Pakistan Policy Note—Expanding access to quality education 79573 Pakistan Policy Note 9 Halil Dundar and Huma Waheed Expanding Access to 1 J une 2 0 1 3 Quality Education Although Pakistan has made some progress in is unlikely to meet the Millennium Develop- improving access to education over the past decade, ment Goal of universal primary education by it still faces major challenges in providing access to 2015. quality education at all levels. It has the world’s sec- ond highest out-of-school population (7 million)— Primary and Secondary Education two-thirds of them girls (though enrollment rates vary appreciably between and within provinces). National Education service delivery up to grade 12 has pri- surveys of student learning suggest that achievement marily been the concern of provincial and area is also very low—a sizable share of school leavers do governments. They are responsible for policy not achieve even minimum mastery of mathematics, formulation, sector financing, and implemen- reading, and language, as defined by the govern- tation through their education departments. ment. The main contributing factors include poor The 18th Amendment, passed in 2010, formal- teacher quality and accountability, inadequate and ized this responsibility. When the amendment inefficient funding, and weak management and gov- was approved, Article 25-A—which guarantees ernance. Since the 18th Amendment was passed in the right to free and compulsory education for 2010, the management and financing of education all children ages 5–16 years—was also added. has been decentralized to the provinces, but national Article 25-A requires further legislation by the THE WORLD BANK GROUP SOUTH ASIA REGION standards need to be set and their achievements moni- provincial assemblies to enforce free and com- tored to address disparities in access to quality edu- pulsory education; the process is under way. cation between provinces. The federal government should play this role and coordinate and facilitate the Role of the Private Sector provision of “education for all.� As in other South Asian countries, Pakistan’s private Pakistan ranks 113 of 120 countries in the United sector is emerging as an alternative to public schools, Nations Educational, Scientific and Cultural Orga- even for poor households. Today, nearly a third of nization’s Education for All Education Development primary and secondary students in Pakistan Index.1 It has the world’s second highest out- attend private schools: 31 percent of boys and of-school population (7 million), two-thirds 33 percent of girls ages 6–10 years and 26 per- of them girls (UNESCO 2012). It has gener- cent of boys and 30 percent of girls ages 11–15 ally performed worse than other countries in years (Figure 2). Over 2004/2005–2010/11, the South Asia and other developing countries (at overall share of children ages 6–10 and 11–15 its level of per capita income; Figure 1). Based enrolled in private schools increased 7 and 6 on current trends, the United Nations Devel- percentage points, respectively. In net terms, opment Programme reports that the country virtually all the gain in school participation in Pakistan Policy Note—Expanding access to quality education Figure Primary net enrollment rate, selected developing countries, 2009 1 100 90 80 Percent 70 2 60 50 Pakistan Bangladesh India Indonesia Sri Lanka Mozambique Philippines Thailand Source: UNESCO Institute for Statistics n.d. the country during the period, especially at the channel public financing for low-cost private primary level, was a result of increased private provision. Evidence from Sindh and Punjab school participation. confirms that the growth of such a dynamic school system is reaching more and more low- This increased enrollment was triggered by low income and rural households. It also suggests public financing and poor public service delivery. that student achievement in private schools In responding to the widespread demand for tends to be much higher than in government greater access to and better education, a size- schools—and that the higher achievement is able and rapidly expanding low-cost private obtained far more cost effectively than in gov- schooling system now serves as a viable alterna- ernment schools, possibly arising from a combi- tive to the government school system for even nation of private schools’ lower labor costs and low-income and rural households. market stimulation. The private sector holds real potential for increasing Government Reform Programs access to quality education. To further increase access, the governments of all four provinces To address the major issues in education, all four have set up “education foundations,� which provinces have embarked on sector reform programs Figure Private school enrollment by gender, selected South Asian countries 2 Male Female 60 40 Percent 20 0 Afghanistan Bangladesh