from EVIDENCE to POLICY Learning what works for better programs and policies February 2015 HAITI: Can Non-Public Schools Fill the Gap for Poor Children? Worldwide, some 57 million primary school age children don’t ous issues before implementing such approaches, from how to have the chance to go to school. Many factors hold these children reach the target population to how to ensure schools maintain back: poverty, poor health and overcrowded or faraway facilities. quality standards, among other things. The problem is most The World Bank has been working with countries to meet acute in Sub-Saharan Af- universal primary school enrollment, one of the United Nations rica, where some 30 mil- Millennium Development Goals that missed the 2015 target lion primary school age deadline. In Haiti, where the overwhelming majority of primary children aren’t enrolled, schools are non-public, fee-charging schools, the World Bank but many countries in and other donors helped the government launch the Tuition other regions are also Waiver Program to improve enrollment. This program gives EDUCATION still struggling to reach schools a per student subsidy for new first grade students, re- universal primary school quiring that the school not charge these students tuition. The enrollment. In some places, education experts and policymakers subsidy continues through sixth grade. An impact evaluation are looking to low-cost private schools—whether run by private found that the program is helping families educate their chil- providers, faith-based groups or non-governmental organizations dren for free, while the number of children who have to repeat —to educate more children, especially where the public system grades has dropped. As the Government of Haiti continues is unable to do so. In order to make private schools accessible, to develop its strategy to ensure all children get a primary governments usually set a per student subsidy that is supposed to school education, the impact evaluation results are help- cover tuition. In turn, schools agree not to charge fees to families. ing development experts and policymakers understand the Development experts and policymakers have to consider numer- potentials and limits of such public-private partnerships. Context Haiti’s school system is dominated by the non-public (also In 2006, the Government of Haiti and the World Bank called private) sector, whether for-profit, faith-based or run developed a program to improve school enrollment by giv- by another non-governmental group. In the school census ing fee-charging schools subsidies to cover the costs for poor just over a decade ago, more than 90 percent of all primary students. The decision to turn to these schools made sense: schools were non-public schools, enrolling more than 80 If Haiti was to reach universal education, it couldn’t very percent of all primary school children. The majority of these well do so without the help of 90 percent of the schools. schools weren’t free. The average cost, about $80 a year before The private schools generally weren’t as crowded as public books, uniforms and transportation, put basic education out schools, and had more flexibility to respond to an increase of reach for many children. A survey a few years later showed in demand. Under the program, tuition subsidies covered that around half of all Haitian children of primary school age new students aged six to eight who entered first grade for the weren’t in primary school, while those in schools were much first time. The subsidies would continue through sixth grade older than they should have been, due in part to families cy- and each subsequent group of first graders would qualify for cling their children in and out, depending on finances. the same subsidy. In order to target schools that served poor students, the self-select out of the program. The program was launched in subsidy was set at $90 a year, slightly above the average tuition 2007 in two parts of the country and then further expanded fee at the time. By setting the tuition at this level, schools in the 2008-2009 school year. Currently, the program covers charging higher fees and catering to wealthier students would over 200,000 students in more than 1,100 schools. Evaluation When the program was expanded to five additional “depart- Researchers used the 2011-2012 school census and the ments” for the 2008-2009 school year, financing was limited 2002-2003 school census to match the original applicant and not every school could be included. The Government schools for the evaluation. They identified 64 percent of ini- and the World Bank agreed that the fairest way of selecting tial applicants in the 2011-2012 census and then 55 percent the 547 schools from the 1,034 eligible schools was through of those were identified in the first census. Because baseline random selection. The sample was then further divided so data wasn’t collected when the program was rolled out, re- that at least one school was picked in each commune (Haiti searchers used the 2002-2003 census as the proxy baseline. is divided into 10 administrative departments and each de- The schools were observationally equivalent at baseline, when partment is subdivided into communes.) This helped further considering things like infrastructure, teaching materials, EDUCATION target the program to rural areas. staffing and number of students. Results Enrollment rose in schools that received the tuition age, these schools lost about 16 students, while schools that subsidies… received the subsidy gained an average of 78 students. The figures indicate that some students may have switched to the The number of students in first through fourth grade in- subsidized schools because of the free tuition. The large in- creased, on average, by 88 students across the four grades, crease in enrollment may show that the program is bringing when compared with schools that weren’t in the program. in children who wouldn’t otherwise be in school, but the eval- This increase reflected students who started with the program uation can’t pinpoint whether the increase is because of that in first grade and were now in fourth grade (and still receiv- or because students are switching from fee-paying schools. ing tuition waivers), and the newer students who had since started school under the same tuition waiver program. The …and there was a drop in the percentage of figures were collected for the 2011–2012 period. However, students who were older than they should have the evaluation couldn’t measure what percentage of these been for their grade. additional students would have otherwise attended a public school or would have paid if necessary for private school. The free tuition reduced the financial burden of schooling When all six grades were looked at over time between the on families, which may have made it possible for them to census of 2002–2003 and the census of 201–2012, there was keep children in school more consistently. In Haiti, fami- a