, basic, and secondary, by authority (2001-14) ....................... 15 Figure 5: Share of total enrolment in kindergarten, basic, and secondary education, by authority (2001- 14) ............................................................................................................................................................... 15 Figure 6: Jordan population projections, 2010-2050 (1,000s) ................................................................... 16 Figure 7: Jordan's performance in international student assessments ..................................................... 17 Figure 8: Variation in 2011 TIMSS scores, by gender and geographic location ......................................... 18 Figure 9: Growth in numbers of schools, classes, teachers, and students, by authority (2001-14) .......... 20 Figure 10: Distribution of school size, Ministry of Education and private schools, by geographic location (2014) .......................................................................................................................................................... 23 Figure 11: Distribution of class size, Ministry of Education and private schools, by geographic location (2014) .......................................................................................................................................................... 24 Figure 12: Distribution of student-teacher ratio, Ministry of Education and private schools, by geographic location (2014) ............................................................................................................................................ 25 Figure 13: Correlation between GDP per capita and government spending on primary education as a share of GDP, 2014 ..................................................................................................................................... 31 Figure 14: Correlation between GDP per capita and government spending on secondary education as a share of GDP, 2014 ..................................................................................................................................... 32 Figure 15: Number of Ministry of Education teachers, 2001-13 ............................................................... 41 Figure 16: Teacher salaries relative to per capita GDP, 2012 .................................................................... 45 Figure 17: Distribution of public school teachers by age and gender, 2010 .............................................. 47 Tables Table 1: Jordan’s gross enrolment ratios for all levels of education, 1970-2012 ...................................... 10 Table 2: Gross enrolment ratios for Jordan and comparators, 2012-13 .................................................... 11 Table 3: Jordan's schools, by authority, type, and gender (2013-14) ......................................................... 14 Table 4: Schools by governorate and authority, 2013-14 (percent share of total in governorate)........... 16 Table 5: Percent growth rate in school size, class size, and student-teacher ratio, by authority (2001-14) .................................................................................................................................................................... 20 Table 6: Trends in school size, class size, and student-teacher ratio, by authority (2001-14) .................. 21 Table 7: School size, class size, and student-teacher ratio, by region and geography (2014).................... 22 Table 8: Characteristics of schools with fewer than 100 students, 2014 .................................................. 26 Table 9: Characteristics of schools with average class size greater than 40 students, 2014..................... 26 Table 10: Tawjihi results, by school characteristic, 2013............................................................................ 28 Table 11: Regression results for tawjihi pass rates and school characteristics, 2014 ................................ 29 Table 12: Ministry of Education expenditures, as share of GDP and total education expenditures (2008- 13) ............................................................................................................................................................... 31 Table 13: Total capital spending, by budget entity (2013) ........................................................................ 33 Table 14: Education sector capital spending, by budget entity (2013) ...................................................... 33 Table 15: Ministry of Education capital spending, by program (2013) ...................................................... 34 Table 16: Total recurrent spending, by budget entity (2013) .................................................................... 35 Table 17: Unit cost, current expenditure (current JDs) .............................................................................. 35 Table 18: Education financial requirements summary table, Jordan Response Plan 2016-2018............... 37 4 Table 19: Ministry of Education recurrent spending, by program (2013) ................................................. 39 Table 20: Ministry of Education worker compensation, by program (2013) ............................................. 40 Table 21: Education sector staff, 2015....................................................................................................... 41 Table 22: Education sector staff basic salary scale, 2012 (JD/month) ........................................................ 42 Table 23: Allowances for education sector staff, 2013 .............................................................................. 43 Table 24: Main components of education sector staff compensation, 2013 ............................................ 43 Table 25: Average education sector staff pay, by governorate (2015)...................................................... 46 Table 26: Share of different education sector staff, by governorate (2015) .............................................. 46 Table 27: Teaching hours per year ............................................................................................................. 48 Table 28: Policy Options for the Short- and Medium-Terms ...................................................................... 58 5 Acronyms EMIS Education Management Information System ERFKE II Education Reform for the Knowledge Economy II GDP Gross Domestic Product GER Gross enrolment ratio JD Jordanian Dinars MENA Middle East and North Africa MOE Ministry of Education MOF Ministry of Finance MOPIC Ministry of Planning and International Cooperation OECD Organisation for Economic Co-operation and Development PER Public expenditure review PISA Program for International Student Assessment SCD Systematic Country Diagnostic STR Student-teacher ratio TALIS Teaching and Learning International Survey TIMSS Trends in International Mathematics and Science Study UNRWA United Nations Relief and Works Agency for Palestine Refugees in the Near East USAID United States Agency for International Development VTC Vocational Training Corporation 6 Executive Summary Jordan invests substantial public resources in providing education services (3.5 percent of GDP for pre- tertiary education) – comparable to international averages and above what might be expected given its per capita GDP. It has made impressive strides in terms of schooling access and attainment – enrolment rates are on par with comparator countries, with the exception of pre-primary education enrolment rates, which remain low at 34 percent. Enrolment rates also show relatively large variation across governorates, as well as gender inequity in favor of females. Yet desired student learning outcomes have not been achieved, remaining below international averages – and here again there is marked variation across governorates. The low education quality has likely contributed to the increasing share of enrolment in private instead of public education, so that today the majority of schools in Amman are private schools, for example. Education policy makers have zeroed in on several structural dimensions of education services in order to address the issue of education quality. Policy makers have sought to address issues such as small schools and rented schools, on the one hand, and overcrowded classrooms and high student-teacher ratios, on the other. The latter concerns have become particularly pressing in recent years with the influx of Syrian refugee students into Jordan’s public education system. Yet the Jordanian system’s basic attributes are sound: on average, schools are smaller than the recommended size but not dramatically so; and class size (27) and student-teacher ratio (16) are comparable to OECD averages. The problems of very small schools, overcrowded classrooms, and overburdened teachers are quite localized and must be addressed as such. This in turn requires detailed planning and careful implementation of capital investments – something that is thwarted by the fact that oversight over capital resources for the education sector is spread across ministries. Another recent focus area for education policy makers has been teacher pay, where a doubling of the base salary was put in place recently. Compensation accounts for a high share of recurrent public education spending (92 percent), which already captures a large proportion of total education expenditure. Using all standard measures, teachers in Jordan today are relatively well-paid: by comparison with other tertiary-educated workers in the economy, relative to per capita, and even taking their teaching and work hours into account. Despite the relatively good pay, there is some evidence that teachers in Jordan do not exert the requisite effort in order to achieve desired learning outcomes. The teacher pay scale and allowances in Jordan reward initial qualifications, seniority, and personal teacher attributes, as opposed to being an instrument for policy makers to incentivize better teaching performance or other desired education sector outcomes – such as teaching of math and science, or deployment in certain geographic regions. Yet, without teachers putting forth appropriate levels of effort in their classrooms, investments in other schooling inputs are unlikely to yield significant improvements in education service delivery and student learning outcomes. As a result of the above summarized findings, this report makes recommendations related to three sets of issues: motivating teachers to perform; realigning capital investments; and addressing regional variation in outcomes. 7 Motivating teachers to perform in Jordan will require a shift towards performance-based education, whereby teacher accountability is at the heart of the sector reform agenda. Mechanisms will need to be put in place to determine actual teacher effort, and this teacher accountability will need to be rewarded through financial incentives, perhaps in the form of bonus pay. Existing allowances that do not incentivize performance will need to be phased out to allow for the introduction of other types of payments that can act as policy levers without further increasing the share of compensation in the budget. In terms of capital investments, the Ministry of Education must have oversight of all capital spending in the education sector even if it does not manage all its implementation. And in order to address infrastructural constraints in the most efficient manner, the Ministry must have access to and make use of detailed geographic data. In addition, it may very well be the case that some areas in the country cannot escape having very small or rented schools – whether for reasons of sparse population, gender segregation, or simply lack of land for construction of large schools. School clustering may therefore be an approach worth consideration for Jordan, whereby individual schools are able to maintain their identity and institution, while at the same time sharing resources (including teachers), experience, and good practices. Finally, the variation across governorates in terms of education access and quality outcomes requires further study and focused policy attention. Yet the Ministry is not structured in a manner to tackle the issue of varying education service needs by socio-economic level, for example, or for lagging regions. Nor does the budget preparation and execution process lend itself to specific focus on certain programs, being instead very much based on historical budgeting with adjustments according to enrolment trends simply. Yet the Jordanian government recently completed an exercise to assess the financial requirements of responding to the influx of Syrian refugees into the country, including for the education sector. This exercise adopted a “projectized� approach that arguably lends itself well to addressing the education service needs of lagging regions in Jordan. In summary, Jordan’s commitment to educating its youth is clear – both in terms of financial resources as well as in terms of the recent education sector reforms put in place. This report puts forth that policy makers have achieved what can be achieved by focusing on the quantity and quality of structural determinants of education services, such as infrastructure and even teachers. It is now time to focus more on the next generation of reforms related to accountability and incentives, on the one hand, and zeroing in on specific segments of the population and country to address their specific needs, on the other. 8 I. Introduction 1. This education sector public expenditure review (PER) aims to respond to policy questions on the mind of the Jordanian Ministry of Education (MOE) using data readily available to the Ministry as well as common analytical techniques. Despite impressive achievements in terms of access to education, policymakers in Jordan are keenly aware of the relatively low student learning outcomes in the country. In their efforts to improve the quality of education in the Jordanian education system, policymakers are quite focused on basic attributes of the system, such as school size, class size, and student-teacher ratios. On the one hand, policymakers are concerned with the number of small schools, and on the other, with the issue of overcrowded classrooms – given the potential negative impact of these on student learning outcomes. In addition, policymakers are intent on the quality of teaching and incentivizing teachers through better pay. The Syrian crisis with the associated influx of Syrian refugee children into Jordanian schools has made these issues all the more pressing for Jordanian policymakers. 2. This PER uses available MOE Education Management Information System (EMIS) and Ministry of Finance (MOF) budget data, in addition to other readily available sources of data, to shed light on the performance of the Jordanian education system and address policymakers’ concerns. The reliance on the MOE EMIS in particular is driven by the desire to demonstrate the value of the data it contains, on the one hand, but also point to additional data needs if MOE is to strengthen its evidence-based decision making. Thus, the EMIS currently includes data by school on numbers of teachers and students; physical attributes such as land area, buildings, and classrooms; in addition to the school’s geographic location, type, and supervising authority. The present EMIS therefore allows for relatively detailed analysis of the Jordanian schooling system, as will be shown in this report. However, the EMIS also entails shortcomings, as it does not allow for calculation of repetition and drop-out rates, for example, nor does it include any school financial data. 3. The present report begins by describing the achievements as well as shortcomings of Jordan’s education system (Section II); Section III describes the current education learning environment, including some detail on the basic system attributes of school size, class size, and student-teacher ratio; Section IV then turns to public education spending and its allocation across capital and recurrent spending, as well as the costs associated with the Syrian crisis; Section V hones in on teacher compensation in particular, given its large share in spending; and Section VI provides resulting recommendations and international good practice for policymakers’ consideration. 9 II. Overview of Education Outcomes Jordan has made impressive gains over the past decades in education enrolment and attainment. Today’s enrolment rates are on par with comparator countries (with the exception of enrolment rates in pre- primary education). Enrolment rates do show relatively large variation across governorates, however, as well as gender inequity in favor of females. Enrolments are increasing and will continue to do so for the foreseeable future based on population projections, but enrolments in the private sector are increasing at a much faster rate than in the public sector. This in turn is likely a reflection of the poorer quality of education in public schools. In general, Jordan’s accomplishments in terms of improving access to education are not matched by improvements in the quality of education: student assessment data show below average learning outcomes. ____________________________________________________________________________________ 4. Jordan has achieved impressive gains in access to education over the past decades. Enrolment in primary education, already universal in the 1970s, has remained universal1, and enrolment in all other levels of education has increased consistently over the decades (see Table 1). The increases in gross enrolment ratio (GER) amount to roughly 30 percentage points between 1971 and 2012 for each of pre- primary, secondary, and tertiary education. This translates to roughly a tripling of the pre-primary and tertiary GERs over the decades, and a 150 percent increase in the secondary GER. Another way to depict the impressive gains in education in Jordan since 1970 is to track the number of years of schooling completed by adults above the age of 15. Thus, according to Barro and Lee (2013), the number of years of schooling in Jordan went up from 3.4 years in 1970 to 9.1 years in 2010. In addition, Iqbal and Kiendrebeogo (2014) track the trends in education attainment between 1970 and 2010 for Jordan and other countries with similar years of schooling to Jordan in 1970, i.e. China, Ghana, Lesotho, Mexico, and Mongolia. They show that Jordan did better in terms of increasing education attainment between 1970 and 2010 than all its 1970 comparators (see Figure 1). Table 1: Jordan’s gross enrolment ratios for all levels of education, 1970-2012 Gross enrolment ratio 1971-80 1981-90 1991-2000 2001-12 Pre-primary 10.4 15.5 25.9 34.0 Primary 109.9 103.4 101.0 102.9 Secondary 61.6 78.4 82.7 91.1 Tertiary 7.4 20.0 21.8 40.2 Source: UNESCO Institute of Statistics 1 The primary gross enrolment ratio is well above 100 percent in the 1970s and today closer to 100 percent, reflecting less under-age and over-age enrolment in primary school years. 10 Figure 1: Average years of school, Jordan and comparators (1970-2010) Source: Iqbal and Kiendrebeogo (2014), page 15 Note: The comparator countries all had average years of schooling within 5 percent of Jordan’s average in 1970. 