Bhutan India Nepal Pakistan Afghanistan Bangladesh Bhutan India Nepal Pakistan 2007/08 2010 2007 2009/10 2009/10 2010/11 2007/08 2010 2007 2009/10 2009/10 2010/11 6–10 years old 11–15 years old Note: Data for Afghanistan are from 2007/08; Bangladesh, 2010; Bhutan, 2007; India, 2009/10; Nepal, 2009/10; and Pakistan 2010/11. Source: World Bank 2013. that aim to increase equitable access to quality educa- were 66 percent, 35 percent, and 23 percent,3 tion. Education reform initiatives address issues or close to those of low-income countries in ranging from merit-based teacher recruitment Sub-Saharan Africa (Figure 3). There are also and mobilization and financing of school man- large inter- and intraprovincial variations—for agement committees to the provision of basic instance, district primary net enrollment rates facilities and free textbooks to schools. Punjab ranged from 14 percent to 90 percent. initiated the reform process nearly a decade ago and is the most advanced in implement- Pakistan’s completion rate for primary education is ing its reforms (Box 1). Balochistan, which lags among the lowest in the world. Less than half the 3 behind the most, is still developing its sector country’s population has completed at least pri- reform program. mary education, making Pakistan unlikely to meet the education Millennium Development Policy Challenges and Issues Goal by 2015. Despite recent achievements in improving access Education participation is inequitable, even at the to education, Pakistan’s education sector performs primary level. Girls, disadvantaged children poorly compared with other South Asian coun- from poor families, rural children, and chil- tries. This undermines its efforts in economic dren from some social groups have very low growth and poverty alleviation. enrollment rates. The child’s age and the household’s socioeconomic status and loca- Low enrollment and major disparities in access tion (urban or rural) appear to matter greatly for participation. The chances of participa- Although Pakistan has made some progress in tion at primary and middle levels increase expanding enrollment since 2000, access to education with age, suggesting late entry into school. remains low at all levels. Major shortfalls persist Children from poor households appear to suf- in school participation.2 In 2010, the estimated fer a large participation disadvantage at all 7 million children still not in primary school levels: only 43  percent of children ages 6–10 accounted for about 10  percent of all out-of- years belonging to the poorest wealth quintiles school children worldwide. Primary, middle, are enrolled. The participation disadvantage and matric net enrollment rates in 2010/11 rises sharply for rural households at middle Box Punjab Education Sector Reform Program 1 In 2003, Punjab launched its Education Sector Reform Program to improve access, quality, and governance in education. Key reforms include: • External monitoring system. This relies on independent collection of data on school performance (such as teacher presence) and ensures that policy makers have reliable data to make informed, transparent, and needs-based decisions. • New teacher recruitment and placement policy. This was introduced to hire new teachers on merit, with preference to local candidates and women. All teacher recruitment is now transparent and has had a major impact on ensuring that teachers with the right qualifications are hired. This has, in turn, helped reduce chronic teacher absenteeism. The government is working on a teacher certification and licensing mechanism to ensure that teachers gain the required competencies. • New textbook policy. The development, printing, and distribution of textbooks are now managed through an open competitive process. • Public-private partnership program. This was established to provide public funding to low-cost private schools. Under this program, the Punjab Education Foundation was restructured as an autonomous but publicly funded institution to support low-cost private schools. In 2012, more than a million students benefited from the program. • Measurement of student learning outcomes. Under this reform, an independent Punjab Examination Commission was established to carry out universal examinations for grades 5 and 8. The program also has tight links with content and quality of teaching and is a strong tool for holding teachers accountable for performance. According to Pakistan Social and Living Standards Measurement Surveys, the primary net enrollment rate increased from 45 percent in 2001/02 to 62 percent in 2006/07 (for girls, from 43 percent to 59 percent; for rural girls, from 38 percent to 55 percent). These gains were helped by improved sector governance, including the establishment of independent bodies and stronger monitoring systems. Pakistan Policy Note—Expanding access to quality education