small dip in enrollment in the control schools—on aver- lies who have trouble paying school fees will often delay This policy note summarizes Policy Research Working Paper 7175. “Increasing Access by Waiving Tuition: Evidence From Haiti,” Melissa A. Adelman, Peter A. Holland. World Bank. January 2015. http://www-wds.worldbank.org/external/default/WDSContentServer/WDSP/IB/2015/01/22/000158349_20150122104022/Rendered/PDF/WPS7175.pdf sending kids to school, and then move kids in and out of the program, because the government didn’t have the ca- school, depending on what they can afford at the time. This pacity to enforce the rules. Almost all of the participat- means that enrolled children are often older than they should ing schools don’t charge parents any fees, keep class sizes be for the grade they’re in, because when they return to school under 45 students, and have a functioning school com- they will return to the grade they dropped out of or missed. mittee. However, some schools fail to meet at least one of Similarly, children are often made to repeat a grade, because the program’s rules. For example, they provide fewer than of insufficient attendance during the year. the required three textbooks per student, they fail to make In schools that qualified for the tuition subsidies, the per- physical improvements in schools, or they don’t maintain centage of students who were two or more years older than the detailed accounting records. age for their grade dropped by 10 percentage points. Interest- ingly, this decline was seen even among fifth and sixth grade students, who didn’t qualify for the subsidies because they were already in school when the program started. Researchers also suggested that schools in the program may have changed their policies to get more children into school at the right age. Most of the schools that qualified for the tuition waiver program were open and educating students three year later. Based on the later census, 76 percent of schools that had been picked to participate in the program three years earlier were still open. The schools that weren’t eligible for tuition waivers were harder to track. Fifty percent were located through the One concern was that schools would overcrowd the census, while the others couldn’t be found. They may have classrooms to get more money but instead, schools shut down, or they may have moved or changed names, as is responded to increased enrollment by hiring more common among fee-paying, non-religious schools in Haiti. teachers, so class size stayed about the same. In either case, it’s possible that the ability to rely on regular tuition payments—the subsidies the government provided, In participating schools, class sizes stayed about the same even starting with that year’s first grade class—helped schools in though total enrollment increased, as schools responded by the program plan for the future and maintain operations un- hiring on average one more teacher. In schools that didn’t re- der the same name and in the same location. ceive tuition subsidies, relatively small declines in enrollment caused class size to drop. So three years into the program par- Schools were also able to meet most of the program ticipating schools had an average of eight more students per rules related to class size, permits and available class than in control schools, but still below the maximum of textbooks. 45 students per teacher set by the program. The goal of the program was to get more students in pri- The program likely has contributed to the increase mary school without ending up with overcrowded class- in Haiti’s national enrollment rate over the past es where children couldn’t learn. In addition to keeping decade or so. classes to a maximum of 45 students, schools had to give students at least three school textbooks, and they couldn’t Before the program started, national net primary school en- have more than two classes per grade that accepted subsi- rollment (which refers to the number of students who are dized students in order to keep program costs in control enrolled and are at the right age for primary school) in Haiti and spread access across schools. was around 50 to 60 percent for children who were of pri- Most schools were able to meet at least some of the mary school age. Five years into the program in 2012, the requirements, but those that didn’t weren’t forced out of enrollment rose to 70 to 80 percent. Conclusion As we get closer to universal primary education, it be- from sending their kids to school even when it is otherwise comes increasingly harder to reach those last kids still out “free”. In addition, many communities, particularly in rural of school. Creative approaches are going to be needed to areas, may not have a school within a reasonable distance. A succeed in bringing everyone in. Haiti’s unusual educational range of approaches may be needed to directly address these landscape—where very few schools are state-run—required barriers, including building new schools or cash subsidies figuring out a way to work with the non-public sector. The paid directly to families. At the same time, enforcement of decision to offer tuition subsidies allowed the government to school standards is necessary to ensure that going to school build on an already existing network of schools, rather than leads to learning. The program in Haiti is clearly helping building a parallel supply structure. This is something that families educate their younger children for free, which can may be applicable to countries where the state school system reduce the financial burden and allow them to spend money is very weak or schools have been decimated in conflicts and to keep their older children in school too. More evaluations weather shocks. Funding non-state schools through targeted are needed to collect evidence on the impact of these pro- tuition subsidies can make education more accessible to the grams on schools that don’t receive subsidies and to under- poor, and tying the money to standards can ensure that stu- stand how to fine-tune these programs to improve access dents have the opportunity to learn. and quality at the same time. EDUCATION However, tuition waivers alone may not be enough to get all primary school aged children in class. For poor fami- “Only when we have succeeded in seating all the children of this country lies, the additional costs associated with schooling, includ- on school benches will we be in a position to offer all Haitians … the same chance at building a future with respect and dignity.” ing uniforms and transport, as well as opportunity cost in President Michel Martelly, August 2014 Haiti Libre newspaper the form of lost labor, may be sufficiently high to keep them The Strategic Impact Evaluation Fund, 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, STRATEGIC IMPACT EVALUATION FUND 1818 H STREET, NW WASHINGTON, DC 20433 Produced by the Strategic Impact Evaluation Fund Series Editor and Writer: Aliza Marcus