5. Today, the Jordanian educational ladder consists of 2 years of pre-school education (KG1 and KG2) for 4-5 year-olds, followed by 10 years of compulsory basic education (grades 1-10) for 6-15 year-olds. Following the compulsory cycle, students can choose to enter either comprehensive secondary education or applied secondary education (the latter provided in training centers and as apprenticeships). Comprehensive secondary (hereafter referred to simply as secondary) education lasts 2 years and includes both a vocational and an academic track that students sort into. The Ministry of Education (MOE) regards academic and vocational secondary education as having a common educational base with additional specialized coursework, whereas applied secondary education is geared towards training and preparation for the workplace. Tertiary education therefore follows comprehensive secondary education only. As Table 2 shows, there is some variation in how Jordan compares across the different levels of education in terms of enrolment rates. Thus, while primary, secondary, and tertiary GERs are roughly equivalent to the rates in comparator countries2, pre-primary enrolment rates in Jordan are on the low side. Table 2: Gross enrolment ratios for Jordan and comparators, 2012-13 Gross enrolment ratio Jordan Tunisia Croatia Thailand Pre-primary 34.2 40.0 63.31 118.52 Primary 98.4 110.3 96.82 92.85 Secondary 87.8 90.6 98.43 86.98 Tertiary 46.6 35.2 61.63 51.23 Source: UNESCO Institute of Statistics Note: “Secondary� here includes both comprehensive and applied secondary education 2 These comparators are a subset of those chosen for the Jordan Systematic Country Diagnostic (SCD). The criteria employed in choosing comparator countries for the SCD were: small, upper-middle/high income, ambitions for manufacturing, dependent on current or capital inflows, and energy importers. 11 Box 1. Early Childhood Education in Jordan The Middle East and North Africa (MENA) region stands out as a region that until now has not invested sufficiently in early childhood education, with gross enrolment in pre-primary education at 27 percent in 20113. The only region with a lower pre-primary enrolment rate is sub-Saharan Africa, at 18 percent. In addition to having low overall enrolment in pre-primary education, the MENA region has the lowest public provision of pre- primary, with only 29 percent of pre-primary enrolment in public programs and 71 percent in private preschools and nurseries. Thus, in Jordan, early childhood education was provided by the private sector until 1999, and the Government of Jordan only recently focused on including kindergarten classrooms in public schools. Today, investments in early childhood development have witnessed considerable growth worldwide due to national and international attention to the importance of the early years and their impact on future welfare outcomes. Though only 22 percent of 3-4 year-olds in Jordan attend early childhood education (KG 1), this share rises to around 60 percent for 5-6 year-olds (KG 2). In addition, a national campaign to universalize pre-primary education has been launched, so that a year of early childhood education (KG 2) will be added to the compulsory education cycle over a five-year period, and starting with underprivileged areas. 6. In terms of variation in GERs across governorates in Jordan, Figure 2 shows that enrolment in basic education is near universal across all governorates (average national GER of 99.4 percent). The kindergarten GER averages 37 percent nationwide, and there is a more than twenty percentage point gap between the top governorate Maan (GER of 52 percent) and the bottom governorate Mafraq (GER of 30 percent). The secondary GER averages 79 percent nationwide, and this time Maan ranks lowest with a GER of 73 percent that is almost ten percentage points lower than the secondary GER in Jarash. There are, in other words, no clear patterns in variation of GER across different levels of education amongst the governorates. But ten to twenty percentage point gaps demonstrate large geographic variation in GERs for a relatively small country such as Jordan. GERs show some variation by gender as well. As Figure 3 shows, GERs are slightly higher for girls than boys at the kindergarten and basic education levels, and by the time students reach secondary education, the gap has widened to an 8.7 percentage point gap in GERs in favor of females over males. Figure 2: Gross enrolment ratios in kindergarten, basic, and secondary education, by governorate (2010-11) 120.00 Gross enrolment ratio 100.00 80.00 60.00 Kindergarten 40.00 Basic 20.00 Secondary 0.00 Mafraq Aqaba Zarqa Madaba Irbid Karak Balqa Ajloun Tafila Jarash Ma'an Amman Total Governorate Source: EMIS, MoE. Governorates are ranked by increasing secondary GER. 3 This write-up draws extensively on El-Kogali and Krafft. 2015. Expanding Opportunities for the Next Generation: Early Childhood in the Middle East and North Africa . World Bank: Washington DC. 12 Figure 3: Gross enrolment ratios in kindergarten, basic, and secondary education, by gender (2013-14) Source: EMIS, MoE 7. Schools in Jordan are run either by the Ministry of Education; other governmental entities (such as the Ministry of Religious Endowments, Ministry of Higher Education, Ministry of Social Development, and Ministry of Defense); the private sector; or the United National Relief and Works Agency (UNRWA), which runs schools for Palestinian refugees. Schools are categorized as either Kindergarten, basic, or secondary schools – based on the highest grade available at the school. In addition, schools can enroll girls only, boys only, or be co-educational. Public schools are co-educational only through grade 4 (and in some communities, only through grade 2) after which girls are taught in separate schools predominantly by female teachers and boys by males teachers. 8. As Table 3 shows, Jordan had a total of 6,614 schools in academic year 2013-14, of which 3,694 (or 56 percent) were run by the Ministry of Education and 2,708 (or 41 percent) by the private sector. The share of the private sector in education in Jordan is, in other words, sizeable. Other governmental entities were only in charge of a total of 38 schools, and UNRWA administered 174 schools. All UNRWA schools but one are basic schools, while the majority of schools administered by other governmental entities are secondary schools. Almost all Kindergartens (1,542 out of a total of 1,544) are run by the private sector – though over 800 MOE basic and secondary schools have Kindergarten classes. Roughly two thirds of MOE schools are basic schools and one third secondary schools. Finally, close to half of MOE schools are co- educational, while the overwhelming majority (98 percent) of private schools are co-educational. 9. In academic year 2013-14, total enrolments across all education levels – kindergarten, basic, and secondary – amounted to 1.8 million students in all four authorities, of which 1.3 million were in MOE schools. Total enrolments grew over the 2001-14 period by 23 percent, reflecting roughly 23 percent enrolment growth in MOE schools, 66 percent growth in private schools, and downward trends in enrolment in UNRWA schools and schools run by other government entities (see Figure 4). The impressive growth in private sector enrolment is even clearer in Figure 5, which shows that the share of private schools in total enrolment is the only one that increased over 2001-14, translating to declines in the shares of the other authorities, including MOE (the latter declining from 70.6 to 68.5 percent of total enrolment). 13 Table 3: Jordan's schools, by authority, type, and gender (2013-14) Absolute numbers Gender Mi ni s try of Educa tion Pri va te School s Other Governmental UNRWA Gra nd Total KG Ba s i c Seconda ry Total KG Ba s i c Seconda ry Total KG Ba s i c Seconda ry Total KG Ba s i c Seconda ry Total KG Ba s i c Seconda ry Total Ma l e - 782 570 1,352 - 8 38 46 - 1 33 34 - 85 1 86 - 876 642 1,518 Fema l e - 246 302 548 - 4 16 20 - 1 2 3 - 71 - 71 - 322 320 642 Co-educa tiona l 2 1,366 426 1,794 1,542 913 187 2,642 - 1 - 1 - 17 - 17 1,544 2,297 613 4,454 Total 2 2,394 1,298 3,694 1,542 925 241 2,708 - 3 35 38 - 173 1 174 1,544 3,495 1,575 6,614 Percent share of type of school, by authority Gender Mi ni s try of Educa tion Pri va te School s Other Governmental UNRWA Gra nd Total KG Ba s i c Seconda ry Total KG Ba s i c Seconda ry Total KG Ba s i c Seconda ry Total KG Ba s i c Seconda ry Total KG Ba s i c Seconda ry Total Ma l e 0.0 57.8 42.2 100 0.0 17.4 82.6 100 0.0 2.9 97.1 100 0.0 98.8 1.2 100 - 57.7 42.3 100 Fema l e 0.0 44.9 55.1 100 0.0 20.0 80.0 100 0.0 33.3 66.7 100 0.0 100.0 0.0 100 - 50.2 49.8 100 Co-educa tiona l 0.1 76.1 23.7 100 58.4 34.6 7.1 100 0.0 100.0 0.0 100 0.0 100.0 0.0 100 34.7 51.6 13.8 100 Total 0.1 64.8 35.1 100 56.9 34.2 8.9 100 0.0 7.9 92.1 100 0.0 99.4 0.6 100 23.3 52.8 23.8 100 Percent share of gender of school, by authority Gender Mi ni s try of Educa tion Pri va te School s Other Governmental UNRWA Gra nd Total KG Ba s i c Seconda ry Total KG Ba s i c Seconda ry Total KG Ba s i c Seconda ry Total KG Ba s i c Seconda ry Total KG Ba s i c Seconda ry Total Ma l e 0.0 32.7 43.9 36.6 0.0 0.9 15.8 1.7 - 33.3 94.3 89.5 - 49.1 100.0 49.4 0.0 25.1 40.8 23.0 Fema l e 0.0 10.3 23.3 14.8 0.0 0.4 6.6 0.7 - 33.3 5.7 7.9 - 41.0 0.0 40.8 0.0 9.2 20.3 9.7 Co-educa tiona l 100.0 57.1 32.8 48.6 100.0 98.7 77.6 97.6 - 33.3 0.0 2.6 - 9.8 0.0 9.8 100.0 65.7 38.9 67.3 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 - 100.0 100.0 100.0 - 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source: Ministry of Education (2015) page 10 14 Figure 4: Total enrolment in kindergarten, basic, and secondary, by authority (2001-14) Source: Education Management Information System (EMIS), Ministry of Education Figure 5: Share of total enrolment in kindergarten, basic, and secondary education, by authority (2001-14) Source: Education Management Information System (EMIS), Ministry of Education 10. A closer look at the breakdown of schools by governorate shows that there is great variation across governorates in terms of MOE share of schools in the governorate. The two main categories are MOE and private schools, as UNRWA and other governmental schools constitute only 2.7 and 2.5 percent of schools nation-wide. Thus, as Table 4 shows, the share of MOE-administered schools in total number of schools is highest in Mafraq at 87.9 percent, and lowest in Amman at 36.7 percent. Amman has by far the highest share of private schools amongst all governorates, amounting to 58.8 percent of all schools. 15 Table 4: Schools by governorate and authority, 2013-14 (percent share of total in governorate) Governorate Ministry of Private UNRWA Other Education Education Governmental Total Mafraq 87.9 11.2 0.4 0.6 100.0 Tafilah 83.1 8.1 0.0 8.8 100.0 Maan 79.2 5.9 0.0 14.8 100.0 Karak 78.4 17.8 0.0 3.8 100.0 Jerash 72.7 23.1 3.3 0.8 100.0 Madaba 70.5 24.7 2.1 2.6 100.0 Ajloun 63.4 34.0 0.0 2.6 100.0 Balqa 55.7 35.6 4.6 4.1 100.0 Irbid 55.1 39.7 2.7 2.5 100.0 Aqaba 54.5 35.5 0.0 9.9 100.0 Zarqa 50.7 44.3 4.0 1.0 100.0 Amman 36.7 58.8 3.9 0.6 100.0 Sub-total 55.8 39.0 2.7 2.5 100.0 Source: EMIS, MOE Governorates are ranked by decreasing share of MOE schools in total. 11. As mentioned above, total enrolment grew by 23 percent over the 2001-14 period, and Figure 6 shows population projections for Jordan through 2050. Narrowing in on the age ranges most relevant for the education system, i.e. age 5-9, 10-14, and 15-19 years, Figure 6 shows that the population is projected to grow significantly through 2020, 2025, and 2030 respectively. Thus, at their peak, 5-9 year-olds are projected to number 7.5 percent more than in 2015; 10-14 year-olds 29.1 percent more than in 2015; and 15-19 year-olds 32.2 percent more. In other words, the education system in Jordan will be expanding substantially in the foreseeable future. Figure 6: Jordan population projections, 2010-2050 (1,000s) 1000 950 900 Population (1,000s) 850 0-4 years 800 750 5-9 years 700 10-14 years 650 600 15-19 years 550 20-24 years 500 2010 2015 2020 2025 2030 2035 2040 2045 2050 Year Source: Population Division of the Department of Economic and Social Affairs of the United Nations Secretariat, World Population Prospects: The 2012 Revision 16 12. It is not clear, however, to what degree the enrolment growth will take place in the public versus the private education sector. As described earlier, though enrolment has grown over the past decade, it has grown at a faster rate in the private than in the public sector, and the share of the private sector in total enrolment is increasing. This is probably the result of households, particularly in Amman, resorting to the private sector as a more likely provider of high quality education. In general, the quality of education in Jordan is relatively low: the country has participated in several international student assessments over the years, including the Trends in International Mathematics and Science Study (TIMSS) and the Program for International Student Assessment (PISA), which aim to measure actual learning achieved and competencies acquired in schools. In TIMSS, Jordan scores below the international average, although better than many of its comparators in science (though not in mathematics). Worryingly, Jordan’s scores in TIMSS declined between 2007 and 2011 (see Figure 7). In PISA, Jordan scores below the OECD average, below most of its comparators, and exhibits a decline between 2009 and 2012 (Figure 7). However, more detailed analysis of the 2009 PISA results for Jordan, for example, shows that there is a statistically significant difference (23 points) in reading performance between private and public school students in favor of the private school students4. Figure 7: Jordan's performance in international student assessments Source: World Bank EdStats Database 13. Jordan also exhibits variation in its 2011 TIMSS scores by gender and geographic location. As Figure 8 shows, girls outperform boys in both math and science. In addition, urban areas stand apart in terms of student learning outcomes in both math and science, with the next ranked grouping being suburban areas and medium size cities, followed after a gap by the final grouping that includes small towns and rural areas. The issue of geographic variation in terms of learning outcomes is one that this report will return to in a later section. 4 OECD. 2010. PISA 2009 Results: What Makes a School Successful? Resources, Policies, and Practices (Volume IV), Table IV.3.9. 17 Figure 8: Variation in 2011 TIMSS scores, by gender and geographic location 500 471 500 466 449 449 420 428 422 428 432 408 409 392 385 371 400 400 300 300 200 200 100 100 0 0 Urban Suburban Medium size Small town Remote rural Math Science city Male Female Math Science 18 III. The Public Education Learning Environment School size, class size, and student-teacher ratio are key education system indicators that the Jordanian Ministry of Education is concerned with – particularly in the context of an influx of Syrian refugee students into the public education system. In fact, school size in Jordan is small on average, and class size and student-teacher ratio are today comparable to OECD averages – and all three indicators are on a decreasing trend. Very small schools are mainly a rural phenomenon but also result from the need to resort to renting buildings due to lack of land for construction of schools. Overcrowded classrooms are very much an urban phenomenon limited to a few governorates. However, tawjihi pass rates are not associated with school size, class size, and student-teacher ratio, while being negatively correlated with location in certain governorates, male schools, and rented schools. __________________________________________________________________________________ 14. Cognizant of the relatively low learning outcomes achieved by the education system, education policymakers in Jordan are concerned with a set of issues revolving around basic attributes of the system, i.e. school size, class size, and student-teacher ratios. On the one hand, policymakers are concerned with the number of small schools, partly as these tend to be costly to operate, but also in the belief that they may potentially have a negative impact on students’ learning outcomes. On the other hand, policymakers are concerned about overcrowded classrooms – again, in the belief that they result in poor learning outcomes for students. The Syria crisis with the associated influx of Syrian refugee children into Jordanian schools has made the issue of overcrowded schools and classrooms all the more pressing for Jordanian policymakers. Finally, Jordanian policy makers recognize that teachers are the key drivers in achieving student learning and want to ensure that a sufficient number of teachers is available to students. 15. According to the Government’s Jordan Response Plan for the Syria Crisis 2016-2018, Jordan hosted about 630,000 Syrian refugees by the end of 2015, 83 percent of whom settled in host communities, particularly in the urban area of Amman and the northern governorates of Jordan, with the remainder in refugee camps. Syrian children, if attending school, are either in schools in the refugee camps, or integrated into Jordanian public and private schools. In most instances where additional Syrian students result in overcrowding in public schools, a second shift is introduced that is dedicated to the Syrian students. Data at the end of 2014 indicate a total of 129,058 Syrian refugee children in public schools in Jordan, of which 23,226 (18 percent) are in schools in refugee camps; 59,627 (46 percent) in public schools (regular shift); and 46,205 students (36 percent) in second shifts in public schools. The number of Syrian refugee children in public schools went up to 145,458 by the end of 2015, but the breakdown of 18 percent enrolled in camps and 82 percent in host communities remained. Therefore, the ratio of Syrian to Jordanian students in host community MOE schools (whether regular or second shift) is roughly 1:10. For students in refugee camp schools and in public school second shifts, additional teachers are hired. For students in regular shift public schools, an important concern is overcrowding and the impact this would have on the quality of education for all students in the class, both Jordanian and Syrian. (Financial implications of the Syrian refugee students will be discussed in the next section.) 