Figure Pakistan net enrollment rates, 2010/11 3 Federal Punjab Sindh Khyber Pakhtunkhwa Balochistan 75 50 Percent 4 25 0 Primary Middle Matric Source: Pakistan Bureau of Statistics 2011. and high education levels: only 36 percent of 3 scored 30 percent on English tests, 29 percent on rural children ages 11–15 years are enrolled Urdu tests, and 38 percent on math tests. Four years in grades 6–10, compared with 56 percent of later, students in grade 3 in the same villages urban children. Although some provinces, scored similarly or slightly lower, at 31 percent, notably Punjab, have made progress in reduc- 27 percent, and 34 percent in the same tests ing the gender gap, female participation in pri- (Andrabi and others 2007). Data from uni- mary and secondary education remains low. At versal testing exercises in Punjab in 2010 and 61 percent, female participation at the primary 2011 suggest that the child’s gender (a general level lags 10 percentage points behind male disadvantage for boys), urban or rural status participation. (a disadvantage for rural children), and dis- trict affect the mean test scores for government Poor learning outcomes school students in grades 5 and 8.5 The above low enrollment and poor learn- International comparisons of student achievement ing outcomes have three main contributing fac- are unavailable because Pakistan has not taken tors, now discussed. part in any international assessments. However, national surveys of student learning outcomes Poor teacher quality and accountability suggest low achievement. A large proportion of school leavers do not achieve minimum mas- Teachers are consistently found to be the most impor- tery of mathematics, reading, and language tant factor in student learning (  Hanushek and (as defined by the national government). The Rivkin 2010; Rivkin, Hanushek, and Kain 2005; ASER4 2011 assessment found that, although Nye, Konstantopoulos, and Hedges 2004; Rock- arithmetic competence was somewhat better in off 2004; Park and Hannum 2001; Sanders and rural Pakistan than in rural India, it was very Rivers 1996). While the quality of education is low in absolute terms (ASER-Pakistan 2011). affected by many factors (teachers, facilities, For instance, only 37 percent of grade 5 stu- textbooks, and so forth), improving teaching is dents in rural Pakistan could divide a three- probably the most effective way to raise school digit number by a single-digit number. By quality. In Pakistan, both teacher effectiveness grade 8, only 72 percent could perform simple and accountability are weak. Key issues include: division. Unlike in rural India, however, recog- • Poor quality of teachers. Teachers across the nition of two-digit numbers in rural Pakistan country have low content knowledge and was widespread by grade 3. weak pedagogical skills. A recent survey showed that only 36 percent of primary A survey of children in rural villages in selected Pun- school teachers in Punjab could explain two- jab districts found that, in 2003, students in grade digit addition, suggesting that they are not competent to teach the curriculum (Word and transfers is the main contributor to the Bank 2013). This is generally attributable to imbalance. Particularly in rural schools, the country’s weak recruitment policies and teachers may be appointed to vacant posi- practices. Two of the larger provinces—Pun- tions, but they are known to use political jab and Sindh—have introduced merit-based pressure to transfer their posts within a teacher recruitment, which could dramati- few months, leaving remote schools under- cally improve teacher quality at entry (Box staffed or sometimes without a teacher 2). Policies and practices in teacher deploy- altogether. Government policies dictating ment are also a serious concern. student–teacher ratios and teaching loads 5 • Weak teacher accountability. Measured by high exist, but enforcement is weak and made absence rates and minimal effort, this is harder by external influences. the main cause of poor learning outcomes even when teachers are sufficiently quali- Inadequate and inefficient fied. A recent survey of primary schools in financing of public education rural Punjab found that 11  percent of the teachers were absent on any given day. Most Although Pakistan is increasing its attention to nonattendance was unexplained, and ill- education, financing has not followed suit. The ness accounted for most explained absences country has one of the lowest rates of public (World Bank 2013). In most parts of the expenditure on education as a share of GDP country, policy dictates that teachers can (1.9 percent; Figure 4). Adequate financing is be dismissed for misconduct, child abuse, a major challenge for the provinces and the absenteeism, and poor performance, while country as a whole. More than 90 percent of teacher absenteeism can theoretically result the