81076 from EVIDENCE to POLICY Learning what works, from the Human Development Network September, 2013 Using Low-Cost Private Schools to Fill the Education Gap An impact evaluation of a program in Pakistan Educating children is a priority across the globe, but develop- go to school, making education a priority for the Sindh govern- ing countries can face enormous challenges. Schools are often ment. Through the International Development Association overcrowded and in disrepair. Teachers don’t always show up or (IDA), the World Bank’s fund for the poorest, the Sindh may not be qualified or interested in teaching. Parents hesitate government received assistance to develop and implement its to send children, especially girls, to schools that aren’t close by or Sindh Education Sector Reform Program to raise enrollment, they may want to keep improve student achievement and reduce social disparities in them at home to help education by improving school performance through more ac- with housework. The countability and better governance. This included a program numbers tell the story: offering cash subsidies to private entrepreneurs to provide free, EDUCATION Worldwide, 57 million co-educational primary schools in villages in remote areas with- children who should be out local schools. To measure the effect, an impact evaluation in primary school are was built into this program. The evaluation found that boys not, despite the global and girls in villages that received program-supported private push for universal pri- schools were more likely to be in school and they did better mary education through on tests than children in villages without such schools. The the United Nations Millennium Development Goals. children also had bigger dreams: Girls in program villages were The World Bank works with governments to develop and more likely to want to be teachers instead of housewives; par- implement innovative methods for expanding access to edu- ents were more likely to want their children to become doctors cation, particularly for girls, and improve school quality. In or engineers instead of farmers. the effort to boost enrollment, raise teaching standards and The World Bank, which is committed to education for strengthen school accountability, policymakers and education all, this year announced support for the second Sindh Edu- experts in developing countries are exploring a variety of ap- cation Reform Program Package, which will help expand proaches, including the use of private and other non-govern- this public-private partnership program and support and mental schools to reach disadvantaged children. In Pakistan’s strengthen the quality of the schools. As this evaluation Sindh province, only about half of primary school age children shows, innovation in education is worth testing. Context Private schools in Pakistan, long catering to children of the perform better, although teachers are paid less and are often country’s elite, have become popular among the poor thanks less educated than their counterparts in government schools.* to the spread of low-cost private schools. More than a third of Pakistani policymakers and education experts are turn- all children are now enrolled in private school, where tuition ing to for-profit education as one route to improve the qual- averages less than $5 a month in rural villages, a small fraction ity of education and expand access for the country’s primary of average household income. Studies of these schools have school age children, more than a quarter of whom aren’t generally found that student learning is higher and teachers in school. While the models differ in details, these public- *See Andrabi, Tahir, Jishnu Das, Asim Ijaz Khwaja. 2008. A dime a day: The possibilities and limits of private schooling in Pakistan. Comparative Education Review 52(3):329–355; Das, Jishnu, Priyanka Pandey, and Tristan Zajonc. 2006. Learning levels and gaps in Pakistan. World Bank Policy Research Working Paper 4067. Washington, D.C.: World Bank. private partnerships generally require that schools waive tu- Foundation, a quasi-governmental agency of the Sindh ition in exchange for receiving a per student subsidy that provincial government, working with a World Bank team. schools use for their costs. The subsidy is set lower than what The project sought to encourage private entrepreneurs to it costs to run a government school, where salaries tend to be build and operate primary schools in underserved villages. much higher and controls are built into the public-private To qualify, schools had to meet minimum facility standards, partnerships to promote educational quality. In Punjab prov- be co-ed, waive tuition fees for all students, and have at least ince, for example, in a program initiated in 2005, existing two female teachers. In return, the private entrepreneurs, low-cost private schools can qualify for a per student subsidy who are vetted by the foundation, receive a per student sub- if they waive tuition and ensure that a minimum percent- sidy. They also get additional non-financial assistance, such age of students pass a standardized academic test offered on as free textbooks and other materials, teacher training, and a regular basis. This program, which has proven successful regular visits by foundation staff to advise on how to im- at raising test scores and student enrollment (see Evidence prove teaching and student learning. to Policy, October 2012), now covers more than one million students in some 2,000 partnering private schools. Hope In Sindh province, education experts also have turned “Education is very important and girls have as much of a right to low-cost private schools to increase access to education to it as boys,” says Saima Kausar, a principal at a program- and improve learning, especially in villages that lacked lo- supported school in Wazir Sangi, the village where she grew up. “I always felt that the girls of my village should learn to read and cal schools. The Promoting Low-Cost Private Schooling in write just as I did.” Rural Sindh program was created by the Sindh Education EDUCATION Evaluation The program started in 10 districts chosen because they part of the program. For the evaluation, 199 of the 263 vil- ranked very low on three school access measures: the num- lages across the three groups were surveyed. ber of children in school, gender equality among children in The schools opened in the summer of 2009, halfway school and the percentage of households less than 15 minutes through the school year. The baseline survey was conduct- from the nearest primary school. The evaluation sought to ed in February 2009, with a follow-up survey in June 2010 test not only the impact of these program-supported private and a second follow-up a year later. The first follow-up was a schools on overall enrollment and learning, but also whether complete census of each village and covered 25,157 children giving some schools a bigger per student subsidy for female age five to 15. The second follow-up covered a representa- students raised enrollment of girls. Two-hundred-and-sixty- tive sample of households with children in each village. In three underserved villages qualified for the first round of the total, 17,721 children age five to 17 were surveyed in the program (since then, the program has expanded). These vil- second follow-up. Researchers surveyed households, and also lages were randomly assigned to one of three groups: villages children, who were given a language and math test at home in which the new private schools received a subsidy of 350 and queried about work they did