47949 The Role and Impact of Public-Private Partnerships in Education Harry Anthony Patrinos Felipe Barrera-Osorio Juliana Guáqueta The Role and Impact of Public-Private Partnerships in Education The Role and Impact of Public-Private Partnerships in Education Harry Anthony Patrinos Felipe Barrera-Osorio Juliana Guáqueta © 2009 The International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org E-mail: feedback@worldbank.org All rights reserved 1 2 3 4 12 11 10 09 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 volume 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 judgement on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions 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; Internet: 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. ISBN: 978-0-8213-7866-3 eISBN: 978-0-8213-7903-5 DOI: 10.1596/978-0-8213-7866-3 Library of Congress Cataloging-in-Publication Data Patrinos, Harry Anthony. The role and impact of public-private partnerships in education / Harry Anthony Patrinos, Felipe Barrera-Osorio, and Juliana Guáqueta. p. cm. Includes bibliographical references and index. ISBN 978-0-8213-7866-3 (alk. paper) -- ISBN 978-0-8213-7903-5 1. Privatization in education. 2. Public-private sector cooperation. I. Barrera- Osorio, Felipe. II. Guáqueta, Juliana. III. Title. LB2806.36.P39 2009 379.1--dc22 2008054279 Cover art: Katerina Mertikas (www.katerinamertikas.com) Cover design: Bill Pragluski, Critical Stages Contents Foreword ix Acknowledgments xi Abbreviations xiii Introduction 1 Private providers are playing an increasingly important role in education 2 The arguments in favor of public-private partnerships 4 The arguments against public-private partnerships 4 The evidence 5 Spotlight on the Netherlands 8 1 Understanding Public-Private Partnerships in Education 9 Management services 10 Professional services 10 Support services 10 Operational services 11 Education services 11 Facility availability 12 Both facility availability and education services (comprehensive contracting) 13 2 International Experience 15 Background and trends 16 The public-private partnership continuum 16 Public-private partnerships and the funding of existing private schools 18 Public-private partnerships that bring alternative operators into the education system 22 Alternatives to conventional systems of public education 28 Conclusions 28 Spotlight on Chile 30 v vi C O N T E N T S 3 What Do We Know about Public-Private Partnerships in Education? 31 The definition and objectives of public-private partnerships 31 Intermediate effects and final outcomes 33 Empirical evidence 35 Channels of impact 41 Conclusions 41 Spotlight on Education for All in Africa 42 4 Designing a Conducive Environment for Education Contracting 43 Common weaknesses in education policy and regulations 43 Improving education policy and regulations 45 Designing public-private partnerships 53 Conclusions 59 A framework for understanding public-private partnerships in education 59 International experiences of using public-private partnerships to fund existing private schools 61 What do we know about public-private partnerships in education? 64 Improving education policy and regulatory frameworks 65 Implementing education public-private partnerships in developed and developing countries 67 Recommendations 68 The role of the World Bank Group 69 Appendix A: Public-Private Partnerships: Program and Policy Descriptions by Contract Type 71 Appendix B: Methods for Evaluation of Public-Private Partnership Programs and Policies in Basic and Secondary Education 83 References 85 Index 93 Contents vii Boxes 1.1 Sources of capital for the nonprofit provision of education 13 2.1 Private schooling in Haiti 19 2.2 Punjab Education Foundation: Foundation Assisted Schools, Pakistan 21 2.3 The differences between charter schools and vouchers: The case of the United States 25 4.1 Summary of actions that governments can take to promote public-private partnerships in education 43 4.2 Principles for designing and implementing public-private partnerships 44 4.3 Registering a private school in Kenya 45 4.4 Registering a private school in Cameroon 48 4.5 New Zealand's Education Review Office 51 4.6 Private school accreditation in the Philippines 52 4.7 Microcredit facilities for education 58 Figures 1 Financing and provision of services in public-private partnerships 3 2.1 The public-private partnership continuum 16 Tables 1 Growing private enrollment rate in education, 1990 and 2005, selected countries 2 1.1 Types of contracts in education 9 1.2 The range of options for public-private partnerships in infrastructure 12 3.1 Expected effects of different public-private partnerships on four main education objectives 32 3.2 The effects of different types of public-private partnership contracts on education outcomes 35 3.3 Studies of vouchers 37 3.4 Studies of private management 39 3.5 Studies of subsidies 40 Foreword Education is widely believed to be critical for any nation's economic, political, and social development. It is widely believed to help people escape from poverty and participate more fully in society and in the market place. These are a few of the reasons why governments around the world assume the responsibility for providing and financing education, espe- cially basic education. But this responsibility is a large and complex one for any government to meet adequately, which is why it is important for governments to explore diverse ways of financing and providing educational services. This book presents the results of the first phase of a multi-year program to examine the role of public-private partnerships in education. It focuses on contracting models at the pri- mary and secondary education levels. It reviews the conceptual underpinnings for why such partnerships might contribute to achieving a country's education goals, reviews empirical evidence, and offers some guidelines for operations. The next phase of this agenda will focus on international and multi-stakeholder partnerships, including philanthropic initia- tives on the one hand and for-profit activities on the other. The book examines five ways through which public-private contracts can help countries meet education goals. First, public-private partnerships can increase access to good quality education for all, especially for poor children who live in remote, underserved communities and for children in minority populations. Second, lessons for innovative means of financing education can be particularly helpful in post-conflict countries undergoing reconstruction. Third, lessons about what works in terms of public-private partnerships contribute to the development of a more differentiated business model especially for middle-income coun- tries. Fourth, the challenge of meeting the education Millennium Development Goals in less than a decade is a daunting one in the poorest countries. Understanding new partnership arrangements within a broad international aid architecture in education can help bring us closer to those goals. Fifth, some very innovative public-private partnership arrangements are happening in Arab countries, and lessons can be drawn from their experience. Evidence is emerging from evaluations of the impact of projects funded by the World Bank, particularly with support from the Dutch government through the Bank-Nether- lands Partnership Program (BNPP) Trust Fund. These evaluations are expanding knowl- edge about the benefits and the costs of these arrangements through rigorous analytical strategies and unique data from investment projects. In education, the BNPP is supporting evaluations in Kenya, Mexico, Pakistan, and the Philippines of initiatives aimed at reducing service provider absenteeism; giving financial and technical support to the expansion of private schooling for the poor; and funding school facilities, new education technologies, and parental participation. Although much is being learned from these ongoing evalu- ations, they are also raising more and deeper questions, helping policymakers and the development community to explore various ways to meet shared education goals. Elizabeth M. King Director, Education Human Development Network The World Bank ix Acknowledgments We acknowledge the significant contributions made by Juan-Diego Alonso, Shaista Baksh, Tazeen Fasih, John Fielden, Norman LaRocque, Michael Latham, Kevin Macdonald, Stefan Metzger, Emilio Porta, and Yidan Wang. Fiona J. Mackintosh and Bruce Ross-Larson pro- vided excellent editing. The team received very useful feedback, guidance, and supervision from Ruth Kagia and Robin Horn. The peer reviewers, Svava Bjarnason, April Harding, Juan Manuel Moreno, and Alberto Rodriguez, all gave excellent comments. We received early feedback during an authors' workshop in May 2008 from Raja Bentaouet-Kattan, Erik Bloom, Marguerite Clarke, Ariel Fiszbein, Jay Kimmelman, Bruno Laporte, Joy Phumaphi, Periklis Saragiotis, and Emiliana Vegas. In addition, thoughtful comments were received at both the concept and final review stages from Helen Abadzi, Jacob Bregman, Samuel Carlson, Elizabeth King, Mamta Murthi, Halsey Rogers, Adriaan Verspoor, and Hana Yoshimoto. The ongoing program on public-private partnerships is undertaken in collaboration with teams across the World Bank Group and with partners in developing countries. We are also collaborating with CfBT Education Trust, whose experts also work on this topic, and with the World Economic Forum's Global Education Initiative. In addition, we have an international steering committee of experts, comprising current and former country policymakers and representatives of multilateral organizations, such as the United Nations Educational, Scientific, and Cultural Organization (UNESCO). xi Abbreviations AVEC Venezuelan Association of Catholic Schools BRAC Bangladesh Rural Advancement Committee BTL build-transfer-lease CARE Cooperation for Advancement, Rehabilitation, and Education ESC Educational Service Contracting (Philippines) FAPE Fund for Assistance to Private Education FAS Financial Assistance per Child Enrolled Basis Program (Pakistan) ICFES Instituto Colombiano para el Fomento de la Educación Superior ICT information and communication technology IS Independent School (Qatar) IFC International Finance Corporation IMF International Monetary Fund ITA Itara-e-Taleem-o-Agahi (India) NGO nongovernmental organization NRSP National Rural Support Program (Pakistan) OECD Organisation for Economic Co-operation and Development PIP Private School Implementation Partner (Pakistan) PISA Programme for International Student Assessment PPP public-private partnership PR Pakistan Railways PSN Pitagoras Network of Schools (Brazil) SIMCE Sistema de Medición de la Calidad de la Educación (Chile) TIMSS Trends in International Mathematics and Science Study UNCTAD United Nations Conference on Trade and Development UNESCO United Nations Educational, Scientific, and Cultural Organization xiii Introduction Despite recent increases, enrollment rates services besides public finance and public remain low in several developing regions. delivery. Although there are many forms of Efficient and equitable access to education PPPs, including partnerships where private is proving to be elusive to many people. organizations support the education sector Often low-income families, girls, indig- through philanthropic activities and high- enous peoples, and other poor and mar- engagement ventures, this study examines ginalized groups have only limited access PPPs in which the government guides pol- to education. Several Sub-Saharan Afri- icy and provides financing while the private can and South Asian countries have yet to sector delivers education services to stu- achieve universal primary coverage, even dents. In particular, governments contract though enrollment rates across all devel- out private providers to supply a specified oping countries increased from 81 percent service of a defined quantity and quality in 1991 to 86 percent in 2006. The quality at an agreed price for a specific period of of education, as measured by standardized time. These contracts contain rewards and tests, is low and represents a major chal- sanctions for nonperformance and include lenge. The majority of students from those situations in which the private sector shares developing countries who participate in the financial risk in the delivery of public international assessments score poorly, and services. this is the case even in the absence of most This partial definition covers several low-income countries, which tend not to types of contracts, depending on the spe- participate in such assessments. cific services provided. The contracts vary Given market failures and equity con- in their degree of complexity. For educa- cerns, the public sector remains an impor- tion, the services provided can range from tant player in providing education services, the construction, management, or mainte- but making high-quality education accessi- nance of infrastructure (often referred to as ble for all in developing countries requires a private finance initiative) to the provision innovative programs and initiatives in of education services and operations, as in addition to public resources and leadership. voucher schemes or charter schools. There are ways in which the public and pri- Building on previous work, the inter- vate sectors can join together to comple- national literature, the results of recently ment each other's strengths in providing completed and ongoing impact evaluations, education services and helping developing and the World Development Report 2004 countries to meet the Millennium Develop- (World Bank 2003a) framework, this book ment Goals for education and to improve presents a conceptualization of the issues learning outcomes. These public-private related to PPPs, a detailed review of studies partnerships (PPPs) can even be tailored with rigorous evaluations, and guidelines and targeted specifically to meet the needs on how to create successful PPPs in educa- of low-income communities. tion. The World Bank has been involved in The concept of a public-private part- exploring the private sector's participation nership (PPP) recognizes the existence of in the provision of public goods for several alternative options for providing education years (see Bell 1995 for a general overview). 1 2 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION PPPs have been studied in depth in health are rigorous. Nevertheless, it is possible to (World Bank 2003b; Harding 2002) and glean some information about promising in education (Jallade 1973; Blomqvist and approaches from a careful review of the Jimenez 1989; Lockheed and van Eeghen existing studies. 