education sector’s recurrent expenditure in penalties such as salary deductions—but is related to remuneration. This leaves a very there is no evidence that these strictures are limited budget for spending on critical non- enforced. salary inputs, such as teaching and learning • Few incentives for teacher performance. Govern- materials, repair and maintenance of school ment school teacher salaries are determined infrastructure, and teacher professional devel- principally by teacher grade and tenure, opment. The quality of public education suf- along with allowances and benefits that fers as a result. relate to the characteristics of the location and school where the teacher works. Impor- The education funding mechanism does not induce tantly, government school teacher sala- behavioral changes through incentive structures and ries are not linked to the levels of teacher accountability mechanisms. These could ensure absence, teacher subject knowledge, and stu- that school inputs are used more efficiently dent achievement. and lead to gains in learning. Establishing a • Discrepancies in teacher placement. The place- link between financing and outcomes is essen- ment of teachers shows wide variations: tial for this. some schools have fewer than five students per teacher, and others more than 100. Weak management and governance Some schools are thus understaffed, and others have more teachers than they need. The country’s capacity to plan, manage, and Political interference in teacher placement monitor the education sector is weak. The Box Merit-based teacher recruitment in Punjab and Sindh 2 The recruitment of new teachers in Punjab follows objective, merit-based criteria based largely on academic and professional qualifi- cations of candidates. However, evidence from the province suggests that credentials do not seem to be associated with improved teacher competency and performance. In addition, the authenticity of professional credentials has been brought into question by preservice diplomas and degrees through nonaccredited programs. Sindh has introduced test-based teacher recruitment, which poten- tially is a far better way of judging whether a candidate has the skills to be a good teacher. Pakistan Policy Note—Expanding access to quality education Figure Public expenditure on education, Pakistan, its provinces, and selected countries 4 Punjab Sindh Balochistan Khyber Pakhtunkhwa Pakistan Bangladesh 6 Philippines Turkey India Indonesia Egypt, Arab Rep. Argentina Thailand Malaysia Mexico Brazil Iran, Islamic Rep. South Africa 0 1 2 3 4 5 6 Percent of GDP Note: Provincial expenditures are shown as a percentage of provincial GDPs. These numbers are not strictly comparable because of the different definitions, and it is imprudent to compare provinces with countries. These comparisons should therefore be considered as indicative only. Source: World Bank data. Ministry of Finance data for Pakistan and provinces. management and financing of primary and • Deficiency in coordination and support at the fed- secondary education are the responsibility of eral level for achieving national education goals. provincial governments, which are below par Provincial governments are responsible for in their capacity to gather and report on key delivering primary and secondary educa- performance indicators on time to effectively tion, including policy making, manage- manage and monitor implementation of edu- ment, financing, planning, and monitoring. cation reforms. The key issues in policy mak- However, in principle, the federal govern- ing, management, and monitoring capacity ment has a pivotal role to play in promoting include: national cohesion of the education system, • Inadequate institutional capacity for primary whether through setting universal teacher and secondary education delivery, hampering the standards or in curricula development. implementation of a consistent and effective edu- Yet the federal Ministry of Education and cation policy. Local capacity to deliver educa- Training, established in 2012 following the tion services is especially minimal at district 18th Amendment, is not involved in either and village levels. activity. • Weak monitoring and evaluation systems at all levels of government to guide reforms. Data, Recommendations for Reform Priorities not always easily accessible, are rarely used either by the system to improve education Making learning outcomes an explicit goal of educa- decision making or by parents to demand tion policy. Although Pakistan has made some school accountability. Student assessment progress in improving access to education, systems are inadequate for providing sys- it faces major challenges in providing access tematic and reliable information on student to quality education at all levels. How can learning outcomes. it address the twin challenges of access and quality, given competing