16. It is therefore important to gain a better understanding of Jordan’s education system in terms of schools, classes, and teachers over recent years. Over the 2001-14 period, private schools show impressive growth in numbers of schools, classes, teachers, and students, and MOE shows sizeable growth as well. Enrolments declined in non-MOE governmental schools as well as in UNRWA, and so schools under these authorities have declined in number over this period as well, as have numbers of classes – 19 though numbers of teachers did increase slightly (see Figure 9). Interestingly, the growth in the number of teachers shows the highest growth rate across all four authorities. Figure 9: Growth in numbers of schools, classes, teachers, and students, by authority (2001-14) Source: EMIS, MOE 17. Beyond changes in absolute numbers of schools, classes, teachers, and students, it is important to consider how the basic efficiency indicators of the system are faring, i.e. school size, class size, and student-teacher ratio (STR). As Table 5 shows, all three indicators decreased in size over the period 2001- 14 for all authorities – with the exception of private school size, which increased. Table 6 provides the actual school size, class size, and STR over the years. It shows that UNRWA schools and schools run by other governmental entities are the largest in size in Joran, followed by MOE schools, and finally private schools. Thus, even with the increases in school size in the private education sector, average school size in 2014 remains relatively small at 167 students. Table 6 further shows that class size is largest in UNRWA schools, followed by MOE and other governmental schools, with the smallest class sizes observed in private schools. Across all authorities, class size has declined over the 2001-14 period. Finally, STR is largest in UNRWA schools, followed by MOE and private schools, and last come other governmental schools. Student-teacher ratios have also declined over the 2001-14 period for all authorities. In other words, for all education system efficiency indicators, the trend is towards smaller values – with the exception of private school size. Table 5: Percent growth rate in school size, class size, and student-teacher ratio, by authority (2001-14) Ministry of Private Schools Other UNRWA Overall Education Governmental School size -1.95 14.34 -14.64 -7.23 -3.40 Class size -7.34 -9.87 -5.12 -13.93 -10.47 Student-teacher ratio -14.76 -11.31 -34.54 -20.60 -15.97 Source: EMIS, MOE 20 Table 6: Trends in school size, class size, and student-teacher ratio, by authority (2001-14) Ministry of Private Schools Other UNRWA Overall Education Governmental School size 349 146 486 712 289 2001-2002 Class size 29 23 28 40 28 Student-teacher ratio 19 17 16 33 19 School size 351 146 485 704 289 2002-2003 Class size 29 23 27 39 28 Student-teacher ratio 19 17 16 32 19 School size 348 148 423 702 287 2003-2004 Class size 29 23 26 39 28 Student-teacher ratio 18 16 14 31 19 School size 343 150 405 732 284 2004-2005 Class size 28 23 26 38 28 Student-teacher ratio 18 16 14 31 18 School size 346 155 374 698 284 2005-2006 Class size 28 22 20 37 27 Student-teacher ratio 18 16 14 29 18 School size 342 163 392 714 285 2006-2007 Class size 28 23 22 38 27 Student-teacher ratio 18 16 13 30 18 School size 338 166 363 714 284 2007-2008 Class size 28 22 23 38 27 Student-teacher ratio 17 16 12 29 17 School size 342 170 353 708 287 2008-2009 Class size 28 23 24 37 27 Student-teacher ratio 17 16 12 28 17 School size 335 171 91 689 279 2009-2010 Class size 27 21 27 35 26 Student-teacher ratio 16 15 11 27 16 School size 333 166 160 682 276 2010-2011 Class size 27 21 45 35 26 Student-teacher ratio 16 15 18 26 16 School size 331 164 417 661 274 2011-2012 Class size 27 21 27 34 25 Student-teacher ratio 16 15 10 26 16 School size 331 163 406 652 272 2012-2013 Class size 26 21 26 34 25 Student-teacher ratio 16 15 11 25 16 School size 342 167 415 660 279 2013-2014 Class size 27 21 27 34 25 Student-teacher ratio 16 15 11 26 16 Source: EMIS, MOE 18. In addition, variation in school size, class size, and student-teacher ratio is very much on a regional basis in Jordan, extending beyond simply the urban-rural divide, for example. As Table 7 shows, the different regions of Jordan – Center, North, and South – vary from each other, whether one compares the urban or rural areas of each region. Thus, for all indicators, and whether urban or rural, the value is consistently highest in the Center Region (which includes the capital Amman), followed by the North Region, and finally the South Region. 21 Table 7: School size, class size, and student-teacher ratio, by region and geography (2014) Region School size Class size Student-teacher ratio Urban Rural Urban Rural Urban Rural Center 595 217 31 19 20 12 North 417 185 27 18 16 11 South 287 157 23 16 12 9 Source: EMIS, MOE Notes: The Center Region includes the governorates of Amman, Balqa, Zarqa, and Madaba; the North Region includes Irbid, Mafraq, Jarash, and Ajloun; and the South Region includes Karak, Tafila, Maan, and Aqaba. 19. Given the above-described downward trend in school size observed in Jordan (with the exception of private schools) the question arises as to the “ideal� school size in terms of achieving better learning outcomes at a lower cost. School size is of course one of many factors that can affect school quality, and debates on school size mostly revolve around the implications of size for school quality and fiscal efficiency. In terms of students’ academic outcomes, school size has an impact on the extent to which schools are able to group students into single-grade groups or into homogenous groups according to their ability or age. One of the comparative disadvantages of larger schools mentioned in the literature is that they may suffer from management difficulties and problems with discipline (Ares Abalde 2014). This is related to another disadvantage found in larger schools, which is that they tend to dilute the possibilities for interpersonal relationships between teachers and students outside the classroom, which could in turn impact learning negatively. On the other hand, larger schools allow for teacher specialization and reduce the administrative burden on teachers as larger schools can invest in specific administrative personnel. In addition, the issue of professional development is a concern for school staff in small remote and rural schools. Finally, small schools may be better at closing the achievement gap between students from different socioeconomic status. 20. One of the most common arguments in favor of large school sizes is that larger schools are more cost-efficient since schools face economies of scale. This applies both to capital spending as well as operating costs (e.g. larger schools can benefit from bulk buying). The results of research on school size and efficiency can be divided into studies that find that schools do indeed face economies of scale (and so larger schools are more efficient) and those that find a U-shaped relationship between size and costs (indicating that while smaller schools initially face economies of scale when increasing school size, these turn into diseconomies beyond a critical number of students). In addition, the definition of what constitutes a small versus a large school is often vague, and varies from one study to the next. Definitions of large schools, for instance, range from schools with 800 students or more to schools with more than 2,000 or even 3,200 students. In addition, optimal school size may vary between primary and secondary education: Leithwood and Jantzi (2009) find that optimal school sizes at elementary and secondary levels are 500 and 1,000, respectively. They also suggest that optimal school sizes are smaller when schools have a large proportion of disadvantaged students (300 for elementary and 600 for secondary schools). 21. Comparing the average MOE school size of 342 in Jordan to either the optimal size of 500 or 1,000 for elementary and secondary schools, respectively, suggests that schools in Jordan are small on average. The MOE categorizes schools as basic or secondary according to the highest grade in the school. In other words, any school that has tenth grade or lower as the highest grade in the school is categorized as a basic 22 school, and any school that includes either eleventh or twelfth grade is a secondary school5. In practice, only about 20 percent of MOE basic schools have tenth grade as the highest grade, while the overwhelming majority of secondary schools have twelfth grade as the highest grade. At the same time, however, only about 9 percent of MOE secondary schools start at the eleventh grade, with as many as 26 percent extending all the way through to Kindergarten, so that it is really not possible to speak of basic and secondary schools as distinct categories in Jordan. Beyond the relatively low average school size, as Figure 10 shows for both basic and secondary schools combined, the distribution of schools is skewed towards smaller school size as well. This is less stark for MOE urban schools and more so for MOE rural schools (as would be expected) as well as private urban schools. Figure 10: Distribution of school size, Ministry of Education and private schools, by geographic location (2014) Min of Education, Urban Min of Education, Rural 400 300 200 100 Frequency 0 Private Education, Urban Private Education, Rural 400 300 200 100 0 0 500 1000 1500 2000 0 500 1000 1500 2000 School size Source: EMIS, MOE 22. Smaller classes are often seen as beneficial because they allow teachers to focus more on the needs of individual students and reduce the amount of class time needed to deal with disruptions. Yet, while there is some evidence that smaller classes may benefit specific groups of students, such as those from disadvantaged backgrounds6, overall, evidence of the effect of differences in class size on student performance is weak. In addition, according to recent findings from the 2013 OECD Teaching and Learning International Survey (TALIS), smaller classes are not necessarily related to greater job satisfaction for teachers. It is not so much the number of students but rather the type of students in a teacher’s class that has the strongest association with teachers’ self-efficacy and job satisfaction. Teachers reported being most satisfied with their job when they have no students with behavioral problems in their classroom and least satisfied with their job when the proportion of students with behavioral problems in their classroom reaches more than 30 percent. A similar decrease in job satisfaction is not seen when 5 The MOE EMIS further classifies secondary schools into academic (84 percent), academic and technical (14 percent), and technical schools (2 percent). Given the low share of purely technical secondary schools, this discussion encompasses all types of secondary schools combined. 6 OECD. 2014. Education at a Glance 2014: OECD Indicators. 23 classes are larger. However, there is also evidence that suggests a positive relationship between smaller classes and more innovative teaching practices7. 23. The ratio of students to teaching staff indicates how resources for education are allocated. Smaller student-teacher ratios often have to be weighed against higher salaries for teachers, investing in their professional development, greater investment in teaching technology, or more widespread use of assistant teachers and other paraprofessionals whose salaries are often considerably lower than those of qualified teachers. As larger numbers of children with special needs are integrated into mainstream classes, the increased use of specialized personnel and support services may limit the resources available for reducing student-teacher ratios. 24. Turning to class size and student-teacher ratio in Jordan, the average MOE class size of 27 is only slightly above class size in Jordanian private schools (21) as well as the OECD average for primary public schools (21) and lower secondary schools (24)8. The average MOE student-teacher ratio of 16, on the other hand, is roughly the same as the student-teacher ratio in Jordanian private schools (15) as well as the OECD average for primary education (15) and secondary education (13)9. Beyond the average, the distribution of class size and student-teacher ratio in MOE urban basic and secondary schools (see Figures 11 and 12) shows a relatively wide spread (compared to urban private schools, for example). Rural MOE class size and student-teacher ratio are similarly varied, and the difference between urban and rural indicators is that the distribution for urban indicators is more skewed towards the larger class size and student-teacher ratio, while it is more skewed towards the smaller values in rural areas (as would be expected). Figure 11: Distribution of class size, Ministry of Education and private schools, by geographic location (2014) Min of Education, Urban Min of Education, Rural 200 150 100 50 Frequency 0 Private Education, Urban Private Education, Rural 200 150 100 50 0 0 10 20 30 40 50 60 70 80 0 10 20 30 40 50 60 70 80 Class size Source: EMIS, MOE 7 OECD. 2014. Education at a Glance 2014: OECD Indicators. 8 OECD. 2014. Education at a Glance 2014: OECD Indicators, Table D2.1. “Average class size, by type of institution and level of education (2012)�, page 450. 9 OECD. 2014. Education at a Glance 2014: OECD Indicators, Table D2.2. “Ration of students to teaching staff in educational institutions (2012)�, page 451. 24 Figure 12: Distribution of student-teacher ratio, Ministry of Education and private schools, by geographic location (2014) Min of Education, Urban Min of Education, Rural 200 150 100 50 Frequency 0 Private Education, Urban Private Education, Rural 200 150 100 50 0 0 5 10 15 20 25 30 35 40 45 50 0 5 10 15 20 25 30 35 40 45 50 Student-teacher ratio Source: EMIS, MOE 25. In other words, in general, the emerging picture for the Jordanian education system’s basic attributes is not one of large schools and overcrowded classrooms. On the contrary, based on both averages and distributions of school size, schools in Jordan are relatively small, whether in urban or rural areas. Regarding class size, 27 students on average is only somewhat larger than the OECD average, and looking at the distributions in Figure 10, it is clear that very small class sizes are more common than very large class sizes, especially in rural areas. Similarly, average STR in Jordan is on par with the OECD average, and the distribution shows a higher prevalence of very small STR than very large STR. 26. It is therefore important to examine at a more granular level the occurrence and attributes of the very small schools and overcrowded classrooms that the MOE policy makers are concerned about. Table 8 provides a breakdown of MOE schools with fewer than 100 students, a total of 774 schools in 2014. As the table shows, by far the largest number of these small schools is in Mafraq in the north, followed at a distance by Amman, Irbid, and Maan. In addition, the overwhelming majority are in rural areas – as would be expected. What is surprising is the disproportionately small number of female small schools, even accounting for the fact that there are fewer female than male or mixed schools in the country. This might indicate that schools for girls are not opened if there are not sufficiently large numbers of girls – and that this does not hold for boys. Most importantly, however, is the fact that roughly equal numbers of small schools are owned by MOE as they are rented. In other words, the concern for policy makers regarding small schools may have more to do with the fact that they result from having to resort to rental of inappropriate small buildings in the absence of land for school construction – for example in densely populated areas such as Amman. 25 Table 8: Characteristics of schools with fewer than 100 students, 2014 Total Geography School gender School ownership Urban Rural Female Male Mixed Owned Rented Ajloun 26 2 24 0 4 22 5 20 Amman 90 8 82 4 29 57 53 37 Aqaba 17 1 16 3 2 12 15 2 Balqa 41 9 32 4 14 23 25 15 Irbid 86 26 60 6 18 62 21 65 Jerash 52 6 46 5 13 34 26 23 Karak 79 17 62 3 25 51 39 39 Maan 81 23 58 8 26 47 45 35 Madaba 50 4 46 1 12 37 32 17 Mafraq 178 6 172 7 51 120 80 94 Tafilah 36 8 28 0 8 28 17 19 Zarqa 38 7 31 0 5 33 18 17 Total 774 117 657 41 207 526 376 383 Source: EMIS, MOE 27. In terms of understanding better where overcrowded classrooms are found and how they might be addressed, Table 9 shows that overcrowded classrooms, i.e. class size of 40 or above, are found predominantly in the urban areas of Amman and Zarqa. Contrary to small schools, female schools are over-represented in schools with overcrowded classrooms. Schools with overcrowded classrooms are also overwhelmingly amongst the schools owned by the MOE as opposed to those it rents. Finally, schools with overcrowded classrooms are predominantly schools that run only one shift. Given these last two observations, it would seem that one trade-off MOE needs to consider is the trade-off between overcrowded single shift schools versus putting in place a second shift. Table 9: Characteristics of schools with average class size greater than 40 students, 2014 Total Geography School gender School ownership Single shift Urban Rural Female Male Mixed Owned Rented schools Ajloun 2 2 0 0 1 1 0 2 2 Amman 76 72 4 27 31 18 75 1 58 Aqaba 4 4 0 0 2 2 4 0 3 Balqa 13 7 6 3 6 4 13 0 13 Irbid 20 18 2 7 10 3 17 2 18 Jerash 1 0 1 0 0 1 0 1 1 Karak 3 2 1 0 2 1 2 1 3 Maan 3 2 1 1 1 1 1 2 3 Madaba 1 1 0 0 1 0 1 0 1 Mafraq 6 2 4 1 1 4 3 3 6 Tafilah 0 0 0 0 0 0 0 0 0 Zarqa 51 46 5 14 21 16 50 1 39 Total 180 156 24 53 76 51 166 13 147 Source: EMIS, MOE 26 28. Returning to the issue of Syrian students causing or adding to the overcrowding of schools and classes, it is clear that this phenomenon is localized. To begin with, the highest numbers of Syrian refugee children enrolled in Jordanian public schools are in the northern governorates of Irbid, Mafraq, and Jarash, as well as the central governorates of Zarqa and Amman. As Table 9 shows, both Amman and Zarqa have relatively higher numbers of schools with overcrowded classrooms, so that additional Syrian influx can pose a serious challenge. By contrast, Mafraq is characterized by a large number of small schools and indeed anecdotal reports do not cite overcrowded classes resulting from the Syrian influx as a cause for concern in that governorate. 29. The question remains how school size, class size, and student-teacher ratios are impacting learning outcomes for students in Jordan. Using MOE data on the pass rate of the final secondary exam, the tawjihi, Table 10 provides average pass rates by different school characteristics for the 2013 tawjihi. The overall pass rate for Jordan was 51.4 percent, with 1,289 secondary schools participating in the exam. Of those, 1,082 were MOE schools and achieved a pass rate of 49.4 percent, while there were 173 private schools with a pass rate of 65.1 percent – another indication of the quality of education differential between public and private schools. Focusing on the MOE schools, the highest pass rates were in Amman and Tafilah (53.4 percent) – though the number of schools varies widely between those two governorates (251 in Amman versus only 28 in Tafilah). Higher pass rates were observed in urban, as opposed to rural, areas; schools owned by MOE, as opposed to rented schools (of which there were only 8, however); female or co-educational schools, as opposed to male schools; and combined academic and technical schools, as opposed to purely academic or technical. 30. The above average pass rates, while indicative, do not control for the different factors at play and their individual association with pass rates. They also do not allow for an assessment of the potential correlation with school size, class size, or student-ratio. Table 11 therefore provides the results of regressing the tawjihi pass rates on different school characteristics. The results show that, controlling for the other factors taken into account, tawjihi pass rates are not associated with school size, class size, and student-teacher ratio. The largest negative association with tawjihi pass rates is found with residing in Aqaba, Jarash, Maan, Madaba, or Zarqa (by comparison with Amman); attending a rented school, as opposed to a school owned by MOE; and attending a male school, as opposed to a female school. In other words, in Jordan, the school attributes of school size, class size, and student-teacher ratio are not predictive of one measure of learning outcomes – tawjihi pass rates – while the school’s governorate, whether it is rented or not, and whether it is a male school are good predictors. Other school attributes either have a negligible or no association with pass rates: residing in a rural versus urban location; attending a co-educational school, compared to a female school; and attending a combined academic & technical, or a pure technical, as compared to a pure academic school. 