1998; James 1993; LaRocque and Patrinos 2006; World Bank 2006). Recent contribu- Private providers are playing tions to the literature are the proceedings from a conference jointly organized by the an increasingly important role World Bank and Harvard University in 2005 in education (Chakrabarti and Peterson 2008; Patrinos Private participation in education has and Sosale, 2007). Also, the World Bank increased dramatically over the last two held a follow-up international conference decades across the world, serving all types on PPPs in 2007 where six rigorous studies of communities--from high-income to low- of PPPs in education were presented. income families. Although governments This book shows how PPPs can facili- remain the main financiers of education (at tate service delivery and lead to additional least of primary and secondary education), financing for the education sector as well in many countries private agents deliver a as expand equitable access and improve sizable share of education (table 1). A num- learning outcomes. It goes on to discuss ber of governments contract with the pri- the best way to set up these arrangements. vate sector to provide some of the services A wide range of education contracting involved in producing education, such as models exists, and all of them have the teacher training, management, or curricu- potential to improve the education system. lum design. Other governments contract However, few existing programs have been with a private organization to manage and evaluated, and too few of these evaluations operate a public school, as is the case with charter and concession schools. Still other contracts require private organizations to Table 1 Growing private enrollment rate in education, 1990 and 2005, selected countries provide education to a specific group of stu- Primary % Secondary % dents by means of a subsidy, a contract, or a Country 1990 2005 % Change 1990 2005 % Change voucher. In the most common type of PPP, Benin 3 12 300 8 25 213 the government provides subsidies to exist- Brazil 14 10 ­29 35 12 ­66 ing private schools or to fund student places. Bulgaria 0 0 0 0 1 100 The continuum of the extent to which coun- tries are using PPPs ranges from those in Chile 39 51 31 49 52 6 which education is provided only by the Colombia 15 19 27 39 24 ­38 public sector to those in which it is largely Indiaa 10 20 100 10 23 130 publicly funded and privately provided. Indonesia 18 17 ­6 49 44 ­10 Some countries make a sharp distinc- Jordan 23 30 30 6 16 167 tion between the role of the public sector as Netherlands 69 69 0 83 83 0 education financier and that of the private Pakistanb 25 27 8 24 25 4 sector as education provider. For instance, Peru 13 16 23 15 22 47 in the Netherlands, all education is publicly South Africa 1 2 100 2 3 50 financed, including private schools, which Thailand 10 16 60 16 13 ­19 enroll more than two-thirds of all students. Togo 25 42 68 17 28 65 In other countries, the private sector plays Tunisia 1 1 0 12 5 ­58 an important role in providing education, but the government only subsidizes some of Ukraine 0 0 0 0 0 0 the students who attend private schools (for United States 10 10 0 10 9 ­10 example, Chile). Several African countries Sources: Kingdon 2007; www.uis.unesco.org; www.worldbank.org/education/edstats; www.oecd.org. have different types of nonpublic schools, Note: Compatibility across countries is limited because of different definitions of education expenditure. However, compatibility within each country across years is ensured. Most recent data available within two including government-subsidized indepen- years of the year indicated. dent schools (for example, the Gambia), a. Rural, based on household surveys. b. Based on household surveys. partially subsidized mission or religious Introduction 3 schools (for example, Lesotho), and at least Figure 1 Financing and provision of services in public-private partnerships partially subsidized community-organized Provision schools (for example, Kenya). Elsewhere, Private Public some countries have public schools that are · Private schools · User fees supported financially by the private sector · Private universities · Student loans (for example, Pakistan). Overall, the private Private sector's participation at the primary school · Home schooling level has grown more than its participation · Tutoring at the secondary level, but there is signifi- Finance · Vouchers · Public schools cant variation across countries. While over- · Contract schools · Public universities all private participation is typically higher Public · Charter schools at the secondary level, private participation · Contracting out at all levels continues to grow. One way to categorize the types of PPPs is to separate Source: Adapted from World Bank 2006. financing from provision (figure 1). The governments of many developed 1991 and 2004, while enrollment in public countries have found a range of different primary schools grew by only 10 percent. ways to leverage the capacity and exper- Globally, there are approximately 113 mil- tise of the private sector to provide educa- lion students in nongovernment schools; 51 tion. In a subset of OECD (Organisation million are at the secondary level. for Economic Co-operation and Develop- Public-private partnerships are also being ment) countries, more than one-fifth of used to build school infrastructure. PPPs are public expenditure is transferred to private a useful way to increase the funding avail- institutions, either directly or by subsi- able for constructing or upgrading school dizing households to pay for the school of buildings and often yield better value for their choice. Moreover, on average, OECD money than traditional public sector invest- countries spend 12 percent of their educa- ments. In such partnerships, the govern- tion budgets in education institutions that ment usually contracts a private company to are privately managed. These governments build and/or maintain school buildings on a have financed a wide variety of schools on long-term basis, typically 25 to 30 years. In a per pupil basis to meet demand for differ- this type of PPP, the private sector supplier ent kinds of schooling. In the United States, assumes responsibility for the risk inherent the number of private companies provid- in the ownership and efficient operation of ing supplemental academic services (aca- the project's facilities. This method of financ- demic tutoring) increased by 90 percent in ing school buildings is used in many OECD just one year, between 2003 and 2004. This countries but most extensively in the United sharp increase was partly driven by the 45 Kingdom. In recent years, several develop- percent increase in federal funds allocated ing countries have also tried this approach, to supplemental education between 2001 though it is too early to see results. and 2005. Private education providers are also In several developing countries, govern- playing an increasingly important role in ments subsidize private schools, mostly delivering education to low-income fami- operated by faith-based nonprofit organi- lies. They include a range of school opera- zations, by financing either school inputs, tors including faith-based organizations, such as teacher salaries and textbooks, or local communities, nongovernmental orga- per pupil grants. Although schools man- nizations (NGOs), and private for-profit aged by faith-based organizations and local and not-for-profit schools. Some African communities are often not considered to be and South Asian countries, where demand strictly private, in this book the term "pri- exceeds the supply of school places and vate" encompasses the whole range of non- public funds are limited, have experienced government providers of education services. growth in the number of private low-cost Across the world, enrollment in private pri- schools that cater to low-income students, mary schools grew by 58 percent between mostly at the secondary level. 4 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION The arguments in favor and choice and expand access to education of public-private partnerships services, particularly for households that The theoretical literature on the topic sug- tend to be poorly served by traditional deliv- gests four positive outcomes of the private ery methods. PPPs also allow governments provision of public services: to take advantage of the specialized skills offered by certain private organizations and · PPPs can create competition in the educa- to overcome operating restrictions such as tion market. The private sector can com- inflexible salary scales and work rules that pete for students with the public sector. In may prevail in the public sector. turn, the public sector has an incentive to Another advantage is that governments react to this competition by increasing the can contract out to the private sector in a quality of the education that it provides. range of initiatives that can include everything · PPP contracts can be more flexible than from nonacademic activities such as food most public sector arrangements. Gener- services and management contracts involv- ally, the public sector has less autonomy ing a few schools, to subsidizing the tuition in hiring teachers and organizing schools at private schools for hundreds of thousands than the private sector does. Public-pri- of students, to long-term, multimillion dollar vate contracts can be a better fit between infrastructure partnerships. For policymak- the supply of and demand for education. ers, contracting is a middle ground between Flexibility in teacher contracting is one government delivery and outright privatiza- of the primary motivations for PPPs. tion and does not attract as much controversy · Governments can choose private provid- and criticisms as privatization. Contracting ers in PPP contracts by means of an open can also enable governments to target initia- bidding process in which the government tives towards particular groups in society or defines specific requirements for the quality to achieve specific outcomes. In addition, it is of education that it demands from the con- a way to bring the private sector's skills and tractor. The contracts often include mea- resources into the education sector (as is the surable outcomes and clauses that specify case of capital investments for school con- the condition to deliver a certain quality struction under private finance initiatives) of education, and the contractor with the and to increase efficiency and innovation in best or lowest cost proposal is then cho- the delivery of education. Contracting can do sen. This one characteristic of the contract all of this while allowing governments to keep alone can raise the quality of education. schools accountable. · PPP contracts can achieve an increased The arguments against level of risk-sharing between the govern- public-private partnerships ment and the private sector. This risk- sharing is likely to increase efficiency in There is a body of literature that argues that the delivery of services and, consequently, there are negative outcomes associated with to induce the channeling of additional the private provision of public services: resources to the provision for education. · PPPs will lead to the privatization of edu- So increasing the private sector's role in cation and thus will reduce the govern- education can have several potential advan- ment's control over a public service. tages over the traditional public delivery · Increasing the educational choices avail- of education. Whether these benefits are able to students and their families may actually realized depends greatly on how increase socioeconomic segregation if well designed the partnership between the better prepared students end up self- public and private sector is, on the regula- selecting into high-quality schools, thus tory framework of the country, and on the further improving their outcomes. capacity of the government to oversee and · PPPs will lead to poorer students being enforce its contracts and partnerships with left behind in the deteriorating public the private sector. When a PPP is imple- schools that lose the support of more mented correctly, it can increase efficiency educated parents. Introduction 5 PPPs may face resistance from certain In countries where PPPs have not been stakeholders. For instance, teachers and extensively tried before, the government other employees may see PPPs as a threat may need to change its education policies to their job stability, while teachers' and and regulatory framework. The government public sector unions may see them as a way must clearly create an enabling framework of diminishing their influence over their that includes: members' terms and conditions of service. · defining the place of private providers in Policymakers need to take these points of the national education strategy; view into account when designing their contracting initiatives. They should con- · setting clear, objective, and streamlined sult with stakeholders and share the con- criteria that the private sector must tract documentation with them. It may meet in order to establish and operate also be useful for policymakers to recruit schools; leading figures in the politics and business · introducing school funding systems communities who understand the potential that integrate public and private schools benefits of PPPs and can use their influence and that are neutral, responsive, and to help to overcome any resistance. targeted; There can also be some challenges and · establishing an effective quality assur- risks involved in PPPs. Inputs to educa- ance system. tion, processes, and outputs are very dif- Good design cannot ensure the suc- ferent and require several different forms cess of a PPP in education as it must also of contracts (including management, sup- be implemented effectively and efficiently. port, professional, operational, educational To ensure this, governments should choose services, and infrastructure). All of these their private partners by means of a trans- variations need to be assessed separately as parent, competitive, and multi-stage selec- they require different approaches in order tion process. Second, they should assign the to be effective. For example, in many coun- roles of purchaser and provider of educa- tries, it is likely that the capacity of public tion services to different entities within the agencies will have to be developed before education administrative agencies. Third, it will be possible to expand the schooling they must ascertain that the private agency options available to low-income students. in question has sufficient capacity for the In some cases, there may even be a need to task at hand. Also, government education build the capacity of private operators to institutions must develop their own capac- deliver high-quality schooling. ity, establish quality assurance mecha- While one advantage of PPPs is that they nisms, develop appropriate performance can be a more cost-effective way to provide measures for contractors, and devise incen- education than the tradition public sector tives to achieve performance targets as well approach, there are some instances in which as sanctions for nonperformance. this may not be the case. For example, con- tracting for facility availability may be more expensive than traditional procurement The evidence methods when the costs of awarding and The existing evidence from around the managing contracts or of private borrowing world shows that the correlation between are particularly high. Also, if poorly han- private provision of education and indica- dled, contracting can even reduce already tors of education quality is positive, which low levels of government accountability suggests that the private sector can deliver and control (Kingdon 2007). It can also high-quality education at a low cost. create opportunities for corruption in the Using data from the OECD's Programme awarding of the contracts. Therefore, part- for International Student Assessment, nerships that provide financing to private Woessmann (2005) showed that publicly schools but do not demand accountability operated schools deliver lower test scores can have negative consequences (Kingdon than privately operated schools, but pub- 2007). licly funded private schools are associated 6 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION with higher academic achievement than Bangladesh. BRAC schools teach the same publicly operated institutions. Therefore, competencies as government schools, but partnerships in which the private sector is they enroll and retain a higher proportion the operator and the public sector is the of hard-to-reach children, such as girls, who financier have the potential to increase constitute 65 percent of students. There was enrollment while keeping the education a boom in the creation of private schools budget in check. in Pakistan between 2000 and 2005, with Also, although more rigorous evidence 15,000 new private schools being set up. is needed, it is clear that PPPs, contract- This increase happened to an equal extent ing, and subsidy arrangements can rapidly in both urban and rural areas and reached expand access to schooling and increase both low- and high-income households its quality, especially if coupled with rig- (Andrabi et al. 2008). The enrollment rate orous quality assurance mechanisms and in private schools of children from the poor- such interventions as teacher training and est households in rural areas jumped from school improvement initiatives. In doing 0 percent to 6 percent. The private schools so, it particularly benefits marginalized charge very low fees, less than 10 cents a day groups and the poor who are ill served by (Andrabi et al. 2006). In this way, private traditionally delivered public services. Pri- provision has increased enrollment in rural vate school contracting programs and pro- areas and among low-income households at grams involving the private management a very low cost. These examples show that, of public schools can provide the poor with when implemented correctly, PPPs can low-cost or free access to education. In fact, help countries to satisfy unmet demand for these contracting initiatives are usually schooling. aimed directly at the poor, including the With regard to the effects of charter schools run by Fe y Alegría, a Jesuit order schools, some useful lessons have emerged that provides education in remote rural from a small set of empirical studies. Based areas, under contract to the governments on evidence from Colombia and Venezuela, of several Latin American countries. it is known that the private management Strategic use of the private sector has of public schools has a positive impact on led to the rapid expansion of access to edu- student test scores. However, we know less cation in several countries. Senegal and about precisely which characteristics of Tanzania deregulated the secondary edu- charter and concession schools (publicly cation sector at a relatively low cost and funded, privately operated schools) make a positive correlation with enrollment. them perform better than public schools, Colombia's targeted voucher program pro- other than perhaps fewer civil service con- vided places in private secondary schools straints, more school autonomy, and the for more than 100,000 students from poor increased length of the school year. None- families. Several rigorous evaluations have theless, it seems from existing evaluations shown the program to be a success (Angrist that