demands on a very schools or government elementary teacher edu- limited national budget? Whereas continuing cation colleges. Managed by the Directorate of efforts are needed to increase participation Staff Development under the Punjab School and make education more inclusive, policy Education Department, this field-based support should highlight the importance of improved system serves as a new and promising conduit learning outcomes for all. In fact, learning out- for providing regular, customized teacher sup- comes need to be the central goal of any edu- port and improving teaching performance. cation policy—an “umbrella priority� under which the other priorities fall. Five are identi- Fourth, teachers need to know that if they acquire new 7 fied below. skills and perform well, they will be rewarded—and vice versa. To improve teacher performance, Improve teacher effectiveness incentives that link teacher pay to performance and accountability need to be built into the system. A good exam- ple is the Improver’s Program of the Punjab Developing and implementing policies that focus Education Sector Reform Program. Launch- on teacher performance and management should be ing strong accountability mechanisms (such as the cornerstone of an education system that is effec- penalties for absenteeism) is also important. tive and efficient in providing quality education to all children. Punjab and Sindh have already Improve governance, efficiency, and launched reforms to improve teacher quality quality, using financing as a tool and accountability, focusing on recruitment and management policies. These policies School finance systems must provide the necessary should be implemented throughout the coun- resources so that all students, regardless of back- try and entail four steps. ground, can learn. All provinces and areas need mechanisms that ensure adequate resources First, clear standards need to be set for recruitment, for all schools. The mechanisms should be deployment, transfers, and postings, with strong safe- aimed at meeting the minimum require- guards against decisions not based on merit. Allo- ments for creating and maintaining accept- cation of teaching posts to schools should be able learning conditions. Minimum funding based solely on student–teacher ratio norms standards for schools should be developed and and minimum staffing criteria (for instance, a implemented—and be objective, criteria-based, ­ minimum of two teachers per school to ensure and linked closely with the needs of schools a functional school). (most specifically, the number of students in the schools). Other factors could include needs Second, preservice and in-service training needs relating to school levels, school buildings, avail- to operate on three levels: r aise teacher sub- ability of basic amenities, and so on. ject knowledge, equip teachers with up- to-date approaches to teaching, and help More funding is required to improve resource-use teachers adopt effective pedagogical methods efficiency and learning outcomes. Pakistan could to enhance student learning. consider financing tools that have shown prom- ise in other countries, such as changes in the Third, programs to provide intensive support to incentives structure for both teachers and teachers in low-performance schools need to be schools. Four promising approaches stand out. designed. These programs should regularly pro- First, accountability systems (such as perfor- vide on-site support to teachers to address the mance-related pay and promotion) based on challenges they face in their particular environ- student learning achievements could help mod- ments (a multigrade classroom, for instance). A ify teacher behavior and stimulate more effort good example is the Punjab program to provide in the classroom. Second, modifying school onsite advisory support to teachers through funding formulas has the potential to cre- a network of field-based district teacher edu- ate incentives for quality improvement. Block cators and teacher educators based in high grants that do not demand accountability from Pakistan Policy Note—Expanding access to quality education schools could be replaced by grants that carry Leverage the contribution a range of incentives for efficiency and equity. of the private sector (Box 3 presents the case of Armenia, where this move has been effective.) Third, financial In the face of capacity and resource constraints, lever- powers need to be decentralized closer to the aging the contribution of the private sector is key to school level. School financing is currently man- meeting the challenges of improved access and qual- aged at the tehsil level, with several hundred ity. Since the private sector has already dem- schools (in the larger provinces) under the onstrated that it can offer access at lower cost 8 purview of the drawing and disbursing officer. and with similar