31. While not being associated with tawjihi pass rates, school size, class size and student-teacher ratio do, however, impact spending on public education in Jordan. As mentioned earlier, larger school and class sizes may lead to economies of scale both in terms of capital as well as operating costs. Larger student- teacher ratios, in turn, have a direct impact on the number of teachers in the system. As the next sections will lay out, teacher compensation constitutes by far the largest share of education spending so that policies that affect teachers will have an important impact on spending. 27 Table 10: Tawjihi results, by school characteristic, 2013 Tawjihi No. of pass rate schools Overall 51.4 1289 Ministry of Education 49.4 1082 Private 65.1 173 Ministry of Education Governorate Ajloun 48.5 41 Amman 53.4 251 Aqaba 40.4 29 Balqa 50.7 89 Irbid 51.4 195 Jerash 43.3 48 Karak 51.4 82 Maan 40.5 49 Madaba 39.6 36 Mafraq 50.8 141 Tafilah 53.4 28 Zarqa 43.4 93 Geographic location Urban 51.6 528 Rural 47.4 554 School ownership Owned 49.5 1063 Owned & rented 49.7 11 Rented 41.3 8 School gender Female 58.3 283 Male 40.1 474 Co-educational 55.3 325 Type of secondary school Academic 48.6 891 Academic & techical 54.2 165 Technical 47.0 26 Source: EMIS and MOE 28 Table 11: Regression results for tawjihi pass rates and school characteristics, 2014 Standard Coefficient t-statistic p-value Error Governorate Ajloun -0.06 0.04 -1.62 0.105 Aqaba -0.19 0.04 -4.66 0 Balqa -0.02 0.03 -0.87 0.387 Irbid -0.05 0.02 -2.6 0.009 Jerash -0.12 0.03 -3.65 0 Karak -0.07 0.03 -2.25 0.024 Maan -0.21 0.04 -5.95 0 Madaba -0.15 0.04 -3.88 0 Mafraq -0.05 0.02 -1.85 0.065 Tafilah -0.06 0.04 -1.31 0.19 Zarqa -0.11 0.03 -4.27 0 Geography Village -0.04 0.02 -2.45 0.014 School ownership Owned & rented 0.01 0.06 0.12 0.902 Rented -0.17 0.08 -2.26 0.024 School gender male -0.18 0.02 -10.9 0 mixed 0.00 0.02 -0.26 0.792 Type of education Academic and technical -0.06 0.02 -3.07 0.002 Technical -0.06 0.05 -1.27 0.204 Total number of students 0.00 0.00 4.59 0 STR -0.03 0.00 -8 0 Class size 0.01 0.00 3.23 0.001 _cons 0.80 0.04 20.68 0 29 IV. Overview of Public Education Spending Jordan spends comparable shares of GDP and total public spending on education as the OECD averages, and higher than expected shares when taking its GDP per capita into account. Capital spending in the education sector is classified both under the Ministry of Education as well as under the Ministry of Planning and International Cooperation, with the bulk being treasury-financed spending (as opposed to loans) under the Ministry of Education. The share of capital spending in total education spending is relatively low. This contributes to the fact that recurrent spending on education in Jordan is substantial and equivalent to the country’s budgetary commitment (on a recurrent basis) to defense, with other sectors lagging significantly behind. Within the education sector, basic education captures the majority of capital spending and exhibits relatively high recurrent unit costs. Unit costs were one way that the Ministry of Education and donors approached the question of estimating the impact of Syrian refugee students on the Jordanian education system. However, the latest thinking adopts a more “projectized� approach that spans a range of interventions, with the largest ticket items being additional infrastructure investments and hiring of teachers. __________________________________________________________________________________ 32. Public education expenditures in Jordan fall primarily under the Jordanian Ministry of Education (MoE), which classifies its spending along eight different programs: 1) administration and support services; 2) kindergarten education; 3) basic education; 4) secondary education; 5) vocational education; 6) special education; 7) social, sport, and educational activities; and 8) eradication of illiteracy and adult education. As is clear from the above program listing, the MOE covers all levels of education in the country with the exception of post-secondary education, for which there is a separate Ministry of Higher Education and Scientific Research, and which is beyond the scope of the present analysis. 33. As Table 12 shows, MOE expenditures amounted to roughly 840 million Jordanian dinars (JD) in 2013, which translates to 3.5 percent of gross domestic product (GDP) and 9.7 percent of total public expenditures. Table 12 further shows that the share of public education spending in GDP has been relatively stable since 2008. On the other hand, the share of total public spending in GDP displays a downward trend, yet education spending seems to have been protected, since its share in total spending increases over the period. By way of comparison, the corresponding OECD average 2011 shares of education spending are 4.2 percent of GDP and 9.5 percent of total public expenditure10, i.e. roughly comparable (OECD 2014)11. 10 OECD 2014. Education at a Glance 2014. Indicator B4, Table B4.1. “Total public expenditure on education (2011)�. 11 These figures do not take into account spending on the Vocational Training Corporation (VTC), a semi- governmental organization almost fully financed by the Government and supervised by a Board chaired by the Ministry of Labor. The emphasis of the VTC is on planning, research, and development in technical and vocational education, and it offers training programs to students, apprentices, instructors, as well as teachers. As such, it is outside the purview of the present report. In 2013, its recurrent spending amounted to 8.9 million JD, roughly equivalent to 38 percent of total recurrent spending on vocational education under the MOE. 30 Table 12: Ministry of Education expenditures, as share of GDP and total education expenditures (2008-13) (JD million) 2008 2009 2010 2011 2012 2013 Nominal GDP 15,593 16,912 18,762 20,477 21,966 23,852 Total public expenditures 6,741 7,487 7,260 8,366 8,473 8,718 -- as % of GDP 43.2 44.3 38.7 40.9 38.6 36.6 Ministry of Education expenditures 525 559 591 689 759 842 -- as % of GDP 3.4 3.3 3.1 3.4 3.5 3.5 -- as % of total public expenditures 7.8 7.5 8.1 8.2 9.0 9.7 Sources: Ministry of Finance. July 2014. General Government Finance Bulletin, Studies and Economic Policies Directorate, Vol. 17, No. 6, pages 10, 46; and Ministry of Finance. June 2014. General Budget Final Accounts for Fiscal Year 2013, pages 503-506 34. There is some evidence that the amount of public resources spent on education depends to some degree on the level of economic development of the country. In other words, countries with different levels of GDP per capita do not spend the same shares of GDP on education. Instead, in richer countries, the share of GDP allocated to public primary education tends to be lower than in poorer countries, while the reverse is observed for secondary education12. Figure 13 provides a scatter plot of World Bank data on GDP per capita against UNESCO Institute of Statistics data on public spending on primary education as a share of GDP, as well as a fitted line representing the expected level of public spending as a function of GDP per capita. The line slopes downwards, indicating a smaller share of GDP for primary education the richer the country. Jordan lies above the fitted line, implying that the country spends more than expected for its level of GDP per capita on primary education. Figure 14 provides the parallel scatter plot and fitted line for secondary education, and again Jordan lies above the line, indicating higher than expected spending on secondary education for its per capita GDP. Figure 13: Correlation between GDP per capita and government spending on primary education as a share of GDP, 2014 Source: Heejin Kim, Janice and Quentin Wodon. 2015. Public Education Cost Benchmarking for Jordan, World Bank, Figure 1(a), page 2. 12 Heejin Kim, Janice and Quentin Wodon. 2015. Public Education Cost Benchmarking for Jordan , World Bank. 31 Figure 14: Correlation between GDP per capita and government spending on secondary education as a share of GDP, 2014 Source: Heejin Kim, Janice and Quentin Wodon. 2015. Public Education Cost Benchmarking for Jordan, World Bank, Figure 1(b), page 2. A. Capital Spending 35. In terms of capital spending, budget data distinguish first between capital spending financed by the treasury versus by loans, with treasury-financed spending being the overwhelming majority in 2013 (over 97 percent). The big spending entities in terms of capital spending out of the treasury are the Ministry of Finance and the Ministry of Public Works and Housing, with the Ministry of Education ranking only 8th at 3.9 percent of capital spending out of the treasury. In addition, as Table 13 shows, the budget indicates that the Ministry of Education has no capital spending out of loans. This, however, is due to the fact that Ministry of Education loan-financed capital spending falls under the Ministry of Planning and International Cooperation (MOPIC). Indeed, as Table 13 further shows, MOPIC is the conduit for the lion’s share of loan-financed capital spending (77 percent in 2013). 36. Table 13 in effect provides a distorted picture on capital spending for the education sector in Jordan, since the data under the MOE are not comprehensive13. For this reason, Table 14 provides the education capital spending, both from the treasury and loans, under both the MOE and MOPIC. Table 14 shows that, in 2013, over 20 percent of capital spending in the education sector was channeled through MOPIC, the bulk of which was in the form of loans (Education Reform for the Knowledge Economy II – ERFKE II). Thus, loan-financed capital spending amounted to 17.9 percent of total education sector capital spending. As it turns out, MOPIC financed construction of 25 schools partially out of the treasury in addition to loans. In other words, the divide between MOE and MOPIC in terms of capital spending in the education sector is not a clean divide of treasury versus loan-financed spending. In addition, total capital spending on education (including MOE and MOPIC) amounts to only 5.7 percent of total spending on 13 Another source of capital spending are USAID grants, which appear neither under treasury- nor loan-financed capital spending. Over the years, USAID support to the Jordanian education sector has been substantial (see Box 2) but remains off-budget. 32 education in 2013. By comparison, the OECD average is higher at 7.4 percent of total education spending on non-tertiary education14. Table 13: Total capital spending, by budget entity (2013) Budget entity Capital expenditures (JD) Percentage share Treasury Loans Total Treasury Loans Total Ministry of Finance/HQ 190,396,091 190,396,091 19.15 0.00 18.65 Ministry of Public Works and Housing 163,082,002 - 163,082,002 16.41 0.00 15.97 Ministry of Local Affairs 89,451,936 3,808,496 93,260,432 9.00 14.12 9.13 Ministry of Water and Irrigation 88,415,738 - 88,415,738 8.89 0.00 8.66 Ministry of Health 83,145,375 - 83,145,375 8.36 0.00 8.14 Ministry of Planning & Int'l Cooperation 53,691,343 20,846,707 74,538,050 5.40 77.28 7.30 Royal Medical Services 45,650,000 45,650,000 4.59 0.00 4.47 Ministry of Education 38,599,038 - 38,599,038 3.88 0.00 3.78 Ministry of Transport 28,357,692 28,357,692 2.85 0.00 2.78 Ministry of Defense 28,000,000 28,000,000 2.82 0.00 2.74 Ministry of Interior/Civil Defense 20,800,000 20,800,000 2.09 0.00 2.04 Ministry of Higher Education and Scientific Research 19,190,267 - 19,190,267 1.93 0.00 1.88 Ministry of Water and Irrigation/Jordan Valley Authority 17,231,676 1,257,781 18,489,457 1.73 4.66 1.81 Ministry of Energy and Mineral Wealth 16,492,625 - 16,492,625 1.66 0.00 1.62 Ministry of Agriculture 13,310,153 621,489 13,931,642 1.34 2.30 1.36 Ministry of Interior/General Security 13,913,579 13,913,579 1.40 0.00 1.36 Ministry of Interior/Darak Forces 10,679,969 10,679,969 1.07 0.00 1.05 All other budget entities 73,649,341 441,823 74,091,164 7.41 1.64 7.26 Total 994,056,825 26,976,296 1,021,033,121 100 100 100 Source: Ministry of Finance. June 2014. General Budget Final Accounts for Fiscal Year 2013, pages 466-7, 480 Table 14: Education sector capital spending, by budget entity (2013) Budget entity JDs Percentage share Treasury Loans Treasury Loans Ministry of Education 38,599,038 79.6 Ministry of Planning & International Cooperation 1,224,243 8,654,043 2.5 17.9 - Basic education school buildings -- Construction of 25 schools 1,224,243 1,371,823 2.5 2.8 -- Education Reform for the Knowledge Economy II 7,282,220 15.0 Sub-total 39,823,281 8,654,043 82.1 17.9 Total 48,477,324 100 Source: Ministry of Finance. June 2014. General Budget Final Accounts for Fiscal Year 2013, pages 473, 302 Box 2. USAID Support to the Education Sector in Jordan Since the 1950s, USAID has built schools, donated libraries, and sent thousands of Jordanians to study in US universities, while others were trained in Jordan15. Between 2002 and 2014, USAID invested US$458 million in the education sector. Recent achievements include construction and furnishing of 27 new schools; rehabilitation and expansion of 97 existing public schools; renovation of playgrounds and/or provision of sports kits at 100 schools; and training of 20,000 educators in induction, in-service, leadership, and mentorship programs. 14 OECD 2014. Education at a Glance 2014. Indicator B6, Table B6.2. “Expenditure by educational institutions, by resource category and level of education (2011)�. 15 https://www.usaid.gov/jordan/education , accessed on June 13, 2016. 33 37. Table 15 shows the largest categories of spending, and they fall under 3 programs: early childhood education, basic education, and secondary education. Under all three programs, “additions to the learning environment� is a relatively large item of capital spending. “Additions to the learning environment� under basic education is indeed the largest item overall (17.3 percent of total MOE capital spending), followed by school feeding, ERFKE II, and school furnishings – all under basic education. As such, basic education captures fully 69 percent of total MOE capital spending, followed by a distant second – secondary education – at 16.8 percent. Table 15: Ministry of Education capital spending, by program (2013) MOE Program Amount Percentage (JDs) share Early childhood education Additions to learning environment 1,931,896 5.0 Basic education Administration 1,305,632 3.4 Additions to learning environment 6,672,136 17.3 Building maintenance and repairs 2,383,682 6.2 School feeding 4,989,470 12.9 School furnishings 4,161,145 10.8 ERFKE II 4,953,508 12.8 Construct 60 schools 2,001,965 5.2 Secondary education Additions to learning environment 2,272,472 5.9 Building maintenance and repairs 1,196,994 3.1 All other -- including administration; 6,730,138 17.4 vocational education; educational, social, & sports activities; special education; and illiteracy and adult education Total 38,599,038 100 B. Recurrent Spending 38. The Jordanian budgetary commitment to the education sector is further demonstrated by education’s share in total recurrent spending. Thus, the 2013 budget lists 56 separate budget entities, of which the Ministry of Education and the Ministry of Higher Education and Scientific Research are two. As Table 16 shows, the highest share of recurrent spending is by the Ministry of Finance (over 40 percent), followed by the Ministry of Defense and the Ministry of Education, at roughly 13 percent of total recurrent spending. The Ministry of Finance’s large share in recurrent spending can be explained by the fact that it includes debt interest payments (over 30 percent of Finance’s spending) and pension payments (over 40 percent of Finance’s spending). In effect then, the budgetary commitment to education in Jordan (on a recurrent basis) is substantial and equivalent to that of defense, with other ministries and budgetary entities lagging significantly behind. 39. Combining data on education current spending and numbers of enrolled students at the different levels of education, Table 17 presents the recurrent unit cost of education at each level. The table takes only recurrent cost into account as capital spending tends to be “lumpy� resulting in potentially wide fluctuations in the unit cost from year to year, or alternatively contested methodologies of amortization. Taking unit costs for a basic education student as the comparator in each year, it is clear that a child in 34 Kindergarten costs substantially less, and even a student at the secondary level expends fewer resources. The discrepancy between the basic education and Kindergarten education is to be expected, though arguably in Jordan it is relatively wide as Kindergarten pupils require a higher teacher-student ratio, thus increasing the unit cost. A lower unit cost for secondary students compared to basic students is, however, unexpected as secondary students require more specialized teaching and learning materials, for example. By way of comparison, the unit cost of a pre-primary pupil in the OECD is 10 percent lower than a primary student, and a secondary student is 10 percent higher16. Table 16: Total recurrent spending, by budget entity (2013) Budget entity Current Percentage expenditures (JD) share Ministry of Finance/HQ 2,448,535,998 40.43 Ministry of Defense 834,000,000 13.77 Ministry of Education 803,890,217 13.27 Ministry of Interior/General Security 502,986,420 8.31 Ministry of Health 407,579,785 6.73 Ministry of Interior/Darak Forces 153,220,031 2.53 Ministry of Interior/Civil Defense 145,400,000 2.40 Royal Medical Services 143,100,000 2.36 Ministry of Social Development 106,910,096 1.77 Ministry of Higher Education and Scientific Research 63,919,313 1.06 All other budget entities 446,578,840 7.37 Total 6,056,120,700 100.00 Source: Ministry of Finance. June 2014. General Budget Final Accounts for Fiscal Year 2013, pages 209-10 Table 17: Unit cost, current expenditure (current JDs) Level of education 2012 2013 Kindergarten 131 115 Basic 538 597 Secondary 484 477 Vocational 782 Source: Ministry of Finance. June 2014. General Budget Final Accounts for Fiscal Year 2013 and MOE EMIS C. Syrian Refugee Students 40. In recent years, the question of the unit cost of a basic student in particular has gained importance as the Jordanian education system struggled to absorb the influx of Syrian refugee children. It became necessary to estimate in some manner the additional financial cost incurred by the Jordanian public education system in order to accommodate the Syrian children, and the approach deemed most appropriate for some time relied on using the number of enrolled Syrian students and the unit cost of a student in the Jordanian system to estimate the total cost of Syrian enrollment. At the time, the MOE 16 OECD 2014. Education at a Glance 2014. Indicator B1, Table B1.1a. “Annual expenditure per student by educational institutions for all services (2011)�. 