flexibility in the contract is an impor- et al. 2002; Angrist, Bettinger, and Kremer tant factor in determining positive educa- 2006). Voucher students were more likely tion outcomes. to pass college entrance exams, had higher As for vouchers, they are associated graduation rates, and scored better on stan- with much controversy. In several coun- dardized tests. The program cost less than tries, governments allow parents to send public secondary schools on a per pupil their children to the school of their choice, basis. In Bangladesh, BRAC's (Bangladesh fund private and religious schools from Rural Advancement Committee) Non-For- the public budget, and allocate resources mal Primary Education Program started in to schools based on enrollment. These 1985 with 22 one-room schools. By 2007, types of programs deliver similar benefits it was serving more than 1.5 million chil- to those offered in voucher programs. dren in more than 20,000 pre-primary and Some of these arrangements are over 100 32,000 primary schools, which accounted years old (such as those in Denmark and for 11 percent of primary school children in the Netherlands) while others are more Introduction 7 recent (such as those in Chile and Sweden). studies that have been carried out so far Colombia's targeted voucher program has suggest that contracting out to the private been subject to extensive analysis because sector can have several benefits, including of its randomized design. These evalu- greater efficiency, increased choice, and ations have shown that the program is wider access to education, particularly for well targeted, effective, and efficient. The those households who have been poorly evidence from Chile's voucher program served by traditional methods of providing is mixed and controversial. Some studies education. In general, private management have found that it has had several positive of public schools tends to be efficient and outcomes, but other studies have chal- yield higher test scores than public institu- lenged this, arguing that the original stud- tions when students reach the end of basic ies had problems of selection and a lack of education. In addition, despite being con- adequate instruments. Furthermore, for troversial, vouchers can improve academic many years following the voucher reform outcomes, especially for the poor. of 1981, overall education quality in Chile However, few of the existing empirical did not improve (Hsieh and Urquiola studies of PPPs can be considered to have 2006). More recently, there have been yielded robust conclusions. There is a need some rapid increases in test scores and an to evaluate how PPPs work most effectively ongoing revision of the school financing in different contexts, particularly where con- formula as an attempt to reduce equity tracting models need to be improved or fine- concerns. Universal school choice (where tuned and in countries where partnerships all parents in a country can choose their are still nascent. While much is known about children's schools by means of a voucher) funding school choice, much less is known in Europe has led to a more competitive about which characteristics of charter and schools market. In most cases, this com- concession schools make them perform bet- petition yields better outcomes overall, as ter than public schools. More research is also would be predicted by theory. Neverthe- needed on universal versus targeted school less, there is much that we still need to choice and on private finance initiatives. learn about school choice and vouchers. These programs should be piloted and rigor- Some of the evidence of the impact of ously evaluated in different settings. Because public provision of private services on edu- of the pressing need to increase the evidence cation outcomes, including measures of base in these areas, this study provides guid- student achievement, is positive but is not ance on how to carry out better evaluations enough to justify either ignoring PPPs or of a variety of aspects of public-private part- expanding them on a large scale. The few nerships in education. spotlight Spotlight on the Netherlands The Netherlands provides a model of school choice that delivers access and quality education; an example of the potential of public-private partnerships in education. O ne of the key features of the exercises ultimate control over both Public and private secondary school Dutch education system is free- public and private schools. Students enrollment in the Netherlands (%) 100 dom of education--freedom to from the Netherlands do exceptionally establish schools, determine the prin- well on international academic achieve- 90 Private ciples on which the school is based, and ment tests such as the Third Interna- 80 organize classroom teaching. In fact, tional Mathematics and Science Study 70 the Netherlands has one of the old- (TIMMS). The Netherlands scored near 60 est national systems based on school the top in reading and math in 2003 50 choice in the world. Although all and was the top performer in mathe- 40 schools in the Netherlands are govern- matics and science achievement for the Public 30 ment funded, most are administered by final years of secondary school in 1995. 20 private school boards. As a result, most The country achieves high scores even 10 children in the Netherlands attend after controlling for national income 0 private schools, a trend that has been and expenditure per student. Thus, the 1850 1900 1950 2000 increasing over the past 150 years. Par- system is not only successful academi- Source: Patrinos 2002. ents can choose among several schools, cally but is also cost effective, yielding and school choice is often promoted good results at relatively low cost. Pre- credence to the arguments of the pro- by the government as a way to increase vious research has found that religious ponents of school choice. However, the competition in the school system. Effi- schools perform slightly better than question remains whether these out- ciency increases as public and private public schools in academic achieve- comes can be expected in other coun- schools try to improve their outcomes ment. More recent research has shown tries or whether the Netherlands is to develop a good reputation and thus that the substantial degree of competi- unique. If they can be generalized, what attract more students. tion in the system is one determinant can other countries do to promote aca- In the Dutch education system, edu- of its high academic achievement rates. demic achievement and to ensure they cation policy is determined centrally Thus, a large school choice system can are accessing all available resources, but the administration and manage- promote efficiency and equity without both private and public? ment of schools is decentralized at the necessarily leading to privatization or Sources: Netherlands Ministry of Education school level. The central government reduced public scrutiny. All this lends 2002; James 1984; Justesen 2002. Understanding Public-Private Partnerships in Education 1 The main rationale for developing public- governments buy the services involved private partnerships (PPPs) in education in producing education (inputs), such as is to maximize the potential for expand- teacher training, management, curriculum ing equitable access to schooling and for design, or the use of a school facility from c h a p t e r improving education outcomes, especially private organizations (Savas 2000). Other for marginalized groups. In this chapter, we governments contract with private organi- show how different types of contracts can zations to provide the process of education, help meet these two objectives in different for example, by managing and operating socioeconomic and political contexts. Spe- public schools. Some other governments cifically, we examine how contracts are used contract with private organizations to pro- to hold all partners accountable and how vide education to specific students (thus, contracts are designed to produce measur- buying outputs). The challenges and poten- able improvements in education outcomes tial benefits of contracting for services that or performance. The analysis considers are inputs, processes, or outputs are very contracting as a distinct instrument from different and are thus discussed separately. any other education accountability mecha- This chapter discusses each type of edu- nisms. We define contracting as the process cation service: (i) management services, (ii) whereby a government procures education professional services, (iii) support services, or education-related services of a defined (iv) operational services, (v) education quantity and quality at an agreed price services, (vi) facility availability, and (vii) from a specific provider. The agreement facility availability and education services between the funder and the service pro- combined. vider is recorded in a contract and is valid for a specified period of time (Taylor 2003; Wang 2000). Table 1.1 Types of contracts in education The World Development Report 2004 What governments contract for What governments buy (World Bank 2003a) concluded that ser- Management, professional, support · School management (financial and human vices can be provided to poor people most services (input) resources management) successfully when citizens, service provid- · Support services (meals and transportation) ers, and governments are accountable to · Professional services (teacher training, curriculum each other. Contracts can improve service design, textbook delivery, quality assurance, and delivery by clearly assigning responsibilities supplemental services) among these actors, identifying objectives Operational services (process) · The education of students, financial and human and outputs, gathering information on the resources management, professional services, and performance and progress of the contrac- building maintenance tor, and ensuring the enforceability of the Education services (outputs) · Student places in private schools (by contracting provisions of the contract. with schools to enroll specific students) Many forms of contracting are currently Facility availability (inputs) · Infrastructure and building maintenance used in education around the world. A Facility availability and education · Infrastructure combined with services (operational range of different services can be procured services (both inputs and outputs) or educational outputs) from the private sector (table 1.1). Some Source: Adapted from World Bank 2006. 9 10 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION Management services critical when buying input services. Simple Weak management is an important con- input services are relatively easy to specify straint to improving public school perfor- in contractual terms, and the performance mance. To deal with this problem, some of contractors can also be conveniently governments have brought in private orga- monitored. In other words, the quality of nizations to manage either a single public service can be specified in the contract and school or an entire public school district. sanctions included if the contractor fails to The responsibilities that the contractor provide that level of quality. In addition, assumes under these contracts usually fall competitive pressures can give providers an into four categories: financial management, incentive to maximize their performance staff management, long-term planning, because contract cancellation is a credible and leadership. Within these contracts, threat as there are plenty of providers of all nonmanagerial personnel continue to input services in most countries. Because be public sector employees. Management an organization can be contracted to deliver contracts have several potential benefits input services to many different schools, for public education, including bringing economies of scale can be achieved. The in professional skills and new ideas from benefits that come from specialization-- the private sector, giving managers the reduced costs and better quality--can also freedom to manage, reducing the bureau- be easily achieved. cratic and union constraints associated If public sector staff originally provided with public service employment, promot- the services, then the handover to the con- ing competition among organizations bid- tractors can be difficult. The early phases ding to win the management contract, and of contracting can be daunting for officials enabling education authorities to specify who may be unfamiliar with the process performance requirements so that they and who may lack the know-how to estab- can change contractors if performance is lish contracts. However, while it can be unsatisfactory. challenging to gain political acceptance for PPPs in the area of management services contracting out support and professional can work, but these services are inherently services, this move usually yields demon- more difficult to contract out than some strably positive results, including cost sav- other services. Specifying and monitoring ings, quality improvements, and more time the performance of managers, as distinct for school officials to devote to education from the organization overall, is diffi- (World Bank 2006). cult. Because many factors contribute to school performance besides the quality of Support services management, it would be inappropriate to Noninstructional activities, including attribute changes in school performance building maintenance, pupil transpor- simply to the effects of the management tation, and school meals, are often very contract. In most countries, the gains from costly for public schools. In the few cases contracting out input services have built where good cost analyses have been done, up over time as the governments gradu- these services have often been found to ally become better at creating these kinds cost significantly more in public schools of contracts. than in private schools (World Bank 2006). The proportion of nonteaching school staff Professional services is often high in public schools, and salary Contracting out professional services such studies in several countries have found that as teacher training, textbook delivery, cur- the wages of support staff are higher in riculum design, quality certification, and public schools than for similar jobs in the supplemental services is straightforward private sector. In response, policymakers in and usually successful. Its main advantage many countries have expanded the extent is that it brings private providers' expertise to which they contract out support services to bear on improving public education. The to increase cost-effectiveness and free up content and oversight of contracts are both the time and resources of school staff and Understanding Public-Private Partnerships in Education 11 education officials so that they can focus on populations (World Bank 2006). Also, this the learning process. Usually, one contract type of contract can be targeted to disad- is tendered to cover multiple schools so that vantaged populations (Barrera-Osorio the contracts are large enough to attract 2007). many bidders. Contracting out support services enables Education services the education sector to take advantage of Instead of engaging a private organization the expertise and the efficient organiza- to operate a public school, some govern- tion of private companies with expertise ments contract out the enrollment of stu- in specific activities, and of the economies dents in private schools, thus, in essence, of scale that result when the same contrac- buying outputs. By paying for students to tor provides services for many schools. It enroll in existing schools, governments can also allows school staff to concentrate on quickly expand access without incurring teaching. Also, in those countries where any up-front expenditure on constructing public sector staff is paid high wages as a and equipping new schools. Other govern- result of belonging to strong unions, there ments contract out students' enrollment in is a cost saving associated with the contrac- specialized services that are not available tor being able to hire nonunionized labor in the public sector. Thus, the concept of (World Bank 2006). Some contracting out contracting out education services involves of support services is done in virtually every using public funds to underwrite individual public education system in the world. For student enrollment in existing schools. This example, public school authorities hardly type of contract can be targeted to specific ever run food services in schools in devel- students and groups, such as low-income, oped countries. disadvantaged, or "problem" students. Contracting for education services Operational services also makes it possible to leverage private In some countries, the education authori- schools' investments in their school capital ties contract private organizations to handle assets by sending publicly funded students a wider range of responsibilities, in essence, to these schools. As a result, the publicly to operate an entire public school. In these funded students receive a higher quality operational contracts, private organiza- education than if the cost of their education tions not only manage the school but staff had been restricted simply to the amount of it as well (World Bank 2006). The aim of public funding spent on them. Also, if the such contracts is most often to free schools contracted schools are willing to subsidize from public service constraints or to give publicly funded students from the fees paid schools more autonomy and to improve by their paying students (as many nonprofit the oversight of the school by tapping into schools do), this form of contracting allows the interest and knowledge of parents and publicly funded students to benefit from other community members. In many cases the higher fees paid by privately funded where schools are allowed to govern them- students (World Bank 2006). selves, communities also contribute to the This type of contract enhances account- construction, upkeep, or improvement of ability in two ways. First, schools are facilities (either in-kind or financially). subject to competitive pressures because Sometimes education authorities initiate parents and students are able to choose a contracting arrangement in response to from among public and private schools. demand from a community organization or Second, in some cases school operators a nonprofit education organization (World are selected through competitive processes Bank 2003a). that give schools an incentive to improve Operational services contracting is usu- their services. Moreover, accountability is ally tried in problem areas, making it a assured by pre-existing school governance viable mechanism for improving schools and oversight arrangements, such as school with performance problems and for ensur- boards, boards of trustees, and parent com- ing service delivery to "hard-to-reach" mittees (World Bank 2006). 