or better outcomes than the Fourth, greater use of public–private partner- public sector, countries will gain by helping it ships—Punjab and Sindh already have them— expand, easing barriers to entry, and carefully could also increase resources for education and designing PPPs. Innovative and cost-effective maximize efficiency, as long as efficiency and programs in Sindh and Punjab could be scaled equity incentives are built into the agreements. up (Box 4). Box Ensuring adequacy with per capita funding, Armenia 3 Armenia has undertaken a series of major (but incremental) education finance reforms—including decentralization and the use of per capita funding—to respond to demographic and political changes. After the Soviet Union dissolved in the early 1990s, Armenia’s education expenditures and enrollment declined but its number of teachers rose. The inefficiency in spending was the result of previ- ous school funding methods, based on a simple adjustment of historic levels—bargaining, and discretionary decisions. To increase efficiency, Armenia shifted the responsibility for managing education resources to administrators at the subnational level through decentralization and to schools through school-based management reforms. A per capita formula made funding more transparent. Resources are now allocated per student and adjusted for rural location, level of schooling, and special needs. These structural reforms, piloted in 1999 and implemented throughout the country by 2007, have improved the efficiency of education finance tremendously, increasing class size to appropriate levels and making surplus teachers redundant. Box The promise of public–private partnerships for improving education quality 4 Several public–private partnerships introduced in Pakistan have generated cost-effective gains in participation and achievement. Introduced in 2005, the Foundation-Assisted School program administered by the Punjab Education Foundation provides condi- tional cash subsidies to low-cost private schools to open up private schooling opportunities for children from low-income households and to raise the level of learning. Per student cash subsidies are provided monthly, with essentially no conditions on how they are to be used. The amount is purposely set low (half the estimated per student cost in the public school system) to ensure that only low-cost private schools self-select into the program. In return for the subsidy, the program school has to waive tuition and other fees for all students and ensure that the school achieves a minimum student pass rate in the Quality Assurance Test (QAT). The QAT is a curriculum-based, multisubject test designed by subject specialists and administered by independent testing agencies. Program schools are also eligible for group bonuses for teachers who achieve high QAT pass rates and for competitive bonuses for schools that rank highest in the QAT. Schools that do not achieve a minimum pass rate twice in succession are dropped from the program. As of June 2010, the Foundation-Assisted School program had proceeded through six phases of expansion and supported about 800,000 students in 1,800 schools in 29 of the 36 districts in Punjab. Rigorous evaluation (Barrera-Osorio and Raju 2010, 2011) found that within two years, the program generated major gains in enrollment and school inputs (roughly 40 percent) and student achievement (0.3–0.5 standard deviation). The Punjab Education Foundation runs a sister program, the New School Program, which supports the opening of new schools in underserved communities. The program provides per student subsidies to new private schools in underserved areas, conditional on a school’s achievement on standardized, competency-based tests. The program (early 2013) covers more than 20,000 students in 230 schools in 16 districts. A similar program, Promoting Private Schooling in Rural Sindh, is run by the Sindh government. It, too, attempts to leverage the private sector to deliver schooling to underserved rural communities. Program schools get grants for construction and other support, as well as per student subsidies conditional on maintaining minimum student achievement. A rigorous evaluation found that the program produced substantial gains in participation and achievement (Barrera-Osorio and others 2011). Improve sector coordination and support References Andrabi, Tahir, Jishnu Das, Asim Ijaz Khwaja, The federal government has an important role to Tara Vishwanath, and Tristan Zajonc. play in promoting cohesion of the education sys- 2007. “Pakistan: Learning and Educational tem nationwide, whether by setting universal Achievements in Punjab Schools (LEAPS): teacher standards or by playing a role in cur- Insights to Inform the Educational Policy riculum development. Coordination among Debate.