35 preferred to calculate the unit cost not only based on recurrent costs, however, factoring in instead estimates on depreciation of physical infrastructure as well as other costs, resulting in higher unit costs than those reported above. With time, the MOE became very concerned with erosion of certain service standards in the education sector in terms of class size and student-teacher ratio as a result of the Syrian refugee influx, taking these issues into account in estimating the financing needs for absorption of Syrian students into the Jordanian system. 41. Thus, the Jordan Response Plan 2016-2018 estimates education sector needs by carrying out an education sector vulnerability assessment at the district level by using the three indicators of school size, class size, and student-teacher ratio. It defines the national standard for class size at 27, thereby identifying vulnerability to crowding in classes to be most severe in seven districts of the four governorates with high concentrations of Syrian refugees, i.e. Amman, Zarqa, Irbid, and Mafraq. The assessment further finds that 150 new schools would be needed to meet a national standard of 19 classes per school, which translates to a school size of 513 (assuming an average class size of 27). These schools would be located primarily in the same four governorates with highest concentrations of Syrian refugees. Finally, the assessment finds that an additional 8,600 teachers would be needed to meet a national standard of 17 students per teacher. In other words, the MOE is aiming to absorb the Syrian students while maintaining its current class size and student-teacher ratio, as described in Section I of this report. In addition, new schools to be constructed are to accommodate above 500 students, which is greater than the current average school size and in line with efficiency and quality gains outlined in Section I. 42. Beyond abiding by national standards on core education system indicators, the Jordan Response Plan 2016-2018 aims to apply lessons learned from past refugee crises and enhance the Government’s ability to respond to emergencies while at the same time strengthening the education system’s resilience. This resilience implies that the education system is able to adapt and maintain quality in the face of potential new crisis scenarios. Rather than relying on the unit cost approach, the Plan proposes a number of projects that aim to improve access to quality and inclusive education for Syrian refugees and vulnerable Jordanians, boosting the capacity of the public education system with additional learning spaces, remedial/catch-up classes for those children who have missed out on weeks or months of schooling, and access to improved and diversified certified alternative learning opportunities for children and youth. Projects are also to deliver capacity building of teachers to safeguard the quality of education. The resulting three-year Plan starts with a baseline number of 156,663 Syrian children enrolled in education services (whether formal, non-formal, or informal), and targets increases in enrolment to 222,000 in 2016; 248,000 in 2017; and 272,800 in 2018. The increased enrolment is not expected to result from influx of additional Syrian refugees into Jordan, but rather from increased enrolment rates of Syrian children already in Jordan. 43. As Table 18 shows, a range of projects are proposed in the Plan, spanning improving the capacity of education authorities to ensure the continuous delivery of quality inclusive education services; enhancing access to safe and protective learning spaces; and increasing provision of adequate, protective, and safe learning spaces and facilities. In terms of associated costs, as Table 18 shows, over US$850 million are estimated needs over the 2016-2018 period, with over US$470 million dedicated to facilities, i.e. construction of 150 new schools and additional classrooms or renovation of 450 prioritized schools. The fact that attention is being paid not only to construction of new schools but also to adding classrooms to existing schools is warranted given the relatively small average school size in Jordan described in Section I. In other words, the Plan recognizes the necessity for a dual approach that is based on detailed analysis of the needs on the ground in specific locations in the country. The next largest project at over US$180 million includes the hiring and pay of teachers – a topic to be discussed further in the next section. 36 44. The question arises why the Jordan Response Plan 2016-2018 was able to stipulate construction of new schools that are larger than the average school in Jordan today, but did not choose to define standards for class size and STR that are higher than current levels in the country. This is likely the result of several factors, including the comparability of current class size and STR to OECD averages, but also the belief that these national standards contribute to quality of education and improved learning outcomes. As the final recommendations section will argue, it is not clear that maintaining these national standards on STR in particular will lead to improved learning outcomes in the absence of other measures related to accountability and performance monitoring of teachers. Table 18: Education financial requirements summary table, Jordan Response Plan 2016-2018 37 Source: Ministry of Planning and International Cooperation, Hashemite Kingdom of Jordan. 2016. Jordan Response Plan for the Syria Crisis 2016-2018, page 26. 38 V. Teacher Compensation Compensation captures a high share of recurrent spending in the education sector in Jordan, leaving few resources for non-wage inputs. The share of teachers in total education sector staff is 10 percentage points higher than the OECD average. The teacher pay scale and allowances in Jordan reward initial qualifications, seniority, and personal teacher attributes, as opposed to being an instrument for policy makers to incentivize better teaching performance or other desired education sector outcomes. Teachers in Jordan are relatively well-paid, based on comparisons with other tertiary-educated workers in the economy, relative to per capita GDP, and taking teaching time into account. __________________________________________________________________________________ 45. Teachers’ salaries represent the largest single cost in formal education and have a direct impact on the attractiveness of the teaching profession. They influence decisions to enroll in teacher education, become a teacher after graduation (as graduates’ career choices are associated with relative earnings in teaching and non-teaching occupations, and their likely growth over time), return to the teaching profession after a career interruption, and/or remain a teacher (as, in general, the higher the salaries, the fewer the people who choose to leave the profession)17. Since compensation and working conditions are important for attracting, developing, and retaining skilled and high-quality teachers, policy makers carefully consider teachers’ salaries as they try to ensure both quality teaching and sustainable education budgets. 46. A closer look at recurrent spending under the Ministry of Education (see Table 19) shows that fully 92.3 percent of total recurrent spending was dedicated to worker compensation in 2013, with non-wage recurrent spending amounting to 7.7 percent. The non-wage recurrent spending includes spending on rent, utilities, maintenance, and cleaning – i.e. basic operations of educational institutions – as well as direct teaching inputs such as stationery, learning materials, and textbooks. By contrast, the 2011 OECD average share of compensation in total recurrent spending was 78.9 percent, leaving 21.1 percent for non-wage inputs18. Table 19: Ministry of Education recurrent spending, by program (2013) Administration and Vocational Educational, social, Special education Early childhood Basic education Secondary Eradicating Sub-total support services education & physical activities education education illiteracy Worker compensation 4.5 2.8 0.0 0.3 0.3 76.4 7.9 0.0 92.3 Salaries, wages, allowances 4.3 2.6 0.0 0.3 0.3 71.4 6.7 0.0 85.6 Social security contributions 0.2 0.2 0.0 0.0 5.0 1.2 6.6 Goods and services 0.6 0.1 0.1 0.1 0.0 1.8 3.2 0.0 5.8 Assistance 0.0 0.0 Assistance/stipends 0.0 0.0 Other 1.6 0.0 0.3 0.0 1.9 Sub-total 6.7 2.9 0.1 0.4 0.3 78.4 11.1 0.1 100.0 Source: Ministry of Finance. June 2014. General Budget Final Accounts for Fiscal Year 2013, pages 158-174 47. Looking at the breakdown of worker compensations (see Table 20), overall – across all levels of education – additional allowances constitute the largest share at 36.6 percent, followed by wages of non- classified employees (26.8 percent), and the cost of living personal allowance (24.1 percent). However, since employees’ wages appear divided into three categories – classified, non-classified, and contract employees – arguably they should count together and so constitute 36.2 percent of total compensation. The additional allowance reflects the decision to double the base salary of all education sector employees, 17 OECD 2014. Education at a Glance 2014. 18 OECD 2014. Education at a Glance 2014. Indicator B6, Table B6.2. “Expenditure by educational institutions, by resource category and level of education (2011)�. 39 beginning in 2013. In effect, since it is an automatic doubling, it is no longer an allowance but constitutes part of the base salary. The largest allowance is therefore the cost of living personal allowance. Table 20: Ministry of Education worker compensation, by program (2013) Administration and Vocational Educational, social, Special education Early childhood Basic education Secondary Eradicating Total support services education & physical activities education education illiteracy Worker compensation Salaries, wages, allowances Classified employees 17.4 12.7 6.8 8.4 13.8 9.3 Non-classified employees 16.0 23.8 35.0 29.6 27.7 25.1 26.8 Contract employees 1.2 0.1 Cost of living personal allowance 18.0 22.1 25.5 27.1 24.7 22.3 24.1 Cost of living family allowance 2.1 1.7 0.7 0.1 1.3 1.5 1.4 Additional work allowance 0.7 0.8 55.0 1.0 3.7 1.2 Additional allowance 36.4 38.9 37.2 36.4 36.8 33.6 36.6 Other allowances 1.5 0.1 Transportation allowance 4.2 0.2 Transfer compensation 1.0 0.1 Field allowance 0.6 0.0 Employee bonus 0.9 0.0 45.0 1.6 0.0 0.0 100.0 0.1 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source: Ministry of Finance. June 2014. General Budget Final Accounts for Fiscal Year 2013, pages 158-174 48. Worker compensation includes spending on teachers as well as non-teachers in the education sector. Countries demonstrate significant differences in distribution of education staff between teaching and other categories, reflecting differences among countries in the organization and management of schooling. These differences reflect the numbers of staff that countries employ in non-teaching capacities, e.g. principals without teaching responsibilities, guidance counsellors, school nurses, librarians, researchers without teaching responsibilities, bus drivers, janitors and maintenance workers, and administrative and management personnel both inside and outside the school. The OECD classifies education sector staff into two broad categories. The first are instructional personnel, i.e. classroom teachers, special-education teachers, teaching assistants, and teachers’ aides. The second category are non-instructional personnel, i.e. school and higher-level management (e.g. principals, superintendents of schools, and commissioners of education); school and higher-level administrative personnel (e.g. bookkeepers and clerks, network administrators); professional support for students (e.g. guidance counsellors, librarians, nurses, and psychologists); maintenance and operations personnel (e.g. maintenance staff, plumbers, grounds staff, and food servers)19. According to this classification, 69 percent of total staff were instructional personnel in OECD countries on average in 2010 (in primary and secondary education); 12 percent were involved in management and administration; another 10 percent in maintenance and operations; and fully 9 percent in professional support for students20. 49. Table 21 shows the distribution of instructional and non-instructional staff in Jordan, i.e. MOE staff (i.e. employees of the Central Ministry in Amman as well as those distributed across the Education Directorates in the governorates); school principals and assistant principals; education supervisors and counselors; laboratory and computer technicians; school administrators; and teachers. Thus, the Ministry in Amman employs a total of 1,002 staff, and overall MOE staff are 5.1 percent of total education sector staff. Principals and assistant principals make up another 5.3 percent of education sector staff, education supervisors and counselors 3.2 percent, and technicians and school administration another 6.5 percent, for a total of 20 percent for non-instructional staff, so that teachers are 80 percent of total education sector staff. The share of teachers in total education sector staff is therefore roughly 10 percentage points higher than the respective OECD average, and this discrepancy is mostly the result of the seeming absence 19 OECD 2012. Education at a Glance 2012. “The Learning Environment and Organisation of Schools,� pages 447 - 8. 20 OECD 2012. Education at a Glance 2012. Indicator D2, Table D2.4a. “Teaching staff and non -teaching staff employed in primary, secondary, and post-secondary non-tertiary education institutions (2010),� page 453. 40 of the category of professional support for students that constitutes 9 percent of OECD education sector staff. Table 21: Education sector staff, 2015 MOE As s i s tant Educa tion Educa tion La b/computer School Pri nci pa l s Tea chers Total a dmi ni s tra tion pri nci pa l s s upervi s ors couns el ors techni ci a n a dmi ni s tra tion Centra l Mi ni s try 976 26 1,002 Amma n 909 632 643 267 495 804 695 18,326 22,771 Ba l qa 352 204 144 79 146 225 242 5,062 6,454 Za rqa 333 272 138 104 211 387 309 8,605 10,359 Irbi d 685 545 331 201 388 724 527 14,840 18,241 Ma a n 261 155 40 46 58 79 103 3,070 3,812 Ka ra k 367 260 109 92 157 218 314 5,482 6,999 Ja ra s h 103 157 67 33 73 112 110 3,017 3,672 Ajl oun 90 101 58 24 67 123 85 2,469 3,017 Ma fra q 261 400 228 73 167 212 190 7,342 8,873 Aqa ba 104 52 49 18 49 78 73 1,643 2,066 Ma da ba 167 119 80 34 60 103 113 2,563 3,239 Ta fi l a h 144 109 61 30 65 110 96 2,351 2,966 Total 4,752 3,006 1,948 1,027 1,936 3,175 2,857 74,770 93,471 Sha re of total (%) 5.1 3.2 2.1 1.1 2.1 3.4 3.1 80.0 100.0 Source: MOE Human Resource Department 50. The number of teachers has grown consistently over the last decade, so that MOE teachers amounted to a total of 54,609 in academic year 2001-02 (of whom 32,184 were female and 22,425 male) and reached 75,401 in 2012-13 (46,120 female and 29,281 male). This translates to an average annual growth rate of 3 percent over the period 2001-13. However, the number of female teachers grew at a faster rate than male teachers, so that the average annual growth rate for female teachers was 3.3 percent compared to 2.5 percent for male teachers. Figure 15: Number of Ministry of Education teachers, 2001-13 80000 70000 60000 Number of teachers 50000 40000 Total 30000 Male 20000 Female 10000 0 Source: EMIS, MOE 41 A. Teachers Are Paid According to Qualification and Seniority 51. Public education employees fall into three categories, depending on whether they hold a BA (Category 1), a diploma (Category 2), or are technical support staff, e.g. drivers (Category 3). Focusing on Categories 1 and 2, under which teachers would fall, the salary progression takes place both in terms of steps within the same category (seven steps in Category 1 and nine steps in Category 2) as well as by years of service (up to a maximum of fifteen years at the highest step in each of the categories). As a result, the lowest base monthly pay for an entry level Category 1 employee (Step 7) is 150 JD, while the highest pay for a Category 1 employee (Special) with 15 years of service is 593 JD, meaning almost four times as much pay. Similarly, an entry level Category 2 employee (Step 9) earns 125 JD per month, compared to a Category 2 Step 1 employee with 15 years of service who earns 353 JD, i.e. almost three times as much (see Table 22). In many countries, and in most OECD countries, teachers’ salaries increase with the level of education taught21. This is not the case in Jordan. Instead, the initial qualification (BA or diploma) determines the starting salary. Table 22: Education sector staff basic salary scale, 2012 (JD/month) Category Level Grade Years at grade level 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Special 425 437 449 461 473 485 497 509 521 533 545 557 569 581 593 First 1 302 312 322 332 342 352 362 372 382 392 2 260 268 276 284 292 3 228 234 240 246 252 One Second 4 202 207 212 217 222 5 181 185 189 193 197 6 165 168 171 174 177 Third 7 150 153 156 159 162 1 269 275 281 287 293 299 305 311 317 323 329 335 341 347 353 First 2 243 248 253 258 263 3 218 223 228 233 238 4 197 201 205 209 213 Two Second 5 177 181 185 189 193 6 161 164 167 170 173 7 146 149 152 155 158 Third 8 135 137 139 141 143 9 125 127 129 131 133 Three 120 123 126 129 132 135 138 141 144 147 150 153 156 159 162 Source: MOE Human Resources Department 52. In all school systems, teachers’ salaries rise during the course of a career, although the rate of change differs greatly. Since Jordan participated in the 2012 PISA, it is possible to compare its salary scale with other participating countries. Thus, Jordan is among the countries where salaries at the top of the scale are considerably higher than starting salaries – on average for this group, which includes Korea, Shanghai-China, Malaysia, Singapore, Romania, and Jordan, salaries at the top of the scale are 2.5 times higher than starting salaries and it takes between 20 and 40 years to reach the top salary22. By contrast, in Denmark, Iceland, Norway, Slovenia, Sweden, Finland, Germany, the Slovak Republic, the Czech Republic, Spain, Peru, Montenegro, and Croatia, teachers’ salaries at the top of the scale are at most 1.4 times higher than starting salaries. Jordan’s approach to teacher compensation therefore encourages longevity in the sector, since rewards with seniority are substantial. Of course, weak financial incentives 21 OECD. 2014. Education at a Glance 2014, page 456. 22 OECD. 2013. PISA 2012 Results: What Makes Schools Successful? Resources, Policies and Practices (Volume IV), Programme for International Student Assessment, OECD Publishing, page 95. 42 make it more difficult to retain teachers as teachers approach the peak of their earnings. On the other hand, there may be benefits to compressed pay scales, e.g. it is often argued that organizations with smaller differences in salaries among employees enjoy more trust, freer flows of information, and more collegiality among co-workers. In Jordan in particular, the pervasive feeling of teachers that they are under-paid might be related to the decompressed pay scale. 