12 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION Facility availability contract involves minimizing the risk of In many countries, governments have the government defaulting and making the managed to mobilize private investment investment safer and, hence, more appeal- to finance needed capital stock in utili- ing to private investors. ties and other public services. Contracting The scope of the responsibilities taken out the provision of facilities is appealing on by the private sector varies by contract, because it relieves governments of having and similar arrangements often have dif- to finance capital investments up-front and ferent names. For example, build-operate- all at once. In the education and health sec- transfer arrangements are often referred to tors, the government is often the major or as design-build-finance-operate (table 1.2). only purchaser of services for the new facil- Under build-operate-transfer, which is the ity, which puts an important burden on the most common type of arrangement, the public purse all at once. In these cases, con- private sector finances, designs, constructs, tracting out the financing and construction and operates a public school facility under of facilities to the private sector allows the a contract with the government for a given government to pay for these capital invest- period of time (for example, 25 to 30 years). ments over time by making periodic pay- At the end of the contract period, the own- ments over the term of the contract. ership of the school facility is transferred to The value of the capital investment is the government. determined completely by the govern- Although arrangements can differ ment's payments. This reliance on a single widely, infrastructure-focused PPPs share customer, subject to changing political and several characteristics: policy priorities, makes investing in social · The private consortium is selected through service facilities extremely risky for private a competitive tender process. investors (World Bank 2006). As a result, · Private sector partners invest in school contracting private institutions to finance infrastructure and provide related services and build schools is much more challeng- (for example, building maintenance). ing than other types of contracting. There- The government retains the responsibil- fore, much of the process and content of the · ity for delivering core services such as teaching. · Arrangements between the government Table 1.2 The range of options for public-private partnerships in infrastructure and the private partner are governed by Type of partnership Features long-term contracts (usually 25 to 30 Traditional design The government contracts with a private partner to design and build years) that specify the services the pri- and build a facility to specific requirements. vate contractor must deliver and the Operations and The government contracts with a private partner to operate a standards that it must meet. maintenance publicly owned facility. · In service contracts, the private organi- Turnkey operation The government provides financing, the private partner designs, constructs, and operates facility for a specified time period, while zation often takes on several functions the public partner retains ownership of facility. such as design, building, maintenance, Lease-purchase The private partner leases a facility to the government for a and employment of some nonteaching specified time period, after which ownership is vested with staff. government. · Payments under the contract are contin- Lease or own- The private partner leases or buys a facility from the government gent on the private operator successfully develop-operate and develops and operates the facility under contract to the government for a specified time period. delivering services of an agreed perfor- mance standard. Build-operate-transfer The private partner obtains an exclusive contract to finance, build, operate, maintain, manage, and collect user fees for a facility for Contracting out facility availability a fixed period to amortize its investment, and at the end of the franchise, the title reverts to the government. can have several benefits. Facilities can be built more quickly than under traditional Build-own-operate The government either transfers ownership and responsibility for an existing facility or contracts with a private partner to build, own, and procurement arrangements, provided that operate new facility in perpetuity. authorities have made a detailed quanti- Source: World Bank 2006. fication of capital costs involved and have Understanding Public-Private Partnerships in Education 13 reasonably good capital planning pro- cesses in place. Using these contracts in BOX 1.1 Sources of capital public education often has positive effects for the nonprofit provision of throughout the sector, not just in the facili- ties involved. For example, private involve- education ment in the financing and construction of There are only few sources of capital education facilities often results in better- funding for the nonprofit provision of maintained buildings that do not require social services. These include: costly renovations. · Publicly guaranteed or subsidized However, the long-term purchasing com- bonds mitments required for contracting out the · Public subsidies financing and construction of an education · Private finance with a government facility are difficult for many governments guarantee (or quasi-guarantee) to manage, and the associated repayment · Retained earnings risks can make loans obtained by the pri- · Donations vate consortiums very costly. Many govern- · Long-term loans (restricted to large, ments find it difficult to set and maintain corporate, nonprofit organizations) service prices high enough to allow consor- Because the first three require public tiums to pay back (equity or debt) capital sector support, they defeat the purpose payments. This difficulty is compounded if of mobilizing nongovernmental finance. the education authorities have either hid- The last three are used to only a limited den or poorly quantified the capital costs of extent in most countries. these public facilities at the outset, which Source: World Bank 2006. can lead to exorbitant payments for pri- vately provided infrastructure. There are often only limited efficiency gains and cost savings from contracting out facility availability because of the high in education, is to contract private firms cost of borrowing for social infrastructure to both provide and operate facilities, in and the limited range of savings associated other words, to undertake all of the activi- with the private design, construction, and ties associated with delivering the needed operation of facilities compared with tradi- services and infrastructure. In essence, tional public procurement. For most social the government simultaneously imple- services, more significant cost savings can ments two forms of contract with the same be gained from contracting out operational operator--a contract for facility financing, services to the private sector. Capital costs, development, and availability and a long- including maintenance, rarely exceed 15 term contract for providing services. The percent of total service costs in education rationale cited most often for this form of and health (World Bank 2006). Thus, non- contracting is that it enables governments profit organizations are often unable to to obtain needed capital investment while participate in contracts for the finance and providing the operator with a considerable construction of facilities because they have incentive to organize and deliver services as less access than for-profit organizations to efficiently as possible. The efficiency gains the large amount of long-term finance that that the private consortium can capture is needed (box 1.1). from both constructing and operating the schools may make up for the fact that they face higher costs of borrowing than the Both facility availability government. and education services Managing these facility availability and (comprehensive contracting) operations contracts is clearly challenging. Another form of contracting that some It is "best practice" for private participation governments have used in the social sec- initiatives to be managed by the central tors, particularly health care, but not yet government, often in a PPP unit attached 14 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION to the finance ministry or treasury (World it difficult for the staff of these two enti- Bank 2006). This is done to ensure that the ties to collaborate. The contracting process government rapidly develops the expertise itself can be expensive, which may dampen that it needs to manage the "transaction" interest among potential private operators or capital part of the initiative. However, and investors because unsuccessful bidders in the social sectors, the service purchas- have to absorb the considerable cost of bid- ing contract is an integral part of the fea- ding. Despite these challenges, comprehen- sibility and attractiveness of the proposed sive contracting combines the advantages private involvement. Officials from both of contracting out both facility availability the central unit and the sectoral ministry and services, and savings and efficiencies must work together effectively to design could result from having the same organi- the two contracts (World Bank 2006). zation design and build a facility in which Nonetheless, their very different priori- it will deliver high-quality services at the ties, training, and perspectives often make lowest possible cost. International Experience 2 While governments remain the main finan- In the most common type of PPP, gov- ciers of primary and secondary education, ernments fund existing private schools, a substantial share of education worldwide mainly to increase access to education but is now delivered by private agents (Lewin also to enhance quality by enabling poor c h a p t e r and Sayed 2005). Private enrollment has students to attend better private schools increased faster than public enrollment and by introducing school competition to in recent years. Enrollment in private pri- promote efficiency. In more recent types of mary education grew by 58 percent between PPPs, governments have contracted with 1991 and 2004 from 39 to 62 million, while private providers to deliver a range of inputs public enrollment grew by only 10 percent and services with the expectation that they from 484 to 530 million during the same will introduce new pedagogical skills and period (UNESCO 2007). Sub-Saharan management efficiencies that the public Africa, the Middle East, and South Asia are sector lacks, thus generating alternatives to the regions with the largest growth in the traditional forms of public education. As dis- private provision of education (UNESCO cussed in the previous chapter, contracts for 2007). education-related services can cover a range To increase access and improve quality of services and inputs including the private in education, many governments are find- management of public schools, subsidies ing it effective to separate the financing of and vouchers, private finance initiatives for education from its provision (World Bank school construction and maintenance, and 2003a). Empirical evidence suggests that professional services such as teacher train- education systems in which schools are ing, curriculum design, and textbook provi- publicly funded but privately operated are sion. The expansion of private participation associated with better student performance in the education systems of both developed (Schütz, West, and Woessmann 2007). So and developing countries is increasingly governments are exploring ways to involve turning them into markets with the potential the private sector in providing education. to develop innovative education methods. This chapter presents a global review of In the following chapters, we discuss public-private partnerships (PPPs) in pri- examples of public-private partnerships mary and secondary education, focusing from around the world. These countries on partnerships in which governments use and programs are described in more detail contracts as instruments of accountability. in appendix A, which contains information The underlying idea behind contracts is on 92 PPP programs and policies across 47 that they introduce a performance-based countries. The PPPs are organized by con- approach to education because they clearly tract type, as defined in chapter 1, and are link funding and provision with education listed alphabetically by country within each outputs and they direct services to under- of these categories. The list is not exhaus- served student populations, especially mar- tive but gives a representative picture of the ginalized groups such as low-income or variety and geographical location of PPPs disadvantaged students. worldwide. 