� www.leapsproject.org/assets/ provinces—a key function of the Ministry publications/LEAPS_Report_ExecSummary. of Education and Training—needs to be pdf. 9 strengthened. The ministry should play a key ASER-Pakistan. 2011. The Annual Status of Educa- role in improving communication between tion Report—Pakistan. South Asia Forum for the provinces/areas to facilitate knowledge Education Development and ITA, Lahore, exchange for improving Pakistan’s education Pakistan. indicators. Barrera-Osorio, Felipe, and Dhushyanth Raju. 2010. “Short-Run Learning Dynamics under Enhance learning-assessment systems a Test-Based Accountability System: Evidence from Pakistan.� Policy Research Working Enhanced learning-assessment systems are nec- Paper 5465, World Bank, Washington, DC. essary to monitor progress in learning outcomes ———. 2011. “Evaluating Public Per-student and in improvements in schooling quality over Subsidies to Low-Cost Private Schools: time. These comprehensive assessments need Regression-discontinuity Evidence from Pak- to cover students in both public and private istan.� Policy Research Working Paper 5638, schools. Yet Pakistan’s student assessment World Bank, Washington, DC. systems, which are at the provincial level, are Barrera-Osorio, Felipe, David S. Blakeslee, Mat- weak and do not provide regular information thew Hoover, Leigh L. Linden, and Dhushy- about student performance. Pakistan needs anth Raju. 2011. “Expanding Educational to strengthen these systems to emphasize Opportunities in Remote Parts of the World: classroom testing. It also needs to include Evidence from a RCT of a Public Private large assessments. Partnership in Pakistan.� In Third Institute for the Study of Labor (IZA) Workshop, “Child Notes Labor in Developing Countries,� Mexico City. Given the differences in challenges, oppor- www.iza. org/conference_files/childl2011/ tunities, and priorities by level of education, blakeslee_d6783. this note focuses on primary and secondary Hanushek, Eric A., and Steven G. Rivkin. 2010. education. “Generalizations about Using Value-Added 1. This is a composite index that provides an Measures of Teacher Quality.� American Eco- assessment of a country’s education system nomic Review 100 (2): 267–71. on four Education for All goals: universal Nye, Barbara, Spyros Konstantopoulos, and primary education, adult literacy, quality, Larry V. Hedges. 2004. “How Large Are and gender. Teacher Effects?� Educational Evaluation and 2. All statistics in this paragraph and the next Policy Analysis 26 (3): 237–57. one are staff estimates based on various Pakistan Bureau of Statistics. 2011. Pakistan rounds of the Pakistan Social and Living Social and Living Standards Measurement Standards Measurement Surveys. Survey 2010/11. Government of Pakistan, 3. The net enrollment rate figures are based Islamabad. on the officially recognized age of chil- Park, Albert, and Emily Hannum. 2001. “Do dren for each education level. Teachers Affect Learning in Developing 4. Assessment, Survey, Evaluation, Research. Countries? Evidence from Matched Student- 5. World Bank staff estimates using student Teacher Data from China.� Paper presented test score data from the Punjab Examina- at the Social Science Research Council Con- tion Commission. ference “Rethinking Social Science Research Pakistan Policy Note—Expanding access to quality education on the Developing World in the 21st Cen- of Tennessee Value-Added Research and tury,� Park City, UT, June 7–11. Assessment Center, Knoxville, TN. Rivkin, Steven G., Eric A. Hanushek, and John UNESCO (United Nations Educational, Sci- F. Kain. 2005. “Teachers, Schools and Stu- entific and Cultural Organization). 2012. dent Achievement.� Econometrica 73 (2): Education for All Global Monitoring Report 417–58. 2012—Youth and Skills: Putting Education to Rockoff, Jonah E. 2004. “The Impact of Individ- Work. Paris. ual Teachers on Student Achievement: Evi- UNESCO Institute for Statistics. n.d. Data Cen- 10 dence from Panel Data.� American Economic tre. http://stats.uis.unesco.org/unesco/ Review 94 (2): 247–52. tableviewer/document.aspx?ReportId=143. Sanders, William L., and June C. Rivers. 1996. World Bank. 2013. “Improving School Qual- “Cumulative and Residual Effects of Teach- ity and Learning Outcomes in South Asia�. ers on Future Student Academic Achieve- Draft Sector Report. Human Development ment.� Research Progress Report. University Unit, South Asia Region. Washington, DC. © 2013 The International Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street NW Washington, DC 20433 USA All rights reserved This report was prepared by the staff of the South Asia Region. The findings, interpretations, and conclusions expressed herein are those of the authors and do not necessarily reflect the views of the World Bank’s Board of Executive Directors or the countries they represent. The report was designed, edited, and typeset by Communications Development Incorporated, Washington, DC.