53. In fact, teacher salaries are one component of teachers’ total compensation. Other benefits, such as regional allowances for teaching in remote areas, family allowances, reduced rates on public transport, and tax allowances on the purchase of cultural materials, form part of teachers’ total remuneration in parts of the world. In Jordan, a teaching allowance was put in place in 2013 that doubled the base salary for all staff – a reform unique to education sector employees in the Kingdom. The cost of living allowance (see Table 23) is also an allowance allocated to all staff without preconditions. Other allowances depend on the particular circumstances of the individual, e.g. whether they are married or single; teaching in a remote area; or had to relocate from their home district to another district, governorate, or even region of Jordan. Table 23: Allowances for education sector staff, 2013 Type of Allowance Amount Notes Teaching allowance 100% of base salary Allocated across the board to all staff since January 1, 2013 Cost of living allowance 135 JD/month Allocated across the board to all staff Family allowance 10-20 JD/month 20 JD if married Location allowance 10-30 JD/month Higher for remote areas Relocation allowance 50-150 JD/month 50, 100, or 150 JD/month for relocation to another district, governorate, or region, respectively Rank allowance 6-32% of base salary e.g. 16% for expert teacher, 15% for MOE Department Head, 25% for school principal 54. From the expenditure data reported earlier (Table 20 in particular) and summarized in Table 24, wages (including for classified, non-classified, and contract employees) constitute 36.2 percent of total education staff remuneration, so-called additional allowance another 36.6 percent, cost of living allowance 24.1 percent, and next in size in terms of share of total is the family allowance at 1.4 percent of total, with all other allowances representing 1 percent or less of the total. This approach to allowances blunts their usefulness in terms of achieving certain desired results, such as location of teachers in certain areas, or teaching of certain subjects, or motivating certain behaviors of teachers. For example, additional payments in OECD countries are most often awarded for particular responsibilities or working conditions, such as teaching in more disadvantaged schools, particularly those located in very poor neighborhoods or those with a large proportion of students whose language is not the language of instruction. These schools often have difficulties attracting teachers and are more likely to have less-experienced teachers. In addition, half of OECD countries provide additional payments for special activities, e.g. sports and drama clubs, and teaching students with special needs in regular schools23. Table 24: Main components of education sector staff compensation, 2013 Education staff remuneration Share of total (%) Wages 36.2 Additional allowance 36.6 Cost of living allowance 24.1 Family allowance 1.4 Source: Ministry of Finance. June 2014. General Budget Final Accounts for Fiscal Year 2013, pages 158-174 23 OECD. 2014. Education at a Glance 2014, page 462. 43 55. Additional payments based on teachers’ qualifications, training, and performance are also common in OECD countries. In other words, while Jordan uses teachers’ qualifications to distinguish between the base salary for those with a diploma versus a BA, OECD countries tend to use an allowance to reward an initial education qualification that is higher than the minimum requirement. In this manner, the initial qualification is rewarded but its importance in teacher total pay recedes with time as the growth potential for all teachers is the same. Moreover, among the OECD countries, 21 countries offer an additional payment to teachers for outstanding performance, and in 17 of those countries, the decision to award the additional payments is made by the school principal (OECD 2014). This type of performance- based pay will be discussed further in the next section. B. Teachers Are Relatively Well-paid 56. Given the above-described salary scale and the automaticity of the doubling of the base salary and the cost of living allowance, the lowest starting compensation for a teacher is for a single teacher with a diploma who is not relocating from home and amounts to 405 JD/month (2xbase salary of 125 JD + 135 JD cost of living allowance + 10 JD family allowance + 10 JD location allowance). If the starting teacher has a BA, then the lowest compensation amounts 455 JD/month. Annually, this translates to 4,860 JD or 5,460 JD for starting teachers with diploma or BA, respectively. 57. One comparison that is often employed to assess relative attractiveness of teacher pay is between teacher salaries and the earnings of tertiary-educated workers in the economy. Ideally, teachers’ salaries here either refers to actual salary, including bonuses and allowances, for teachers aged 25-64 or to statutory salary after 15 years of experience and minimum training. The comparison is made then relative to full-time, full-year workers with tertiary education. For the OECD countries, teachers in pre-primary, primary, lower secondary, and upper secondary education earned on average 0.80, 0.85, 0.88, and 0.92 times the amount earned by full-time, full-year workers with tertiary education24. The country with the highest relative teacher pay was Korea, where pre-primary teachers earned 1.32 times the earning of other tertiary-educated workers in the economy, and all other teachers earned 1.36 times as much. 58. In the case of Jordan, using data from the Jordan Employment and Unemployment Survey and Household Income and Expenditure Survey, Assaad et al. (2014) provide first-job wages for individuals between the ages of 25 and 40 in 2012 who graduated in either commerce/business or information technology (IT) from a four-year higher education institution and live in urban areas. The average monthly wage is found to be JD 342, rising to JD 561 five years later (where the individual did not go on to further higher education). This translates to JD 4,104 in 2012 JDs, or JD 4,334 in 201325. By comparison, a starting teacher with BA earned 5,460 JD, or 1.26 times the earnings of a commerce or IT graduate working in the private sector. While the comparison is not completely parallel with the OECD comparison described above, it does provide an indication that teachers are relatively well-paid in Jordan. 24 OECD. 2014. Education at a Glance 2014: OECD Indicators, Table D3.2. “Teachers’ salaries relative to earnings for full-time, full-year workers with tertiary education (2012)�, page 469. 25 2013 inflation rate of 5.6 percent, source Department of Statistics accessed on at http://www.dos.gov.jo/dos_home_a/jorfig/2013/1.pdf 44 59. Another method used often to assess whether teacher pay is adequate or not is to compare teacher pay to a country’s per capita GDP. Thus, per capita GDP in Jordan in 2013 was 3,653 JD26, while the minimum starting salary for a teacher was 4,860 JD (see above), i.e. 1.3 times the per capita GDP. However, this comparison is more typically done not for starting teachers but for teachers with 15 years of experience and minimum training – separately for lower secondary and upper secondary education. Using the 2012 PISA findings, the OECD average for lower and upper secondary is 1.24 and 1.29, respectively. For Jordan, the ratio is the same for both levels of education and stands at 2.15 – the highest ratio amongst all countries participating in PISA in 2012 (see Figure 16). In other words, a teacher with 15 years of experience in Jordan is earning more than twice the per capita GDP according to PISA data. Figure 16: Teacher salaries relative to per capita GDP, 2012 Source: OECD. 2014. PISA 2012 Results: What Students Know and Can Do – Student Performance in Mathematics, Reading, and Science (Volume I), Revised Edition, page 96. 60. The above analysis of teacher salaries has focused on statutory pay using available data on the pay scale. However, actual average teacher pay in Jordan may well be quite different, as the average teacher may not have 15 years of experience and possess minimum qualifications. Table 25 provides data on total salaries and allowances at the governorate level. The wage data are not available by type of education sector employee, i.e. whether they are teachers, principals, supervisors, etc. If there are significant differences across governorates in the breakdown across the different types of education sector employees, then taking a simple average will make the comparison across governorates very 26 Ministry of Finance. July 2014. General Government Finance Bulletin, Studies and Economic Policies Directorate, Volume 16, Number 6, page 10. 45 tenuous. But as Table 26 shows, while there is some variation across governorates in terms of the share of different types of education sector staff in total staff in the governorate, the variation is not great. Therefore, 6,800 JD probably represents a good estimate of the average pay for education sector staff in Jordan in 2015 – although not for teachers specifically. Therefore, given that Jordan’s per capita GDP was 3,811 JD in 201427, average education sector pay equals 1.78 of per capita GDP, i.e. relatively high, but not as high as the ratio reported in comparison to other PISA countries. This, in turn, indicates that the average education sector staff in Jordan has fewer than 15 years of experience on the job, a fact confirmed by Figure 17 from the 2010 SABER report on teachers. Table 25: Average education sector staff pay, by governorate (2015) Governorate Total governorate Total staff Average wage wages and allowances Amman (incl. MOE) 162,220,932 23,774 6,823 Balqa 43,917,540 6,453 6,806 Zarqa 69,781,632 10,359 6,736 Madaba 21,924,624 3,239 6,769 Irbid 127,378,668 18,241 6,983 Mafraq 58,181,928 8,873 6,557 Jarash 25,154,568 3,672 6,850 Ajloun 20,585,724 3,017 6,823 Karak 47,644,260 6,999 6,807 Tafilah 19,880,508 2,966 6,703 Maan 24,860,952 3,812 6,522 Aqaba 13,679,316 2,066 6,621 Total 635,210,652 93,471 6,796 Source: MOE Human Resource Department Table 26: Share of different education sector staff, by governorate (2015) MOE As s i s tant Educa tion Educa tion La b/computer School Pri nci pa l s Tea chers Total a dmi ni s tra tion pri nci pa l s s upervi s ors couns el ors techni ci a n a dmi ni s tra tion Centra l Mi ni s try 97.4 2.6 100.0 Amma n 4.0 2.8 2.8 1.2 2.2 3.5 3.1 80.5 100.0 Ba l qa 5.5 3.2 2.2 1.2 2.3 3.5 3.7 78.4 100.0 Za rqa 3.2 2.6 1.3 1.0 2.0 3.7 3.0 83.1 100.0 Irbi d 3.8 3.0 1.8 1.1 2.1 4.0 2.9 81.4 100.0 Ma a n 6.8 4.1 1.0 1.2 1.5 2.1 2.7 80.5 100.0 Ka ra k 5.2 3.7 1.6 1.3 2.2 3.1 4.5 78.3 100.0 Ja ra s h 2.8 4.3 1.8 0.9 2.0 3.1 3.0 82.2 100.0 Ajl oun 3.0 3.3 1.9 0.8 2.2 4.1 2.8 81.8 100.0 Ma fra q 2.9 4.5 2.6 0.8 1.9 2.4 2.1 82.7 100.0 Aqa ba 5.0 2.5 2.4 0.9 2.4 3.8 3.5 79.5 100.0 Ma da ba 5.2 3.7 2.5 1.0 1.9 3.2 3.5 79.1 100.0 Ta fi l a h 4.9 3.7 2.1 1.0 2.2 3.7 3.2 79.3 100.0 Sha re of total 5.1 3.2 2.1 1.1 2.1 3.4 3.1 80.0 100.0 Source: MOE Human Resource Department 27 Ministry of Finance. January 2016. General Government Finance Bulletin, Studies and Economic Policies Directorate, Volume 17, Number 12, page 18. 46 Figure 17: Distribution of public school teachers by age and gender, 2010 Source: World Bank. 2010. Jordan Teachers SABER Country Report, Figure 4, page 3. 61. Yet another angle to assessing teacher pay in Jordan is the following: in addition to class size, student-teacher ratio, and teachers’ salaries, the number of hours of student instruction and the amount of time teachers spend teaching also affect the financial resources countries need to allocate to education. In Jordan, students spend on average 6 hours per day, 5 days a week in school – adding up to 30 hours of instruction per week. For teachers, the average workload is 24 lessons a week, although there is some variation: for basic education, teachers’ workload is 24-26 lessons a week, each lesson 45 minutes long; and for secondary, the weekly workload is 18-20 lessons, each 55 minutes long. The relationship between class size, student-teacher ratio, student instruction time, and teaching time can be described as: Class size = student-teacher ratio x student instruction time / teaching time per teacher Using available data (see Table 7) on average class size and student-teacher ratio, this relationship holds for Jordan using the average student instruction time of 30 hours per week and teaching time per teacher of 18 hours per week (24 lessons each 45 minutes long). 62. Given that the school year in Jordan consists of 195 days (slightly above the OECD average of 180- 183 depending on the level of education), the annual teaching workload is provided in Table 27 for basic and secondary levels of education, using the different weekly teaching workloads (and keeping in mind that lessons in basic education are 45 minutes long, whereas they are 55 minutes long in secondary education). As Table 27 shows, annual teaching hours are roughly comparable with averages observed in OECD countries28. In other words, teachers in Jordan are not teaching above average annual hours so that above average pay is warranted. 63. In most countries, teachers are formally required to work a specified number of hours per week, including teaching and non-teaching time, to earn their full-time salary. Some countries also regulate the time that a teacher has to be present in the school. In fact, more than half of OECD countries specify the time during which teachers are required to be available at school, for both teaching and non-teaching activities, at one or various levels of education. Although teaching time is a substantial component of teachers’ workloads, assessing students, preparing lessons, correcting students’ work, in-service training and staff meetings should also be taken into account when analyzing the demands placed on teachers in 28 OECD. 2014. Education at a Glance 2014, page 474. 47 different countries. The amount of time available for these non-teaching activities varies across countries, and a large proportion of statutory working time spent teaching may indicate that less time is devoted to activities such as assessing students and preparing lessons. Table 27: Teaching hours per year Level of education Jordan teaching workload OECD average Weekly Annual Annual Basic 26 761 782 (primary) 24 702 694 (lower secondary) Secondary 20 715 655 (upper secondary) 18 644 655 (upper secondary) Source: World Bank SABER 2010; Ministry of Education; OECD 2014 page 474 64. In Jordan, the Ministry of Education has outlined teachers’ duties as: teaching, grading assignments, supervising students, integrating difficult students, mentoring fellow staff members, standing in for absent teachers, carrying out administrative functions, collaborating on the school plan, and taking part in the internal school evaluation. The Ministry of Education further determines teachers’ working time, stipulating that teachers spend the school day on the school premises. Since the school year consists of 195 days with 6 hours of school per day, this amounts to 1,170 hours of teacher working time annually. While such a definition is more favorable than limiting working time only to hours spent directly in the classroom, it does not go far enough in recognizing that lesson planning and grading may take place outside of official school hours. Nonetheless, the resulting teacher working time required at school in Jordan is almost identical to the OECD average for lower secondary (1,173 hours), below the OECD average for primary (1,200 hours) and above the OECD average for upper secondary (1,142 hours). Therefore again, teachers in Jordan are working average numbers of hours annually and receiving relatively high pay. 48 VI. Recommendations 65. As described in this report, the Jordanian government’s commitment to education is apparent in the amount of resources it has dedicated to education over the years. These investments have led to a system that has made great strides in terms of schooling access and attainment. At the same time, education policy makers are aware that quality of education remains below international standards. In order to improve quality, basic system attributes such as school size, class size, and student-teacher ratios have been an important concern for policy makers – particularly in the current context of influx of Syrian refugee students into the system. In addition, policy makers have sought to address the need for better teaching by implementing an across-the-board doubling of teacher base salaries. 66. Arguably, these measures do not tackle sufficiently the shortcomings that the Jordanian education system is exhibiting today. One set of important issues revolves around paying teachers in a manner that incentivizes better performance on their part, leading to better teaching and learning, as well as matching their performance better with students’ needs. Another set of issues concerns capital investments in the education sector, and the need to tackle the small school size, rented schools, as well as localized overcrowding. And a third set of issues revolves around the persistent regional differences that are observed – whether in terms of access or learning outcome indicators. A. Motivating Teachers to Perform 67. Teacher quality is the key driver in achieving student learning29. It is true that students’ family background (parent education, socioeconomic status, and conditions at home such as access to books) is the strongest predictor of students’ learning outcomes. But in terms of schooling inputs, the single most critical factor is teacher quality. Research in the United States on the “value added� of individual teachers over the course of a single school year has shown that students with a weak teacher may master 50 percent or less of the grade curriculum; students with a good teacher get an average gain of one year; and students with great teachers advance 1.5 grade levels or more30. Beyond technical knowledge, teacher quality requires adequate levels of effort to be exerted to meet professional standards. Yet, at times, absent monitoring and incentive mechanisms, teachers may exert suboptimal levels of effort, thereby undermining the quality of education service delivery31. 68. In terms of motivating teachers to perform, incentives may be classified into three broad categories: (a) professional rewards, e.g. recognition and prestige; (b) accountability pressure; and (c) financial incentives32. Cross-country studies suggest that no education system achieves high teacher quality without aligning all three types of incentives. Given the nature of this report, the discussion here is focused on financial incentives. Across-the-board salary increases – as was recently implemented in Jordan – are inefficient: for the same fiscal expenditure, school systems can achieve higher quality by 29 Bruns, Barbara and Javier Luque. 2014. Great Teachers: How to Raise Student Learning in Latin America and the Caribbean. World Bank. 30 Bruns, Barbara and Javier Luque. 2014. Great Teachers: How to Raise Student Learning in Latin America and the Caribbean. World Bank. 31 World Bank. 2015. Accountability: The Last Mile on the Route to Quality Service Delivery – Evidence from Jordanian Schools & Primary Health Centers. 32 Bruns, Barbara and Javier Luque. 2014. Great Teachers: How to Raise Student Learning in Latin America and the Caribbean. World Bank. 49 raising average salaries through a pay scale differentiated by performance. This avoids overcompensating weak performers, can keep overall pension liabilities lower, and creates stronger incentives for the most talented individuals. 