15 16 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION Background and trends and regulatory frameworks to take advan- The private sector can play a different role tage of the growing capacity and expertise depending on the socioeconomic and polit- of the private sector to enhance public edu- ical scenario. Countries such as Denmark cation. For example, contracts to attract pri- and the Netherlands have used the private vate funding to build and maintain school sector to provide basic education for more infrastructure are spreading in European than 100 years by financing a wide variety countries. Also, the governments of Colom- of schools on a per pupil basis to meet the bia, Qatar, and the United States have con- demand for a wide variety of different kinds tracted with private partners to manage of schooling. More recently, in some Afri- public schools to cater to the differentiated can and Asian countries, there has been a demand for education, in some cases using a growth of low-cost private schools aimed at franchising model to take advantage of good students who cannot pay the high tuition practices and economies of scale. In several charged by elite schools or who fail to meet countries in the OECD (the Organization the eligibility requirements of high-qual- for Economic Co-operation and Develop- ity public or government-funded private ment), including Denmark, New Zealand, schools (Lewin and Sayed 2005; Andrabi Norway, and the United Kingdom, more et al. 2007). In the former example, the pri- than 20 percent of public expenditure is vate sector was introduced to the education transferred to private organizations--either sector by policy design whereas in the lat- directly or through households--to pay for ter it emerged by default in order to fulfill education services and maximize school a need. choice (OECD 2007b). The rise of the private sector's involve- ment in the education sector reflects a The public-private broader shift of public service responsi- partnership continuum bilities to the private sector. For instance, The PPP continuum depicts the main between 2003 and 2004, the number of forms of publicly funded and privately pro- approved private providers of supplemental vided education across the world. It ranges services in basic and secondary education in from systems where all provision is strictly the United States increased by 90 percent, public to systems where it is largely pub- from 997 to 1,890, while the amount of licly funded and privately provided. This federal funds available for private contract- conceptual framework helps to identify the ing increased by 45 percent between 2001 extent of a country's engagement in PPPs and 2005 (Burch, Steinberg, and Donovan in education (figure 2.1). The continuum 2007). In response, governments are devel- assumes that the responsibility for funding oping institutions, funding mechanisms, largely remains with the public sector. Figure 2.1 The public-private partnership continuum Low PPP High PPP Lacks Nascent Emerging Moderate Engaged Integral 100 percent public 100 percent private Contracts with Strictly public Subsidies to inputs private schools to Private management Vouchers; Funding systems (regulation, Private schools exist in private schools provide a portion of of public schools follows students finance, provision) education Source: Authors' compilation. International Experience 17 Countries in which the government is of them are heavily subsidized in the form fully responsible for education and related of teacher salaries but the government's services and assumes all regulatory and supervision of the use of resources is weak financing functions have no PPP environ- (World Bank 2003c). ment. Countries that allow private schools A "moderate" PPP environment is evi- to operate within a centrally determined dent in those countries where the govern- regulatory framework but provide them ment enters into contracts with private with no funding from the public budget schools that require them (and pays them can be described as having a "nascent" per pupil) to educate a specified number of PPP environment. In countries with a students for a specified length of time, usu- nascent PPP environment, public and pri- ally an academic year (World Bank 2006). vate schools are independently responsible As instruments of accountability, contracts for hiring their own teachers, providing establish the conditions under which the education and related services, and build- school must operate and specify the out- ing school infrastructure. Students can puts that the schools are expected to pro- choose between public and private schools duce. Contractual instruments are different and, in some cases, among public schools. from subsidy systems in that they introduce However, they may be constrained by their a risk-sharing element between the public families' ability to pay, academic require- and private sectors. In turn, the private ments for entry, and geographical barriers sector faces the risk of financial loss for to access. Mexico is an example of a coun- noncompliance and incentives to improve try with a nascent PPP environment where its performance (World Bank 2006). The 83 percent of schools at the basic level are objective of the contract is to guarantee publicly operated and no public funding is education for low-income students when given to privately operated schools (Woess- the public system lacks the capacity to do mann 2005). so itself (World Bank 2006). Countries where the government sub- In countries with an "engaged" PPP sidizes private schools to support their environment, private organizations sign an capacity to educate more students can be agreement with the government to man- described as having an "emerging" PPP age and operate public schools in exchange environment. In these countries, a lump for payment from the public budget. The sum from the education budget is trans- objective of these operational contracts is to ferred to entitled institutions based on cri- enhance the supply of education by allow- teria that take into account, among other ing private organizations to take over failing factors, the socioeconomic context of the public schools or to open new schools that school, the number of students enrolled, take in public students. Operational con- and their for-profit or not-for-profit sta- tracts also aim to promote innovation on tus. The subsidy is normally based on the the supply side and to increase efficiency by cost of educating a student, but, because allowing the contractors flexibility in how it does not necessarily follow the students' they manage their human and financial school of choice, it does not foster competi- resources, and by relieving these schools tion. The funds can also be applied to cover from bureaucratic constraints (Gill et al. school inputs such as teachers' salaries or 2007). Communities or the contractors textbooks. The subsidy and the way it is themselves may assume the costs of infra- applied in a school's budget vary by coun- structure and educational inputs, and the try. For example, in Argentina, 13 percent government then reimburses them for that of the education budget of the local prov- expenditure. Under the Concession Schools inces is transferred to private schools with model in Colombia, the state provides the no objective criteria to guide the expendi- school infrastructure and selects the stu- ture, and 85 percent of that money is tar- dents (Barrera-Osorio 2007). In contrast, in geted to primary schools (Villa and Duarte the U.S. state of Minnesota, charter schools 2005). In Bangladesh, almost 97 percent may own school infrastructure as long as it of secondary schools are private, and most is not purchased with state funds; they may 18 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION also lease property from school boards or has had public choice since the 1990s, and nonprofit organizations (Minnesota House a high degree of autonomy is devolved to of Representatives 2005). Private contrac- schools (Rinne, Kivirauma, and Simola tors usually receive payments equivalent 2002), which means that the education sys- to the per student cost of providing educa- tem is under public control but has strong tion (World Bank 2006). Other examples of autonomy and accountability features. countries with an "engaged" PPP environ- ment include Qatar, with its independent Public-private partnerships and the schools, and the various Latin American countries where the Fe y Alegría network funding of existing private schools operates (Allcott and Ortega 2007; Brewer Public-private partnerships are widespread et al. 2007). in demand-side financing of private school In the strongest or "integral" PPP envi- operations, including vouchers, subsidies, ronment, the public sector funds private capitation grants, and stipends; and, more schools by providing students with vouch- recently, in contracts for the provision of ers that will pay for their education at the education (World Bank 2006). The main school they choose to attend, thus encourag- objective of these PPPs is to increase access ing student choice and school competition. by giving families money to invest in their In these countries, governments largely rely children's schooling by compensating them on the private sector to provide and admin- for the cost of education (Patrinos 2000). ister education but retain regulatory and In addition, demand-side mechanisms financing responsibilities. The rationale is promote parental choice, school competi- that parents can send their children to the tion, and school accountability (Gauri and most productive school based on their pref- Vawda 2004). Parents can choose the best erences (Hanushek and Woessmann 2007). schooling alternative for their children, If private schools are more productive, which may induce pressure on schools to then their enrollment will tend to increase increase enrollments and to achieve better while improvements will take place in all academic results at a lower cost (Hanushek schools as they try to compete for poten- and Woessmann 2007; Hoxby 2000). tial students. Countries with an enabling Moreover, by funding parental choice, environment have also devolved autonomy schools become accountable to families to schools on the grounds that keeping the rather than to the central government, thus locus of decision-making as close as pos- giving them incentives to develop inno- sible to the locus of schooling produces the vative approaches to learning (Hanushek best learning outcomes because this makes and Woessmann 2007). In Haiti, where schools accountable for their actions and socioeconomic and political problems have outputs (World Bank 2003a). Examples weakened the public sector's capacity to include countries such as Belgium and the deliver adequate education services, 80 per- Netherlands, where private schools receive cent of education providers are private. One public funding and where 68 percent and 83 alternative way to improve access, quality, percent of secondary education enrollments and equity outcomes is to channel public are in private schools (World Bank 2008), funds to nonpublic education providers and Chile, where one of the largest univer- within a regulatory framework that holds sal vouchers programs covers 38 percent of schools accountable for the funds that they the student population (Bellei 2005). receive and to offer support (in the form Good outcomes can also be obtained in of training and technical assistance) that countries that rely on public provision. In strengthens schools' educational and man- Finland, for instance, 97 percent of enroll- agerial capacity (box 2.1). ments are in public schools and the country Critics of such policies argue that, when is one of the top performers on the OECD's public funding flows to private schools, international student assessment known the distribution of socio-economic and as Programme for International Student educational characteristics is important in Assessment (PISA) (OECD 2007a). Finland determining parental choice and that this International Experience 19 BOX 2.1 Private schooling in Haiti In Haiti, 80 percent of all primary stu- that poor students who are not enrolled otherwise remain out of school. Approved dents attend nonpublic schools, which in school can attend nonpublic primary proposals are sent to the Department of are financed by parents, religious asso- schools for free. Eligibility is based on Administrative Affairs, which then trans- ciations, and nongovernmental orga- proposals submitted by the schools and fers funds to the schools' bank accounts. nizations, among others. The quality are evaluated based on six criteria: (i) Those private schools that receive funds of instruction and learning tends to be geographic location and related poverty are required to submit a simple financial poor, and the school-based management classification, (ii) the quality of the educa- and technical report (using a basic tem- capacity is extremely weak. tion provided, (iii) governance, (iv) com- plate) to account for their use of the funds The World Bank's Haiti Education For mitment to maximizing the enrollment and to indicate the numbers of students All Adaptable Program Grant gives the capacity of the school, (v) the age of entry that they have enrolled. management committees of eligible pri- of students, and (vi) a demonstrated com- vate schools a $90 subsidy per student so mitment to reaching children who would Source: World Bank 2007b. arrangement can lead to students being ences of specific communities. There is segregated by income level and academic an important presence of religious-ori- achievement, with no improvement on ented private schools. average academic achievement (Hsieh and · Parents are free to choose between public Urquiola 2006). Some studies suggest that and private schools and, in some cases, in large-scale voucher programs, the posi- among public schools. tive effects of competition are limited to · Finance and provision are separate. high-achieving students and that not all · Private schools must comply with edu- parents choose their schools based only cation standards defined at the central on academic criteria (Andersen, 2008; level (Andersen 2008). McEwan 2001). There has been a move toward school- Universal voucher programs to increase based management, in which governments access and introduce school choice devolve some or all autonomy to schools Several high-income countries have and allow them to manage and allocate long had education systems that rely on their own resources to stimulate innovation voucher-like mechanisms, and most of the (European Commission 2007). As part of children in these countries who attend pri- decentralization reforms in the 1980s and vate schools receive vouchers. In the Neth- 1990s, Chile, the Czech Republic, Hungary, erlands, 69 percent of primary enrollment and Sweden introduced system-wide vouch- is private; in Belgium, 54 percent; and in ers that promote parental choice and enable Denmark, 12 percent (World Bank EdStats private schools to receive public funding. 2008). These de facto systems have been In Chile, 94 percent of schools receive pub- in operation for more than 100 years and lic funding, and over 50 percent of urban fit the theoretical characteristics of more schools are private and for-profit (McEwan, recent voucher programs designed explicitly Urquiola, and Vegas 2007). Private schools to promote choice and competition (Ander- can choose their students and can be for- sen 2008). The most prominent features of profit or not-for-profit. Almost 90 percent de facto voucher systems include: of subsidized schools receive co-funding from parents (Contreras, Bustos, and Sep- · Funding is based on expressed demand. ulveda 2008). In Sweden, the reform autho- · All private schools share the risk that rized student choice and public funding for if they cannot attract enough students, a wide variety of operators, including for- they will have to close. profit corporations. Unlike in Chile, Swed- · Private schools have a diverse student ish public and private schools are subject to body because they reflect the prefer- the same rules and receive the same amount 20 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION per pupil, and private schools do not charge for girls, disadvantaged, hard-to-reach, fees, making them a real option for poorer and minority students (Angrist et al. 2002; students (Ahlin 2003). In the Czech Repub- Carnoy and McEwan 2001). Bangladesh lic and Hungary, market incentives intro- (1994­2001) had a program that gave sti- duced after communism led to a growth in pends to girls who had demonstrated high private schooling, mostly at the secondary attendance rates, scored high in school level (Filer and Münich 2000). exams, and stayed unmarried until the age Several African countries subsidize pri- of 18 or until they had obtained a second- vate schools, mostly faith-based nonprofit ary school certificate. The program sub- organizations, either with school inputs stantially increased girls' enrollment but (such as teacher salaries and textbooks) no effort was made to increase the number or through per pupil grants. The Gambia, of teachers to avoid overcrowded classes Mauritius, and Zimbabwe rely substan- (Raynor and Wesson 2006). A similar pro- tially on private schools to deliver public gram in Pakistan helped to solve the under- education (LaRocque 2008). Recently, the supply of education services in urban areas attempt to achieve universal enrollment in by encouraging existing private schools to basic education coupled with limited pub- open new facilities and thus create econo- lic funding has increased demand across mies of scale, but the program was less suc- Africa to such an extent that this has fueled cessful in rural schools, which had more a growth in the number of private low-cost difficulty in hiring teachers and suffered schools that cater to low-income students, from a higher turnover (Orazem 2000). mostly at the secondary level (Lewin and Colombia's secondary school voucher Sayed 2005). This has given rise to a two- program, Programa de Ampliacion de tier system, with a few well-funded private Cobertura de la Educacion Secundaria, schools that cater to high-performing stu- which benefited 125,000 students between dents and many private schools with no 1991 and 1997 in low-income neighbor- government support that do not perform as hoods, yielded several good practices. The well (Verspoor 2008). Although many Afri- program increased secondary enrollment can countries recognize the importance of rates by allowing parents to choose among private schools in meeting demand and private schools and by providing a renew- have found ways to expand access to educa- able voucher as long as the student met the tion, the quality of the education and equity academic requirements needed to move on of access remain challenges (Verspoor to the following grade (Angrist et al. 2002). 