apart from school, enroll- rupees ($4.70) per student, regardless of gender (100 villag- ment status and their desired professions. Schools were also es), villages where schools received an additional 100 rupees surveyed to gather information on their operations, facility ($1.34) for each female student (100 villages), and 63 control and staff. Teachers were interviewed and attendance lists were villages where private schools were not being established as checked to cross verify reported enrollment and attendance. *This policy note is based on the research paper “Leveraging the Private Sector to Improve Primary School Enrollment: Evidence from a Randomized Controlled Trial in Pakistan,” (May 2013) by Felipe Barrera-Osorio, David S. Blakeslee, Matthew Hoover, Leigh L. Linden, Dhushyanth Raju, Stephen Ryan. Findings Enrollment among children in villages with program- 21 percentage points (66% of questions answered correctly supported private schools jumped by 30 percentage in program villages compared with 45% in control villages). points compared with children in control villages. For those children who had enrolled in school as a result of the program, the impact was a large 63 percentage points. The primary school enrollment rates in Sindh are among the The findings indicate that program-supported private schools lowest in the country. Prior to the launching of the program, 53 were very effective in raising student learning. percent of boys and 41 percent of girls age five to nine were in primary school. In villages targeted by the program, enrollment Program schools not only do a better job teaching— was even lower, with about 30 percent of all children in school. despite having less experienced teachers than One year into the program, which brought private schools government schools—but they provide their to these remote villages, enrollment rose by 50 percentage students with better facilities. points compared with the average enrollment rate in control villages. In other words, more than 80 percent of primary Similar to findings from surveys of low-cost private schools in school age children in program villages were in school, compa- other parts of Pakistan, the schools in Sindh province did a rable to Pakistan’s major cities. In the program’s second year, the better job of keeping the schools staffed and running. In line government started to re-open shuttered schools in some con- with the emphasis the Sindh Education Foundation places on trol villages, raising enrollment there by 20 percentage points. private schools providing quality infrastructure, the schools As a result, while enrollment remained high in program vil- lages, the program’s impact declined to 30 percentage points from the first year’s 50 percentage points. There was also a dramatic jump in girls’ enrollment. The impact on enrollment of girls in program villages was very high. At the end of the first academic year, the impact was 5.2 percentage points higher than that of boys. By the end of the second academic year, the impact on boys and girls was the same in program villages. Giving schools a larger per student subsidy for girls didn’t show an additional impact over giv- ing schools the same per student subsidy for both boys and girls. Based on this finding, the foundation doesn’t offer new schools different subsidy levels. Primary school age children not only were more usually had more classrooms than government schools, and likely to be in school, but they did much better on were more likely to have sufficient desks for students, drink- standardized achievement tests. ing water, electricity and a toilet. The program required that these schools hire at least two Math and language tests were administered to children age female teachers, and the total number of female teachers five to 10 years old in household samples in both program was higher than in government schools. The private school and control villages as part of the second follow-up, roughly female teachers generally had less teaching experience than two academic years after the start of the program. The per- their counterparts in government schools, and they were paid centage of questions answered correctly was calculated for less, but there wasn’t any evidence that this hurt the quality each test-taker. The impact of the program on test scores was of education. Children who go to school develop the skills and lies in the control villages. They were also less likely to knowledge needed to do better in life. It also gives imagine a future for their children where the boys were children—and their parents—the chance to build farmers and the girls were housewives. Families also raised aspirations. The evaluation found that going to a their expectation of the ideal education level by one-and- program-supported private school didn’t just give a-half years. children a better education, but it gave them the The children themselves also wanted more: Girls were chance to dream bigger. less likely to report wanting to be housewives and more likely to envision their future as teachers. Both sexes want- Families in program villages were more likely to want their ed more years of education, compared with those in the children to be doctors or engineers, compared with fami- control villages. Conclusion Low-cost private schools in Pakistan are proving very suc- In urban areas of Punjab, World Bank researchers are plan- cessful at attracting students—boys and girls—and teaching ning to evaluate the impacts of vouchers on enrollment of them effectively for less money than it costs to run a gov- out-of-school slum kids, an initiative of the Punjab govern- ernment school. Some of the lower costs come from hiring ment under its second Education Sector Reform Program, teachers who receive lower salaries than government school supported by the Bank through IDA financing. The vouch- teachers, but this doesn’t appear to be hurting the quality ers will be good for low-cost private schools, which will of education. On the contrary, given the stronger account- have to meet minimum levels of learning to receive voucher ability and greater teaching and learning support offered to children. Another evaluation, supported by the World Bank program-supported private schools in Sindh, students did Strategic Impact Evaluation Fund (SIEF), will measure the substantially better on tests than children whose only op- impacts of a program to encourage improved functioning tion was a government school. of low-cost private schools and expand their use in Pakistan There is still more to learn to create and support pro- through special grants, loans and equity financing, doing grams that expand educational access and improve quality. away with the need to rely on government money. The Human Development Network, part of the World Bank Group, supports and disseminates research evaluating the impact of development projects to help alleviate poverty. The goal is to collect and build empirical evidence that can help governments and development organizations design and implement the most appropriate and effective policies for better educational, health and job opportunities for people in developing countries. For more information about who we are and what we do, go to: http://www.worldbank.org/sief. The Evidence to Policy note series is produced by SIEF with generous support from the British government’s Department for International Development. THE WORLD BANK, HUMAN DEVELOPMENT NETWORK 1818 H STREET, NW WASHINGTON, DC 20433 Produced by Office of the Chief Economist, Human Development Network, Communications/Aliza Marcus