69. A recent World Bank study33 finds that teachers’ effort to meet professional standards in Jordan is seemingly low. The study focuses on four measures of teacher effort that it derives from the Jordanian teacher professional standards as defined by the Civil Service Bureau. According to these standards, teachers are required to (i) provide continuous feedback to students; (ii) reply to students’ questions in a manner that supports their learning; (iii) provide varied forms of student assessment; and (iv) take into account students’ performance and needs in lesson design. The study reports that only 20 percent of teachers in the early grades provide continuous feedback in their students’ copybooks, while ro ughly 25 percent mark only a few pages, and 3.4 percent do not mark a single page. Most teachers’ reaction to a student who is unable to answer a question is to simply repeat the question to the same student, and 5.4 percent of teachers even punish the student for being unable to provide the answer. Almost two-thirds of teachers rely on a limited number of student assessment methods and only a quarter of teachers make use of the assessments for lesson plans. 70. In other words, though the teacher professional standards exist in Jordan, teachers in effect have very little incentive to abide by these standards and are not held accountable to them. As described earlier in this report, the teacher salary pay scale is based on initial credentials and seniority, with no financial incentives for performance. At the school level, principals are not empowered to manage teachers for performance – they do not have the authority to hire and fire staff -- nor do they have the budget to provide additional pay. On the other hand, the study finds that principals in Jordan seldom rely on non-financial incentives to encourage teachers to perform, instead making use of mechanisms to sanction, if at all34. 71. Therefore, it is arguably the case that Jordan has achieved an acceptable threshold in terms of “hard� structural inputs for its education system, but remains behind on the necessary “soft� accountability and incentive measures that will maximize the benefits out of the structural inputs (see Box 3 for recent innovations in accountability in the education sector in Jordan). There is therefore a need for a shift towards performance-based education, whereby teacher accountability is at the heart of the sector reform agenda. Moving toward such a performance-based accountability system requires35: (i) selection of adequate indicators to measure provider performance; (ii) collection of the relevant standardized data for these performance indicators; and (iii) design of effective rewards and sanction schemes against these performance indicators. This in turn would mean that performance indicators would inform the design of professional development programs, so that the teacher hiring and deployment system is re-oriented towards enhancing teacher performance. 33 World Bank. 2015. Accountability: The Last Mile on the Route to Quality Service Delivery – Evidence from Jordanian Schools & Primary Health Centers. 34 World Bank. 2015. Accountability: The Last Mile on the Route to Quality Service Delivery – Evidence from Jordanian Schools & Primary Health Centers. 35 World Bank. 2015. Accountability: The Last Mile on the Route to Quality Service Delivery – Evidence from Jordanian Schools & Primary Health Centers. 50 Box 3. Strengthening Accountability in Jordan’s Education System The World Bank is supporting the Jordanian Ministry of Education in strengthening the governance of the education sector through the establishment of an accountability system that incentivizes stakeholders in the system to improve education quality in public schools. The accountability process works in the following manner: Schools undergo a process of self-evaluation and are also assessed against a holistic quality framework that is underpinned by the four critical domains of teaching and learning; student environment; community relationships; and leadership and management. The external assessment is conducted by a highly trained team of professional assessors, and the results are made public in the form of a school report. Schools and field directors are held accountable for addressing weaknesses identified in the assessments, and the community is made aware of their school’s strengths and weaknesses. This public and professional accountability is expe cted to improve learning and education quality. The ongoing project has achieved major milestones to date, including 1) passing of legislation ensuring that the Education Quality and Accountability Unit is institutionalized in the education system; 2) training and formal appointment of 20 assessors to the Unit (target is 80); and 3) development of technical materials necessary for the operation of the Unit. The process of school assessment and follow-up is in the pilot phase, with plans for national scale-up in 2017-2018. All schools in Jordan will be subject to external assessment, and all schools will be held to the same high standards. 72. The institutionalization of mechanisms to motivate high teacher performance is therefore an area that needs to take center stage in the teacher reform agenda in Jordan. In a positive step in this direction, Jordan has introduced a teacher appraisal system which requires teachers to take part in annual performance evaluations that include assessments from their school principals and Education Directorate supervisors. These appraisals are intended to inform decisions on promotions and raises on the salary scale at the central level. Yet, anecdotal evidence suggests that the existing teacher appraisal system does not effectively distinguish among high and low performing teachers and is unable to link career opportunities and compensation to teachers’ performance. The above-mentioned World Bank study36 conducted a number of case studies in Jordanian schools and concluded that across all visited schools, there is a prevalent belief by teachers that they will receive an automatic promotion and salary increase after four to six years, regardless of their performance. One potential driver behind this belief is that criteria to assess teacher performance tend to center on issues pertaining to teacher attendance, timeliness, and perfunctory reviews of lesson plans, while often overlooking teacher pedagogical practices, content knowledge, and student learning outcomes. According to these authors, these rather administrative indicators are insufficient measures to effectively distinguish among high and low performing teachers, resulting in most teachers being graded as “excellent� in their annual performance evaluations. Thus, providing any type of incentive mechanism at the central level becomes increasingly difficult in a national pool where all or most teachers are ranked as excellent. 73. Strengthening the ability of the teacher performance evaluation system to better capture measures of teacher performance that lead to improved student learning should be a priority area for education reform. The case of Ontario’s Teacher Performance Appraisal (TPA) system may provide important insights in this regard. The TPA includes systematic steps that principals and teachers need to follow. First, a pre-observation meeting is held between the principal and the teacher to discuss how the teacher’s performance will be assessed, including a review of the competencies that will form the basis of the teacher’s performance appraisal. Second, the principal conducts one or more classroom observations 36 World Bank. 2015. Accountability: The Last Mile on the Route to Quality Service Delivery – Evidence from Jordanian Schools & Primary Health Centers. 51 to assess teachers’ skills, knowledge and attitudes in her classroom. Third, a post-observation meeting is held to review the results of the classroom observation and identify specific areas for training and development. Finally, a summative report is produced by principals to document each teachers’ appraisal. The report includes a record of meetings and classroom observation dates, the principal’s comments regarding the teacher’s competencies, the principal’s overall rating of the teacher’s performance, and recommended professional growth goals and strategies for the teacher to pursue. The principal must also collect evidence of the teachers’ performance and attach it to the summative report. Whereas new teachers are given a “development needed� or “satisfactory� score, experienced teachers are rated as either “satisfactory� or “unsatisfactory�. A “development needed� or “unsatisfactory� rating requires that an improvement plan be developed identifying very specific areas in which the teacher must improve. Where steps have been taken to provide support to a teacher, but a teacher’s performance is still rated as “unsatisfactory�, the principal recommends the termination of the teacher’s employment37. 74. UNRWA provides an example closer to home38. UNRWA operates one of the largest non- governmental school systems in the Middle East. It manages nearly 700 schools, hires 17,000 staff, educates more than 500,000 refugee students a year, and operates in five areas (Jordan, Lebanon, Syria, West Bank, and Gaza). UNRWA students outperform public schools in Jordan, West Bank, and Gaza – by a year’s worth of learning – as evidenced by TIMSS 2007 results. Thus, UNRWA students in Jordan, West Bank, and Gaza on average achieve scores 23 to 80 points higher than their peers at public schools, even after controlling for student characteristics and for urban or rural contexts. The reasons include: • UNRWA selects, prepares, and supports its education staff to pursue high learning outcomes • Time-on-task is high in UNRWA schools and this time is used more effectively than in public schools • UNRWA schools have a world-class assessment and accountability system UNRWA has a well-defined and carefully implemented accountability system to assess and support teaching and learning. Assessments are priority and are disseminated to teachers (to inform lesson plans and instructional practices) and to policy makers. Teachers are also required to participate in internal and external monitoring and evaluation. While both public and UNRWA evaluation systems include classroom observations and multiple criteria, UNRWA evaluations are more rigorous and frequent. To remain in the teaching profession in UNRWA, professional development and performance evaluations are requirements, with incentives for good performance and sanctions for poor performance39. 75. The two main strategies for differentiated financial rewards for teachers are career path reforms and bonus pay. Career path reforms typically make permanent promotions contingent on teachers’ skills and performance rather than on seniority and expand salary differentials across different grades. Bonus pay programs are politically and technically easier to implement than career path reforms and do not have long-terms fiscal or pension implications. Bonus programs typically offer a one-time reward for teachers (or schools) for specific results achieved during the prior school year. The evidence to date provides some evidence on short-term impacts on teacher and school performance40. 37 Ministry of Education of Ontario. 2010. Teacher Performance Appraisal: Technical Requirements Manual, retrieved from http://www.edu.gov.on.ca/eng/teacher/pdfs/TPA_Manual_English_september2010l.pdf . 38 World Bank. 2014. Learning in the Face of Adversity: The UNRWA Education Program for Palestine Refugees. 39 World Bank. 2014. Learning in the Face of Adversity: The UNRWA Education Program for Palestine Refugees. 40 Bruns, Barbara and Javier Luque. 2014. Great Teachers: How to Raise Student Learning in Latin America and the Caribbean. World Bank. 52 76. In this regard, the case of Chile’s teacher merit pay program is particularly informative. Since 1997, Chile instituted a group incentive pay program (Sistema Nacional de Evaluación del Desempeño de los Establecimientos Educacionales or SNED), whereby all teachers within a school may be granted a bonus on top of their base salaries, contingent –primarily– on their students’ performance on national standardized tests. Other secondary criteria for SNED bonuses include the schools’ student grade repetition, dropout rates, equity policies, and working conditions41. To award the bonuses, the program compares schools with similar student and school-level characteristics, in order to ensure a level playing field for all.42 The incentive bonus has been estimated to effectively raise teachers’ average monthly salaries between 5 to 11 percent annually43. The program has indeed motivated teachers to perform better, resulting in improvements in students’ standardized language and math scores, of up to 0.23 and 0.25 of a standard deviation in language and math, respectively44. 77. Ensuring that teachers are efficiently deployed to where their skills are most needed is another priority area for teacher reform in Jordan. Absent a purposeful allocation, teachers tend to gravitate towards better-off schools and easiest to teach subjects, widening learning inequalities and creating shortages in critical teaching areas. With regards to the former, Jordan has already begun to address this challenge by providing monetary bonuses and subsidized housing to incentivize teachers in taking up posts in hard-to-staff schools45. Yet, further incentives might be needed to attract the best teachers to low- performing schools. For example, high-performing education systems including Australia, New Zeeland, and the United States provide scholarships, travel stipends, and promotion opportunities as additional incentives to attract talented teachers to disadvantaged areas.46 78. The case of the State of California, US, may be particularly informative for the design of innovative incentive schemes in Jordan. From 2000 to 2002, California introduced the Governor’s Teaching Fellowship –an incentive aimed at attracting highly talented, newly licensed teachers to staff low- performing schools. Application to the fellowship was competitive, entailing a selective screening process whereby prospective students submitted their school transcripts, letters of recommendations, resume, and statement of interest, and also participated in an interview. Successful applicants received 20,000 USD conditional on a four-year commitment to teach in a low-performing school. The fellowship amount offered a 15.1 percent annual premium on the starting salary of these teachers, and was estimated to 41 Alger, V. 2014. Teacher Inventive Pay that Works: A Global Survey of Programs that Improve Student Achievement . Barbara Mitchell Centre for Improvement in Education. 42 Alger, V. 2014. Teacher Inventive Pay that Works: A Global Survey of Programs that Improve Student Achievement . Barbara Mitchell Centre for Improvement in Education. 43 Manzi, J., K. Strasser, E. San Martín, and D. Contreras. 2008. Quality of Education in Chile. Measurement Center, Catholic University of Chile; Mizala, A. and P. Romaguera. 2005. “Teachers’ Salary Structure and Incentives in Chile,� in Emiliana Vegas (editor) Raising Student Learning in Latin America: The Challenge for the 21st Century , pages 103- 150, World Bank; Mizala, A. and B. Schneider. 2014. “Negotiating Education Reform: Teacher Evaluations and Incentives in Chile (1990–2010),� Governance 27, 1 (January): 87–109; Vegas, Emiliana and Ilana Umansky. 2005. “Improving Teaching and Learning through Effective Incentives,� in Emiliana Vegas (ed.) Incentives to Improve Teaching: Lessons from Latin America, World Bank, pages 1–20. 44 Rau, T. and D. Contreras. 2011. Tournament Incentives for Teachers: The Case of Chile , unpublished paper, University of Chile. 45 World Bank. 2010. Jordan Teachers SABER Country Report. 46 OECD. 2011. Establishing a Framework for Evaluation and Teacher Incentives: Considerations for Mexico , OECD Publishing. 53 increase the likelihood of similar teachers taking up posts in low-performing schools by 28 percentage points47. 79. As for hard-to-staff subjects, Jordan has identified important teacher shortages in critical subjects such as mathematics, science and technology. However, mechanisms to incentivize teachers to take up teaching posts in these subjects are absent.48 Teacher reform should focus on introducing effective incentive packages to offset shortages in these critical subjects by attracting highly qualified teachers and even mid-career professionals. A number of education systems have designed and implemented incentives of this type, offering Jordan a wealth of experiences for guiding its policy reform. For example, in the US State of Utah, mathematics and science teachers receive a 5,000 bonus on top of their base salary. In the State of New York, teachers for shortage subjects (mathematics, science, and special education) receive housing assistance of up to 15,000 USD. Similarly, the United Kingdom covers a significant portion of qualified mathematics, science, special education and technology teachers’ tuition expenses, as an incentive to attract them into these subjects49. B. Realigning Capital Investments 80. This report has documented the relatively low level of capital spending in total education spending in Jordan, as well as its dispersion across the Ministry of Education and Ministry of Planning and International Cooperation. In addition, policy makers are concerned about multiple infrastructural aspects of the education system: overcrowded classrooms, partly resulting from the influx of Syrian refugee students; and the need to resort to renting inappropriate buildings for lack of land for constructing schools. At the same time, available school buildings are on average small. 81. As a first priority, there is a need to consolidate at the Ministry of Education, if not management of capital spending, at least the accounting of capital spending. Currently, no single government entity, neither the MOE nor MOPIC, has a complete picture of capital spending for the education sector, i.e. including financed by the treasury as well as loans and donors. While there may be reasons for managing funds from different sources by different ministries, the Ministry of Education is the appropriate government entity for oversight of all capital spending, including determining priority needs and deploying resources accordingly. 82. The Ministry of Education today has an impressive EMIS that has supplied much of the data used in this report. It contains data by school in terms of numbers of students, teachers, and classrooms, amongst other information. However, in order to make decisions regarding expanding existing schools or building new ones, the MOE requires a geographic database that maps schools across the country in addition to the EMIS. Ideally, it also requires relatively detailed population data and population projection data. The MOE has begun working on a school mapping as it recognizes that overcrowded classrooms are not the norm but rather specific to certain areas of the country. However, the question remains whether the data from this school mapping will be used for purposes of decision making regarding expansion of 47 Steele, J. L., R. J. Murnane, and J.B. Willett. 