2008). To ensure accurate targeting, the program The experience in Africa demonstrates required students to prove that they lived in the importance of strengthening the capac- a low-income neighborhood and that they ity of the public agencies responsible for had already been admitted into a participat- regulating, monitoring, and contracting ing private school. An alternative method private schools. It is also important to facili- of targeting is to use funding formulas that tate the access that private operators have to favor students from lower-income families. capital and technical assistance to improve For instance, in South Africa, public and their education and management practices private schools are categorized by their pov- and to create institutions to implement erty level and receive subsidies depending PPPs and guarantee flows of information to on the level of tuition fees that they charge parents on school characteristics (Verspoor their students (Lewin and Sayed 2005), 2008). with the poorest schools receiving the high- est subsidies. Targeted voucher programs can reduce inequity Education service contracts include Targeted voucher programs are a use- quality output specifications ful way to widen access to higher quality When governments contract out education schools, and to reduce inequity and con- services, they contract with existing pri- straints to access and achievement gaps vate schools to educate a specific number International Experience 21 of students in exchange for a per pupil pay- scheme. The private schools must be cer- ment. The contract introduces accountabil- tified and meet several criteria to receive ity and risk-sharing between governments subsidies, including meeting input specifi- and private providers in the provision of cations and quality indicators and having education. More countries have subsidized prior experience in the education market private schools or adopted voucher pro- (Sakellariou and Patrinos 2008). Uganda's grams than have experimented with con- universal secondary education policy was tracting out education services. introduced in 2007 to boost enrollment at Contracting out basic education services the secondary level by contracting out the is part of Colombia's strategy to increase education of students that are not served coverage of vulnerable populations (World by public and government-aided schools Bank 2006). Local governments are respon- in exchange for fixed a per student fee. sible for managing and supervising these Each party's responsibilities are speci- contracts within parameters established at fied in a memorandum of understanding the national level. The local governments that requires private schools to provide carry out a tendering process and encour- authorities with performance data on a age competition by requesting proposals range of agreed indicators, to submit prog- from private operators. They then assign ress reports, and to be subject to periodic beneficiary students to selected schools reviews and assessments of academic perfor- except in Cali, where families are allowed mance (LaRocque 2008). In the province of to select the private school of their choice, Punjab, Pakistan, the Foundation Assisted which encourages schools to compete to Schools Program introduced vouchers to attract students (World Bank 2006). increase enrollment and improve quality The government of Côte d'Ivoire pays in poor areas using accountability mecha- private secondary schools a fixed amount nisms that link increases in access with to educate a student under a contracting quality measures (box 2.2). BOX 2.2 Punjab Education Foundation: Foundation Assisted Schools, Pakistan The Punjab Education Foundation was The accountability components a special institution include less established in 1991 and restructured include: bureaucratic pressure on schools in 2004 into an autonomous and inde- from traditional government pendent institution to promote high- Requirements for Eligibility. At least institutions and the potential to quality education for the poor through two-thirds of students have introduce special management partnerships with the private sector. It is to score at least 33 percent in practices in contracting. funded by the government of the Punjab an academic test as a pre- Incentives and Sanctions Related to province of Pakistan and is headed by requisite to receive vouchers. In Performance. The program includes a 15-member, government-appointed addition, schools have to meet performance-based incentives board of directors, the majority of whom other basic school input require- at the school and teacher levels. are from the private sector. ments, largely of a nonquantifi- Monetary awards are granted to The Foundation Assisted Schools Pro- able nature, that are evaluated by the school that demonstrates the gram aims to improve education quality inspectors. highest pass rate, and actual test by taking full advantage of the capacity A Specialized Institution to Manage scores are taken into account. of the mushrooming number of private PPPs. The program is managed Teachers in schools with high pass schools in Punjab. Approximately 33 per- by an independent institution, rates are entitled to direct mone- cent of children aged 6 to 10 who attend the Punjab Education Founda- tary bonuses. If schools fail to meet school are enrolled in private schools, and tion, which is fully funded by the minimum academic, infrastructure, private enrollment shares are on the rise. provincial government and whose or teaching requirements for three The program attempts to improve quality mandate is to use public-private consecutive years, they are banned through three fundamental components: mechanisms to increase access from the program. vouchers, teacher training, and monetary to and improve the quality of the incentives to schools for improved aca- province's low-cost private educa- Sources: World Bank staff; Punjab Education demic performance. tion sector. Advantages of having Foundation Web site (www.pef.edu.pk). 22 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION Public-private partnerships The Ministries of Education in the that bring alternative operators Dominican Republic and El Salvador into the education system have also contracted out services to the Escuela Nueva Foundation to support and Involving private organizations in activi- strengthen the program's implementation. ties beyond providing education services The Escuela Nueva Foundation usually has expanded the education market and assumes part of the total value of the con- produced new forms of public-private tract by donating textbooks or by making engagement. Outsourcing education-re- some other in-kind contributions. lated services is justified because private Pakistan has also recently developed expertise and education innovations can ways to support private schools that serve add value to public education, but there are low-income students. The Cluster-Based two more advantages to contracting exter- Teacher Training Program was introduced nal providers to support different aspects to improve teaching practices in Punjab of the operation of public schools. First, by engaging specialists to conduct content competition between multiple providers knowledge training for teachers in clusters can improve the quality of the services that consisting of one public and two private they provide and can reduce costs. Second, schools. Similarly, the Quality Assurance economies of scale can result when contrac- Certification Program categorizes schools tors service multiple schools (World Bank using quality criteria to inform parents' 2006). choice among schools while also providing tailored capacity-building programs in pub- Professional and support services lic and private schools (LaRocque 2008). Governments can hire private organiza- The private sector can introduce effi- tions to provide a range of support services ciencies in public education management. to public and private schools that cater to Private organizations can advise public low-income students. In many countries, schools in pedagogical and management the capacity of the public sector to deliver issues for a specified period of time, under high-quality education is compromised by contract stipulations and with the possibil- a lack of knowledge of effective pedagogi- ity of transferring the school back into pub- cal practices. To mitigate this, governments lic management. When a private contractor can contract with private organizations that provides schools with technical assistance have had proven successes with their educa- and has the ability to influence school deci- tion methods to provide certain key services sion-making, this can help reduce ineffi- such as teacher training, curriculum design, ciencies and thus improve the management textbook provision, and supplemental ser- of the school. vices for public or private schools educating In Pakistan, Punjab's Directorate of Edu- poor students. cation has contracted with Idara-e-Taleem- For example, in Colombia, public o-Agahi, a Lahore-based NGO, to serve as authorities contract with the Escuela Nueva a temporary technical adviser on pedagogi- Foundation to train rural school teachers, cal and human resources matters under an distribute textbooks, and update curricula. adopt-a-school program. The period of The Foundation also provides technical engagement is one year to carry out the assistance to rural schools to help them core work of the contract and an additional to implement the Escuela Nueva model, two to three years to transfer back these which is a multi-grade school model that responsibilities to the directorate. Through has improved core education practices in a memorandum of understanding, the Colombia's rural areas (Benveniste and organization took over failed public schools McEwan 2000). The objective of the Escuela and agreed to transfer knowledge and skills Nueva Foundation is to assure quality con- in planning, budgeting, education manage- trol and promote the sustainability of the ment information systems, and pedagogy Escuela Nueva model as it expands nation- to the schools' managers (Sarwar 2006). ally and internationally. This approach addresses the weaknesses International Experience 23 of traditional adopt-a-school programs, in information technology training, and ancil- which a private sector organization donates lary services can be outsourced, whereas money or makes other contributions (such key services such as budget approval can- as volunteer staff time) but with no binding not. There are an estimated 8,000 education agreement between the parties that gives providers in the United Kingdom, and $789 the private organization authority over key million was spent on private sector consul- aspects of school management. tants in 2002­03 (Hatcher 2006). Alternatively, schools can contract with Governments can contract with private private organizations to provide a package companies to provide academic support of services that may include, for example, for disadvantaged students. School dis- management training for principals, educa- tricts in the United States are required by tional materials, teacher training, and tech- federal law to provide supplemental educa- nology services. In Brazil, public and private tion services both to schools that have not schools subscribe to the Pitagoras Network, made adequately yearly progress for three and receive integrated advice on manage- consecutive years, and to schools with high ment procedures and pedagogic methods percentages of poor children (more than through yearly contracts for a cost equiva- 40 percent) or students with special needs lent to the cost of buying a set textbooks for (Burch, Steinberg, and Donovan 2007). every pupil (Rodriguez and Hovde 2002). These supplemental services include after- Aligning the three core pedagogical ele- school tutoring, remediation, and other ments (curriculum, teacher training, and academic support activities that take place pedagogic techniques) with school man- outside regular school hours (United States agement is one of the strengths of Pitago- Department of Education 2007). As a result, ras's services towards quality improvement the number of students receiving supple- (Rodriguez and Hovde 2002). Pakistan has mental education services increased from a similar program, Aga Khan Education 117,000 in 2004­05, to 430,000 in 2005­06 Services, which works with the Directorate (General Accounting Office 2006), and the of Private Education to strengthen instruc- supplemental services industry grew by an tional practices and management in low- average of 14 percent annually between cost private schools (LaRocque 2008). 2000 and 2003 (Hentschke 2005). The pro- Many of the functions traditionally car- gram requires private providers to offer ried out by local education authorities can high-quality and research-based services, also be outsourced to the private sector. and imposes few barriers to entering the The United Kingdom authorized the con- market (Burch, Steinberg, and Donovan tracting out of local authority functions in 2007). 2002 as part of a reform to introduce mar- While theoretically, this U.S. initiative is ket dynamics into the education system. designed to encourage competition between Local education authorities are responsible providers and increase the access of low- for funding and managing state school ser- income students to high-quality academic vices for a local area (Hatcher 2006), but support, critics argue that larger firms have private organizations are contracted by so far been in a better position to acquire the government to provide education ser- greater market shares, hence undermining vices if the local education authorities are competitiveness. Other criticisms that have found to be failing in their performance or been voiced are that providers have few if they voluntarily decide to outsource these incentives to serve special needs students functions for efficiency reasons (LaRocque because of the high cost of providing this 2008). The United Kingdom's legislation kind of education (Burch, Steinberg, and envisions the new role of local education Donovan 2007), that it is too difficult to authorities as brokers between schools and monitor and evaluate the service providers private organizations as opposed to simple effectively (most evaluations are based on service providers (Hatcher 2006). Services school visits and self-reports rather than such as pedagogic support, curriculum performance indicators), and that providers advice, school improvement strategies, have to deal with complicated contracting 24 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION requirements (General Accounting Office et al. 2006). Managed schools operate in 2006). 29 U.S. states and increased from 135 in 1998­99, to 521 in 2005­06 (Molnar et al. The private operation of public schools 2006). The second kind of private manage- Governments can also contract out the ment involves charter schools, which are entire school operation, including the edu- public schools that have been contracted cation of students, school management, out to a private operator for management financing, staffing, the provision of profes- purposes. In 2007­08, there were 4,147 sional services, and building maintenance. charter schools in 40 U.S. states, up from One argument in favor of publicly funded 253 in 1995­96 (Center for Education but privately managed schools is that they Reform 2007). Because of the decentral- have the potential to improve quality and ized nature of the education system in the increase efficiency because they have more United States, the degree of autonomy var- autonomy than traditional public schools, ies by state. which means that they are subject to fewer In the United Kingdom, academies are constraints such as bureaucratic require- independent schools operated by an auton- ments and pressure from teachers' unions omous private consortium in partnership (Gill et al. 2007; Hatcher 2003). In addi- with the central government and local tion, in schools that are publicly funded education stakeholders. The government but privately managed, decisions about provides most of the funding for these acad- school management are made at a level emies, with the private consortium expected that is closer to the beneficiary than in to contribute 20 percent. The academies are other public schools (World Bank 2003a). free from any regulations imposed by local When governments make such operation education authorities regarding education contracts with private organizations, they and staffing issues. The consortium can are leveraging not only the organization's engage in trade (to accumulate funds from expertise but also its innovative instruc- private or public sources) to generate profits tional and management practices. Publicly for the academy (OECD 2004b). Similarly, funded private schools can transform the the government of Qatar introduced the education system from the outset, simply by Independent School Program in 2004 as providing a wider range of schooling alter- part of a decentralization reform aimed at natives. Moreover, because they must offer transferring the management of all public free education and enable school choice, schools to independent operators by 2011, they provide additional places for students at introducing school accountability, and who are traditionally under-served. at boosting academic performance. Pri- With its highly decentralized education vate operators either revamp weak public system and an active capital market that schools or establish new schools (Brewer et invests in for-profit education manage- al. 2007). ment organizations and institutions that Latin America has two examples of pri- channel funds to education businesses, the vately managed public schools. The first is United States is the country with the most Venezuela's Fe y Alegría network, which experience with contracting out the opera- provides free education to poor commu- tion of public schools to the private sector nities in under-served areas and receives (Hatcher 2006). funding (85 percent of the operational There are two kinds of private man- costs) from the government through an agement of schools in the United States-- agreement between the Ministry of Edu- managed schools and charter schools. The cation and the Venezuelan Association of first kind exists when school districts allow Catholic Education. Fe y Alegría schools Education Management Organizations, for- account for 8 percent of total enrollments profit firms authorized to manage schools in Venezuela (Allcott and Ortega 2007). receiving public funds, to take over public Concessions schools in Colombia are the schools, usually failing ones, managed by second example. This concept was intro- school districts or charter holders (Molnar duced by the government in 1999 as a way International Experience 25 to provide high-quality education to low- income and high-risk students (Barrera- BOX 2.3 The differences between charter schools Osorio 2007). Concession schools are public and vouchers: The case of the United States schools managed by private school opera- tors with a record of scoring above-average In the United States, charter schools · Charter schools are not allowed on the national secondary exit examination give parents a choice among schools to promote religion, while schools for five consecutive years. Private operators as vouchers do, but there are three educating voucher students often are granted autonomy over school manage- main differences between these two have a sectarian affiliation. ment and receive a per pupil payment. In systems: · Charter schools are accountable Bogota, there are 25 public schools run as · A governmental body must for the academic results of their students on state and federal concession schools under 15 year contracts approve the establishment and tests, whereas schools educating (Villa and Duarte 2005). The program continued operation of a charter school, while schools educating voucher students are not. sought to overcome the limitations of some voucher students do not need demand-side programs, such as the lack of explicit permission to operate. Source: Gill et al. 2007. a requirement to demonstrate improved outcomes before being allowed to continue receiving public funds, by requiring con- cession schools to score above-average on that apply to public schools, such as those the annual national academic test (Villa related to pupil-teacher ratio and minimum and Duarte 2005). pass rates, and must subscribe to centrally Privately operated schools have more determined academic standards. autonomy than traditional public schools The fact that private operators can be to introduce innovations and to make their for-profit, not-for-profit, or community own decisions about staffing, curricula, and organizations sets incentives to attract pedagogical methods as long as they follow highly qualified organizations to run fail- national labor laws and national academic ing public schools. For instance, in an standards. For instance, Fe y Alegría schools attempt to diversify the education market, are considered to be successful in improving the Qatari government sought to attract a education outcomes because of their decen- variety of potential operators of indepen- tralized and autonomous nature, which has dent schools, including foreign education been replicated in other Latin American management companies, by allowing them countries (Allcott and Ortega 2007). Pri- to make a reasonable profit to operate sev- vately operated schools provide a free edu- eral schools at once to realize economies of cation but also allow school choice, thus scale (Brewer et al. 2007). encouraging competition between schools The World Development Report 2004 and more parental accountability in similar (World Bank 2003a) discusses the lack of ways as voucher programs (box 2.3). systematic learning from innovations and School operators are granted contracts insufficient replication of successful prac- for a fixed term, ranging from three (United tices. Contracting out the operation of States) to fifteen years (Colombia). These schools to the private sector can reverse contracts stipulate clear responsibilities this problem by building incentives into and objectives, and can allow governments the contracts that encourage operators to to collect information on education indica- replicate and scale up good practices. The tors that can be used to assess school perfor- good practices to be replicated should be mance (World Bank 2003a). Contracts also identified either through local research or include cancellation guidelines if school through statistical analysis. operators fail to meet education, perfor- mance, or management benchmarks (Fitz Private sector involvement in building and Beers 2002). For instance, in Colombia, school infrastructure contracts stipulate that concession schools The United Kingdom's private finance ini- must score above average on national stan- tiative allows partnerships consisting of dardized tests. Moreover, these schools private consortiums and public authori- must adhere to many of the regulations ties to construct and maintain education 26 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION facilities. The increased interest in private million. The success of this experience led finance initiatives shown by governments in to a second round of the program in which recent years has been accompanied by sub- 10 schools are to be completed between stantial growth in the global pool of capital 2006 and 2009 for $168 million (The Audit available for investment in infrastructure. Office of New South Wales 2006). Three Infrastructure funds manage an estimated other provinces, Queensland, South Aus- $133 billion worldwide, 77 percent of which tralia, and Victoria, are in the early stages was raised between 2006 and 2007 (Palter, of contracting with private consortiums to Walder and Westlake 2008). finance, build, design, repair, and main- There are three main arguments in tain school buildings, but leaving the pro- favor of private finance initiatives. First, vision of education to public authorities these arrangements enable governments to (LaRocque 2008). attract private investment, which benefits The Egyptian government has signed those whose public resources for infra- 15- to 20-year contracts with private part- structure are declining (HM Treasury ners to design, finance, and maintain 300 2008). Second, the private partner takes schools in 23 governorates and expects to on a share of the responsibility and risk for expand the initiative to 2,210 schools by the infrastructure project as a condition 2011 (LaRocque 2008). Several European of the contract (PricewaterhouseCoopers countries are also adopting this procure- 2005). Third, there is a fiscal incentive to ment procedure. Belgium's Flanders prov- circumvent regular budgeting procedures ince approved a PPP project that will select because only the annual rents that the gov- a single consortium to finance, construct, ernment pays to the private contractor are and maintain all schools built under private deducted from the annual budget instead of finance initiatives in order to benefit from the entire amount of the investment (Sadka economies of scale (OECD 2006). Germany 2006). From an education perspective, pri- has engaged in a private finance initiative to vate finance initiatives help governments to renovate, maintain, and manage 90 schools provide appropriate school buildings and to in the Offenbach province for 15 years as relieve teaching staff and school adminis- well as an initiative to refurbish and oper- trators of maintenance duties that are out- ate seven schools in Cologne for 25 years side of the primary scope of their work, (LaRocque 2008). Also, the Canadian prov- allowing them to concentrate on meeting ince of Alberta has authorized 32 PPP fixed- the learning needs of students. price contracts for 30 years, under which The United Kingdom leads the world in the private contractor assumes the risks of infrastructure PPPs, with 10 to 15 percent the costs of inflation and of any construc- of its public sector capital investment made tion delays (LaRocque 2008). through private finance initiatives (Inter- Most of the evidence about school facil- national Financial Services London 2008). ity PPPs come from the United Kingdom, Education projects account for about 19 where a focus on due diligence has kept percent of infrastructure private finance the number of projects exceeding time and initiative contracts and 8.5 percent of their cost estimations to a minimum and where value (LaRocque 2008). Recent increases in the private lenders' assumption of risk has the amount and value of its deals make it reduced the government's losses when initia- clear that the government heavily empha- tives have failed (PricewaterhouseCoopers sizes this procurement delivery model. As 2008). In 2003, 73 percent of construction of 2008, private finance initiatives in edu- projects under traditional procurement in cation attracted capital investment total- the UK exceeded the contract price, and 70 ing $11.6 billion, and this is projected to percent were delivered late. In comparison, increase to $16 billion by 2010 (HM Trea- only 22 percent of projects under private sury 2008). finance initiatives exceeded the contract Australia's first PPP infrastructure price, and only 24 percent were delivered program, which ended in 2005, built nine late (PricewaterhouseCoopers 2008). Simi- schools in New South Wales for $129 larly, a strong emphasis on analyzing and International Experience 27 allocating risk has increased discipline in must be recorded on the books (Pricewa- procurement (PricewaterhouseCoopers terhouseCoopers 2008). This reduces gov- 2008). However, critics have argued that the ernments' fiscal incentive to use private design and eventual construction of private finance initiatives, leaving only the long- finance initiative schools is of low quality, term incentives for good performance and that users were not always satisfied with increased discipline in contracting that specific aspects of the building, and that PPPs encourage (PricewaterhouseCoopers there is no evidence that private finance ini- 2008). The rationale for recording both tiatives are less expensive than traditional the incremental payments and the poten- direct government financing (United King- tial future costs of PPPs in fiscal accounts dom Audit Commission 2003). is to assure transparency, ensure debt sus- The literature is consistent in emphasizing tainability, and conduct appropriate fiscal that the main reason to adopt private finance planning (IMF 2004). Moreover, abiding initiatives is value for money, which can be by accounting standards reduces the pos- defined as the optimum combination of ser- sibility that expenditure controls can be vice quality and cost (over the whole life of bypassed and reduces hidden costs relative the contract) to meet user demands. How- to traditional procurement (IMF 2004). ever, this does not necessarily imply lower Very few private finance initiatives are kept costs (HM Treasury 2008). A better way to off the books in Australia, and this demon- estimate value for money is by comparing strates that a robust PPP policy and fiscal the net present value of private finance initia- accountability are possible (Pricewater- tive proposals with public sector benchmarks houseCoopers 2008). that represent the cost that the government Nonetheless, private finance initiative would otherwise have incurred in the pro- procurement is contentious. The primary curement project (Hurst 2004). criticism is that the high interest rates Ireland's public sector benchmarks are charged by commercial banks to private not public, so there is no evidence that pri- borrowers for infrastructure make the ini- vate finance initiatives were better values tiatives more expensive for governments than direct public financing. In fact, they (Jones, Vann and Hayford 2004). This is may have been more expensive given the because banks think that the government's high costs of private financing and of the ability to rely on tax revenues if a project tendering process (Hurst 2004). In Austra- fails means that it will not default on the lia, the cost of the public sector comparator loan (PricewaterhouseCoopers 2008). exceeded the net present cost of the private While this ability does exist, it is an inappro- sector by $9.8 million in the first round of priate criterion by which to evaluate private the New South Wales project and by $48.8 finance initiatives because it does not con- million in the second (Audit Office of New sider that public borrowing must ultimately South Wales 2006). The savings were partly be funded by taxpayers or that the sustain- due to the economies of scale achieved by ability of public debt depends on the ability assigning the management of nine schools of taxpayers to bear it. In other words, if to one private contractor, a clearly defined a government has reached its prudent level business case proposal, a competitive ten- of borrowing, a private finance initiative der process, and sound performance and can be a useful way to avoid increasing the evaluation systems (OECD 2004a). public debt, even though the cost of private While one of the main advantages of financing is higher than public borrowing infrastructure PPPs was to enable govern- (PricewaterhouseCoopers 2008). ments to avoid reporting the entire cost of Another argument in favor of private the infrastructure project in the budget at financing relates to the opportunity cost of one go, the United Kingdom government investing in infrastructure projects. A gov- recently announced that infrastructure ernment can choose not to use its available PPPs must follow international financial funding for infrastructure projects because reporting standards, which means that it forgoes the opportunity to buy invest- future private finance initiative contracts ments in a broad capital market portfolio 28 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION and to earn returns equivalent to other already operated education institutions that investors, thus capitalizing on the advan- scored above the average in national exami- tage of the lower cost of public funding nations, which has led to concession schools (PricewaterhouseCoopers 2008). having different observable characteristics, A key characteristic of private finance such as better infrastructure and academic initiatives is that the private sector is achievement outcomes, than neighboring involved in delivering services and sharing public schools. In Qatar, the government risk beyond the construction phase (HM received proposals from international as Treasury 2008). Private contractors expect well as national bidders and allowed them a return in exchange for accepting risk and to make a reasonable profit as an incentive managing a project (HM Treasury 2008). to run independent schools. Public authorities will, in all cases, assume Many of the PPP models involve a trans- an important part of the risk by placing a fer of decision-making power to the school, high value on the service to be provided if thus making providers directly accountable facility construction is delayed or of poor to the users of the service and increasing quality. In other words, governments have their efficiency (World Bank 2007a). One an interest in guaranteeing the infrastruc- explanation behind the success of privately ture needed to provide high-quality edu- operated public schools in increasing aca- cation for all. Contracts are critical for demic achievement rates, despite spending properly allocating risk between private the same or less money per pupil than pub- and public sectors, but the challenge is to lic schools, is their greater autonomy over do so in such a way as to create the right decisions about pedagogical methods and incentives for the private sector to deliver financial and human resources manage- the desired outcomes at an optimal price ment (Allcott and Ortega 2007; Barrera- (Hurst 2004). Some examples of projects Osorio 2007). Ladd (2003) adds the caveat that have failed in this regard are schools in that making individual schools account- Belfast, Brighton, and Clacton in the United able to their consumers may cause inequity Kingdom that were forced to close before because they tend to select better qualified the contract was completed due to insuffi- students, who are cheaper to educate, and cient enrollment, leaving public authorities because incentives to reduce operational with heavy