2009. "Do Financial Incentives Help Low-Performing Schools Attract and Keep Academically Talented Teachers? Evidence from California," NBER Working Paper 14780, Cambridge, MA: National Bureau of Economic Research (NBER). 48 World Bank. 2010. Jordan Teachers SABER Country Report. 49 OECD. 2011. Establishing a Framework for Evaluation and Teacher Incentives: Considerations for Mexico , OECD Publishing. 54 existing schools or establishment of new ones. Anecdotal evidence suggests that too often in Jordan, communities insist on expansions of schools and construction of new ones and politicians comply rather than risk losing favor with communities. 83. Other factors that lead to existence of small schools are at play as well, such as the lack of a clear consistent policy on school transportation or busing – an approach that would serve to reduce the need for construction of new schools potentially (although it would be associated with increased recurrent transportation costs). The same issue applies regarding the policy on co-educational schools, which in some communities means that schools are co-educational through grade 4 while in others only through grade 2. Having separate schools for girls and boys has clear repercussions on the numbers of schools needed in a community, and the earlier the separation, the greater the loss in terms of economies of scale. 84. Even if maintaining smaller schools remains a necessity in Jordan – perhaps in rural or more remote areas where commuting times may be so long that they would negatively impact students and teachers by increasing fatigue – school clustering may be a viable option. School clustering brings together groups of small schools and has become a common response to the disadvantages associated with smaller schools. Clustering allows individual schools to maintain their identity and institution, while at the same time sharing resources, experience, and good practices50. Schools in a cluster maintain their institutions and establish formal links of cooperation with other institutions, sometimes sharing a common management or direction. School clusters can implement a more horizontal management model, one that encourages peer-level exchanges. Different countries and regions have adopted different forms of clustering. 85. In Portugal, school clusters were formed not only as a means of formal cooperation and coordination between different schools (so as to rationalize the administration and management of schools), but also to provide a sequential and coordinated path between the different levels and cycles of compulsory education51. Portuguese school clusters are composed of pre-school establishments and one or more education cycles, with a common administrative and management body. This facilitates transition between different levels of education in one specific geographical area, and it also helps overcome the isolation of the different establishments and rationalize the use of resources. 86. In Catalonia and other autonomous communities in Spain, clustering schools share teachers who travel between the different rural schools in the cluster, a system that helps overcome the scarcity and the high costs of specialized teachers in rural areas52. Such a system might be applicable for Jordan in order to address the dearth of male science and math teachers, with incentives provided to compensate them (and the fact that they are male means that it is more culturally acceptable for them to travel to more distant schools). In Catalonia, each cluster is coordinated by a common direction body including a cluster principal (one of the principals of the clustering schools), a chief of studies, and a secretary. Teachers will commonly teach in one school per day, to avoid travelling between schools during the day. 50 Ares Abalde, Macarena. 2014. “School Size Policies: A Literature Review,� OECD Education Working Papers, No. 106, OECD Publishing. 51 Ares Abalde, Macarena. 2014. “School Size Policies: A Literature Review,� OECD Education Working Papers, No. 106, OECD Publishing. 52 Ares Abalde, Macarena. 2014. “School Size Policies: A Literature Review,� OECD Education Working Papers, No. 106, OECD Publishing. 55 In most clusters, teachers come together from the different schools every two weeks to plan school activities. 87. Separate from the question of small schools, Jordan also faces the issue of lack of land for school construction in densely populated areas such as Amman, so that the MOE has to resort to renting inappropriate buildings. These rented buildings seem to be negatively associated with student learning outcomes, as was shown in the analysis of the tawjihi pass rates. At the same time, the issue of overcrowded classrooms is very much an urban phenomenon and in schools with only a single shift. The MOE regards double shifting as an attribute of its education system that is to be avoided, even if it means that inappropriate buildings need to be rented instead. In the continued absence of land for construction of new schools, and the apparent negative effects of rented buildings on learning outcomes, the policy on double shifting may need reconsideration. There is now recent experience as a result of the Syrian crisis with double shifting, given that Syrian students are often in the second shift in public schools, and valuable lessons learned. These could inform a revision of the policy and implementation of double shifting for Jordanian students. C. Addressing Regional Variation 88. This report has documented variation across governorates in terms of enrolment rates in pre- primary and secondary – though not basic – education, as well as the fact that governorate of residence is associated with tawjihi pass rates (and more so than school attributes such as school size, class size, or student-teacher ratio). There is a clear need for the MOE to try and understand the reasons behind this geographic variation better and to try to address them. The fact is that governorates do vary widely in terms of population density, for example, with 70 percent of the population estimated to reside in just 3 governorates (Amman, Irbid, and Zarqa). At the same time, it is not only population density that plays a role in terms of learning outcomes, as the governorate with lower tawjihi pass rates vary in this respect as well. 89. Yet the MOE treats all governorates in the same manner and its departments are organized by level of education and pedagogical theme instead of including consideration of lagging regions, for example. Thus, the MOE consists of the following departments: Minister’s Office; Director-General’s Office; Buildings and International Projects; Planning and Education Research; Special Education; Education; Private Education; Human Resources; Queen Rania Center for Educational Technology and Information; Examinations and Testing; Vocational Education; Monitoring, Inspection, and Quality Assurance; Legal Affairs; Financial Affairs; Cultural and International Relations; Procurement; Curriculum and Textbooks; Educational Activities; Educational Media and Social Communication; General Diwan; Secretariat of the National Jordanian Board for Education, Culture, and Science; Secretariat of the Education Board and Planning Committee; Center for Educational Training; and the Unit for Development Coordination. In none of the above-mentioned departments is there particular attention paid to education of the socio-economically disadvantaged, for example, or of the needs of remote or lagging areas. 90. In addition, discussions with MOE staff indicate that education sector budget preparation is done very much on an annual historical basis and based simply on demographic needs. The budget preparation process may be described as follows: in April each year, the MOE Finance Unit solicits budget requests for the following year from the 43 Education Directorates in Jordan. The Directorates prepare their requests based on the student numbers they anticipate for the following year. The Finance Unit compiles 56 all requests and provides them to the individual Ministry Departments for their review, e.g. vocational schools to the Vocational Department, etc. The Departments in turn may revert to the Directorates for clarifications or modifications – but mostly if something seems particularly out of the ordinary. In July, a first draft of the Ministry budget is then submitted to the Central Planning Agency. In other words, Education Directorates mostly request additional teachers and other standard inputs based on projected enrollment for the following year. 91. In addition, MOE’s requested budget is always revised downward in discussions with the Ministry of Finance. And in terms of budget execution, every month the Education Directorates submit their payment requests, which are compiled in the Ministry and submitted to the Ministry of Finance. Every year, certain line items run out of funds and transfers from other line items that have under-disbursed are made. For example, in 2014 the funds for water and electricity costs were depleted, and transfers were made out of the funds for textbooks. Here again, there is little effort at prioritization of certain line items above others – with the exception of salaries, which are always the first responsibility in terms of disbursement. 92. The budgeting process described above is not conducive to tackling the issue of geographic variation in education sector outcomes. Instead, arguably a more “projectized� approach, one that the Ministry has recently employed in addressing the Syrian refugee students crisis, is likely more appropriate. Given the recent experience with devising a 3-year plan for the Syrian refugee students, education sector stakeholders in Jordan seem well poised for similar planning for lagging regions in the country. 93. Table 28 provides succinctly short- and more medium-term measures that could be undertaken to implement policy reforms in each of the areas of motivating teachers to perform, realigning capital investments, and addressing regional variation. 57 Table 28: Policy Options for the Short- and Medium-Terms Considerations for the Short-Term Considerations for the Medium-Term Policy Area (1-2 years) (3-5 years) 1. Select adequate indicators to measure 5. Informed by results from the impact teacher performance that have been evaluation, adjust the pilot design and proven to drive student learning and that recalibrate the incentive amount and lie within teachers’ span of control53 scheme as need be 2. Establish a standardized system to collect 6. Roll out the teacher bonus program such indicators, with school principals at across governorates the frontline of monitoring endeavors, and 1.1 Motivating Teachers to Education Directorates and MOE 7. Ensure continuous monitoring of the Perform conducting independent verification54 program by an oversight body housed inside the MOE (potentially the newly 3. Design and pilot a teacher bonus program established Accountability Unit) that is tied to the above indicators 8. Use collected teacher performance 4. Conduct an impact evaluation to measure indicators to design and continuously the effect of the bonus program on teacher update professional development performance and, to the extent possible, on programs student learning 1. Conduct/update a detailed mapping of 4. Informed by the evaluation results, hard-to-staff schools in the country adjust the piloted incentives package as 1.2 Effectively Deploying need be Teachers to Hard-to-Staff 2. Design and pilot an incentives package to Schools attract high quality newcomers as well as 5. Roll out the incentives package and experienced teachers to these schools update the school mapping annually to ensure proper targeting 53 See World Bank. 2015. Accountability: The Last Mile on the Route to Quality Service Delivery – Evidence from Jordanian Schools & Primary Health Centers, for further detail and examples of indicators to measure teacher performance in Jordan. 54 See World Bank. 2015. Accountability: The Last Mile on the Route to Quality Service Delivery – Evidence from Jordanian Schools & Primary Health Centers, for further detail on the monitoring role of school principals in Jordan. 58 3. Conduct an evaluation of the effectiveness of the incentives package in staffing these schools 1. Design and pilot an incentives package to 3. Informed by the evaluation results, recruit talented candidates for taking up adjust the piloted incentives package as teaching posts in identified shortage 1.3 Effectively Attracting need be subjects (e.g. mathematics and science) Teachers to Hard-to-Teach Subjects 4. Roll out the incentives package and 2. Conduct an evaluation of the effectiveness update shortage subjects annually to of the incentives package on teacher ensure proper targeting recruitment in these subjects 1. Upon consultations with relevant stakeholders, consolidate oversight and possibly also accounting of capital 5. Informed by the evaluation results and spending at the MOE the complete school mapping, consider rolling out school clusters and double 2. Use available information on school shifts to increase economies of scale mapping to identify candidates for school 2. Realigning Capital clustering and double shifts 6. Institute mechanisms to ensure all Investments capital spending, including the 3. Initiate school clustering and double shifts definition of priority needs and the in a small sample of schools and conduct an deployment of resources, are overseen evaluation of its effectiveness by MOE and conducted based on identified need through the EMIS 4. Undertake a complete school mapping exercise that is linked to the EMIS 59 1. Assign the mandate of improving educational outcomes in lagging schools/areas to an existing or new department at MOE. 3. Execute the developed plan and 3. Addressing Regional 2. Develop a 3-5 year plan (to be executed by evaluate regional variations in Variation the designated department) on how to educational outcomes annually strategically address educational needs in lagging schools/regions, including the necessary steps for budgetary planning that takes into consideration regional variation in educational outcomes 60 References Alger, V. 2014. Teacher Incentive Pay that Works: A Global Survey of Programs that Improve Student Achievement. Barbara Mitchell Centre for Improvement in Education. Ares Abalde, Macarena. 2014. “School Size Policies: A Literature Review,� OECD Education Working Papers, No. 106, OECD Publishing. Assaad, Ragui, Caroline Krafft, and Djavad Salehi-Isfahani. April 2014. “Does the Type of Higher Education Affect Labor Market Outcomes? A Comparison of Egypt and Jordan,� Economic Research Forum Working Paper No. 826. Barro, R. and J. Lee. 2013. “A New Data Set of Educational Attainment in the World, 1950-2010,� Journal of Development Economics 104:184-198. Branch, G., E. Hanushek, E., and S. Rivkin. 2013. “School Leaders Matter: Measuring the impact of effective principals,� Education Next, Vol. 13, No. 1. Connecticut State Department of Education. 2016. Connecticut Administrator Test, retrieved from http://www.sde.ct.gov/sde/cwp/view.asp?a=2641&q=320420 . Darling-Hammond, L. 2010. The Flat World and Education: How America’s Commitment to Equity Will Determine Our Future, Teachers College: New York, NY. El-Kogali, Safaa El Tayeb and Caroline Gould Krafft. 2015. Expanding Opportunities for the Next Generation: Early Childhood Development in the Middle East and North Africa. Directions in Development, Human Development, World Bank Group: Washington, DC. Heejin Kim, Janice and Quentin Wodon. 2015. Public Education Cost Benchmarking for Jordan. World Bank. Iqbal, Farrukh and Youssouf Kiendrebeogo. 2014. Education Attainment in the Middle East and North Africa: Success at a Cost, World Bank Group, Middle East and North Africa Region, Office of the Chief Economist. Leithwood, K. and Jantzi, D. 2009. “A Review of Empirical Evidence about School Size E ffects: A Policy Perspective,� Review of Educational Research, Vol. 79, pp. 464-490. Magno, C. 2013. Comparative Perspectives on International School Leadership: Policy, Preparation, and Practice, Routledge: New York, NY. Manzi, J., K. Strasser, E. San Martín, and D. Contreras. 2008. Quality of Education in Chile. Measurement Center, Catholic University of Chile. Marsh & McLennan. 2013. International Study into Pay and Non-pay Conditions in Schools. Marsh & McLennan Companies. 61 Ministry of Education, Kingdom of Jordan. 2015. Statistical Report for Academic Year 2013/2014, Department of Planning and Educational Research, Unit for Management of Information System. Ministry of Education of Ontario. 2010. Teacher Performance Appraisal: Technical Requirements Manual, retrieved from http://www.edu.gov.on.ca/eng/teacher/pdfs/TPA_Manual_English_september2010l.pdf . Ministry of Finance. July 2014. General Government Finance Bulletin, Studies and Economic Policies Directorate, Volume 16, Number 6. Ministry of Finance. January 2016. General Government Finance Bulletin, Studies and Economic Policies Directorate, Volume 17, Number 12. Ministry of Planning and International Cooperation, Hashemite Kingdom of Jordan. 2015. Jordan Response Plan for the Syria Crisis 2016-2018. Mizala, A. and P. Romaguera. 2005. “Teachers’ Salary Structure and Incentives in Chile,� in Emiliana Vegas (editor) Raising Student Learning in Latin America: The Challenge for the 21st Century, pages 103-150, World Bank. Mizala, A. and B. Schneider. 2014. “Negotiating Education Reform: Teacher Evaluations and Incentives in Chile (1990–2010),� Governance 27, 1 (January): 87–109. OECD. 2009. Evaluating and Rewarding the Quality of Teachers – International Practices. OECD. 2010. PISA 2009 Results: What Makes a School Successful? Resources, Policies, and Practices (Volume IV), PISA, OECD Publishing. OECD. 2011. Establishing a Framework for Evaluation and Teacher Incentives: Considerations for Mexico, OECD Publishing. OECD. 2011. PISA IN FOCUS 7, Private Schools: Who Benefits? OECD. 2013. PISA 2012 Results: Excellence Through Equity – Giving Every Student the Chance to Succeed (Volume II), PISA, OECD Publishing. OECD. 2013. PISA 2012 Results: What Makes Schools Successful? Resources, Policies, and Practices (Volume IV), PISA, OECD Publishing. OECD. 2014. Education at a Glance 2014: OECD Indicators. OECD. 2014. PISA 2012 Results: What Students Know and Can Do – Student Performance in Mathematics, Reading, and Science (Volume I), Revised Edition. Rau, T. and D. Contreras. 2011. Tournament Incentives for Teachers: The Case of Chile, unpublished paper, University of Chile. 62 Steele, J. L., R. J. Murnane, and J.B. Willett. 2009. "Do Financial Incentives Help Low-Performing Schools Attract and Keep Academically Talented Teachers? Evidence from California," NBER Working Paper 14780, Cambridge, MA: National Bureau of Economic Research (NBER). Vegas, Emiliana and Ilana Umansky. 2005. “Improving Teaching and Learning through Effective Incentives,� in Emiliana Vegas (ed.) Incentives to Improve Teaching: Lessons from Latin America, World Bank, pages 1–20. World Bank. 2010. Jordan Teachers SABER Country Report. World Bank. 2014 (spring). Jordan Economic Monitor – Resilience Amid Turmoil. Poverty Reduction and Economic Management Unit, Middle East and North Africa Region. World Bank. 2014. Learning in the Face of Adversity: The UNRWA Education Program for Palestine Refugees. World Bank. 2015. Accountability: The Last Mile on the Route to Quality Service Delivery – Evidence from Jordanian Schools & Primary Health Centers. 63