financial commitments (House costs and to attract more students may of Commons Education and Skills Com- prompt them to pursue profit at the expense mittee 2006­07). of educational quality. Although the char- ter school system requires open enrollment Alternatives to conventional and free education, schools are allowed to systems of public education adopt tailored curricula that target specific Publicly funded private schools can be an populations, such as likely dropouts or stu- improvement over traditional public sys- dents with a particular interest (Hoxby and tems because new operators have autonomy Rockoff 2004), which may generate student over the selection and implementation of selection at different levels. their educational strategies, thus leading to innovation and experimentation. In addi- Conclusions tion, contracts for operational service tend Theory suggests that PPPs can increase access to attract a wide range of private partners and improve quality in education in a num- that diversify the supply of education. The ber of ways: (i) by allowing school choice, (ii) governments of Colombia, Qatar and the by putting competitive pressure on private United States have explicitly pursued this schools to remain in the market, (iii) by mak- goal and have created incentives to attract ing school operations more flexible, (iv) by high-performing or specialized education setting quality-driven output specifications, organizations to drive up quality, diversify and (v) by ensuring an optimal level of risk- the supply, and increase choice. In Colom- sharing between the public and private sec- bia, organizations bidding to run conces- tor (Patrinos 2000). Public funding of private sion schools had to demonstrate that they schools is justified by the argument that poor International Experience 29 students will benefit from the opportunity to Program has an incentive and sanction enroll in private schools of superior quality component that rewards schools with than the public schools that would other- monetary awards for good performance wise be their only option. Studies demon- and revokes their subsidies if they are strate that private schools are more effective operated poorly. than traditional public schools in delivering · The efforts of private schools to improve higher-quality education outcomes in India, the quality of their education should Indonesia, Pakistan, and Tanzania (Andrabi be supported, and governments should et al. 2007; Bedi and Garg 2000; Cox and therefore consider adding capacity- Jimenez 1991; Muralidharan and Kremer building components to voucher pro- 2006). However, ensuring academic quality in grams. Some private schools may lack education systems in which the public sector the capacity to improve education quality funds private schools and service operators because of unqualified teachers, a short- remains a challenge. International experience age of resources to enhance materials and with PPPs yields five recommendations. textbooks, and inadequate knowledge of effective teaching techniques and man- · Contracts for education services should agement processes. Some of the support include output measures and quality that has been given to private schools to indicators to track the progress of the overcome this problem includes facilitat- contractors in improving quality and ing their access to capital and arranging increasing school efficiency. These per- preferential loans to improve infrastruc- formance indicators can be quantitative, ture and buy other critical inputs, as in such as standardized tests or enrollment the case of Mauritius. Contracting out figures, and/or qualitative, such as school technical assistance to enhance finan- and parental surveys and school inspec- cial management, instructional delivery, tions (World Bank 2006). Evidence from and school leadership has been tried in Colombia shows that for contracts to be Colombia and Pakistan. effective, education authorities must have · In some countries, special agencies have sufficient capacity to carry out monitoring been created to manage private school and evaluation, perform periodic reviews operations and the flow of funds from of school performance, and enforce com- the government to privately run public pliance with the contract's quality mea- schools, and to enforce qualifying cri- sures (World Bank 2006). teria and regulations. Examples include · Operating requirements and perfor- the Sindh Education Foundation and mance standards should be defined for the Pakistan Education Foundation in private schools and operators. Belgium Pakistan, both of which are government and New Zealand require private schools agencies that manage PPPs in education that receive public funding to meet eli- and channel funds to private schools. gibility criteria (including infrastructure Another example is the Private Second- and staff requirements), follow national ary School Authority in Mauritius, which core curricula, and meet performance is an enforcement agency that oversees benchmarks. the operation of private schools and · Innovation and quality improvements manages disbursement grants (Mohadeb should be rewarded to prevent schools and Kulpoo 2008). The advantage of spe- from reverting to negative practices, cialized PPP agencies is that they may such as lobbying for extra funding, concentrate expertise on education PPPs in cases where competition for stu- and centralize the management of con- dents results in reduced public funding tracts and fund transfers, thus promot- (Gauri and Vawda 2004). For example, ing greater efficiencies in the interactions Pakistan's Foundation Assisted Schools between public and private entities. spotlight Spotlight on Chile The case of Chile provides an example of a universal voucher scheme that presents mixed evidence of the impact of public-private partnerships in education. I n the 1980s, Chile introduced a to compete, as very few of them were 2000 and 2006, making them the top universal voucher system with the closed despite declines in their enroll- Latin American country participating objective of making the educa- ment rates. On the other hand, private in PISA, and ranking just behind Tur- tion system more efficient. The reform schools responded to competition by key in the overall list of participants. In enabled students to select the school of exercising their ability to attract and reading, for example, Chile increased its their choice, either public or private, select students. Recent research shows score by 33 points, equivalent to 0.3 of and tied per-student public funding that private subsidized schools have an a standard deviation, although it could to school enrollment. The rationale academic advantage, which seems to be argued that this improvement was behind this policy was that student be associated with their ability to select driven by the more able students. The choice would encourage school compe- the most able students and those with a Chilean experience suggests that it may tition and increase accountability at the greater ability to pay. The beneficiaries take some time for school choice poli- local level by making schools responsive of the reform--those who were able to cies to yield improvements in average to parental preferences. The provision transfer to private schools during the academic achievement. The government of public funding to private schools led basic education cycle (or "switchers")-- is currently introducing policies that to the development of a school market have had higher labor market returns to address the problem of student segre- in which more than 20,000 new private schooling investments than their peers gation, including preferential subsidies schools were created and private enroll- who continued in public schools. The to increase options for students from ment rates increased from 32 percent switchers mostly belonged to the mid- poor families and the elimination of of all enrollments in 1985 to 51 percent dle class and had better cognitive skills student selection by subsidized schools. in 2005. In 2007­08, approximately 94 than their peers who remained in pub- New agencies are being established to percent of all schools in Chile, of which lic schools. On the positive side, there is take responsibility for setting national 36 percent are private, received voucher some evidence of improved household standards, tracking student progress, funding. welfare due to parents being able to and enforcing regulations relating to Nonetheless, empirical research select the school of their choice based schools' academic performance. The indicates that more than 20 years of on their revealed preferences and the gap between the theory and the prac- reform did not lead to improvements government's policy of targeting the tice of school choice still needs further in average academic achievement. poorest schools to improve their edu- exploration. Open access to informa- Moreover, there is evidence of the exis- cational quality. tion has made it possible to analyze the tence of student segregation by socio- However, more recent evidence effects of the Chilean experience; this is economic level and a public-private gap shows that after years of stagnation, another lesson from which other coun- in academic performance that favors results may be improving. Chilean tries can learn. those students who were able to trans- students demonstrated significant fer from public to private subsidized improvements in their reading perfor- Sources: Hsieh and Urquiola 2003; McEwan et al. 2008; Patrinos and Sakellariou 2008; schools. As it turns out, public schools mance in the Program for International Contreras et al. 2008; OECD 2007; World Bank in Chile did not have a strong incentive Student Assessment (PISA) test between 2008. What Do We Know about Public-Private Partnerships in Education? 3 In chapters 1 and 2, we have shown that Based on chapter 2's definition of PPPs public-private partnerships in the provision and four different types of contracts-- of education are growing rapidly in several vouchers, subsidies, the private man- countries. Unfortunately, there are still agement of schools, and private finance c h a p t e r few empirical evaluations of these experi- initiatives--in this chapter we discuss PPP ences. This chapter presents the results of programs in the light of four main objec- the rigorous empirical studies of these part- tives--access, quality, cost, and inequal- nerships and discusses their strengths and ity. The ways in which the different types weaknesses. of PPPs can affect education outcomes are We selected which studies to examine briefly discussed. As the definitions of the based on how they overcame the problem four types of contracts show, PPPs are com- of self-selection--by students, families, plex interventions, and their effects can be or schools--in most PPP programs. Self- numerous and difficult to measure. selected beneficiaries may have different characteristics than those who do not apply The definition and objectives to or do not benefit from the program. Con- sequently, simply comparing an outcome of public-private partnerships (for instance, dropout rates or test scores) PPPs can be defined as a contract that a between beneficiaries and nonbeneficiaries government makes with a private service will not accurately reflect the impact of the provider to acquire a specified service of a program because any observed difference defined quantity and quality at an agreed between these two groups may be driven price for a specified period (Taylor 2003). not only by the program itself but also by This definition covers several different the initial differences in the observable and types of contracts, which may procure unobservable characteristics of the two different services and vary in complexity. groups. The services include education services Evaluations of voucher programs, one (management, maintenance, and support common type of PPP, have to deal with services like transportation); operation ser- the problem of endogeneity. Voucher pro- vices, such as pure management; and infra- grams usually require students to apply, but structure (in what is often referred to as a the students who apply are likely to be bet- private finance initiative) (LaRocque and ter informed or more motivated than their Patrinos 2006). This review of the empiri- peers. Comparing, for instance, school cal literature focuses on three types of edu- enrollment rates of recipients and non- cation services and operations--vouchers, recipients may not accurately reflect the subsidies, and the private management of impact of the program because differences schools--and private finance initiatives for in enrollment rates may be due to these school construction. inherent differences in characteristics and Education operations contracts are not due to the actual efficacy of the pro- generally complex. The delivery of edu- gram (Nechyba 2000; Epple and Romano cation can be measured as the number of 1998). students enrolled in any given school, but 31 32 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION the number of students attending school particularly for people who are poorly does not in itself mean that the students are served by traditional methods. Increased learning anything. Observing the inputs private involvement in education, through associated with these contracts is extremely contracting or vouchers, has the additional difficult. Moreover, how much students advantages of bringing specialized skills learn depends heavily on their family back- to bear in the operation and manage- ground, a factor that the school cannot ment of public schools and of circumvent- control. In short, the parameters of these ing the inflexible salary scales and work contracts are difficult to establish and usu- rules that tend to prevail in public sector ally require long-term commitments. employment. Construction contracts are complex as The final objective of PPPs is to increase well. The private provider has to commit to the enrollment rates and improve the edu- investing over several years, and contracts cation outcomes (such as standardized test have to stipulate who owns the infrastruc- scores and dropout rates), particularly of ture. These contracts are often build-op- students from low-income families. From erate-transfer contracts, which implies an the government's point of view, reducing eventual transfer of infrastructure from the costs alone can be an important objec- private to the public sector. Construction tive. Table 3.1 presents information on the contracts also require a long-term commit- strengths of the four types of contracts ana- ment from both partners. lyzed in this chapter--vouchers, subsidies, Each type of contract works differently private management, and private finance depending on the technical capacity and initiatives--with regard to the four main the rule of law that prevail in a country. objectives of PPPs: increasing enrollment, Less complex contracts can work more improving education outcomes, reducing efficiently in low-capacity countries, while inequality, and reducing costs. more complex contracts require a higher In terms of enrollment, vouchers and degree of legal and technical development. subsidies can in theory deliver very sig- Contracting as a means of increasing nificant positive outcomes as long as there the private sector's role in education can is an adequate private supply of school have several benefits over the traditional places. However, these contracts may also public delivery of education. These benefits reallocate students between public and include greater efficiency, increased choice, private schools, and therefore, the net and wider access to government services, gain in enrollment can be small. Private Table 3.1 Expected effects of different public-private partnerships on four main education objectives Effect on increasing Effect on improving Effect on reducing Effect on reducing Contract enrollment education outcomes education inequality costs Vouchers Strong: number Strong: school choice Strong when targeted Strong when of students who private sector is receive the voucher more efficient Subsidies Strong: use of Moderate: limited Strong when targeted Moderate already built private by available places infrastructure and quality of service delivered in the private sector Private Moderate: limited Moderate: limited by Strong when targeted Moderate management by the supply of available places in and operations private school the private sector operators Private finance Moderate: limited Low Strong when targeted Strong initiatives by financial constraints Sources: Authors' compilation based on World Bank 2003a, 2006; Harding 2002; Latham 2005; LaRocque and Patrinos 2006. What Do We Know About Public-Private Partnerships in Education? 33 management and private finance initiatives sector (as the operator of schools) and the require partners to make large initial capital public sector (as the financier of schools) investment in the construction of schools, can increase enrollment while keeping limiting their ability to produce substantial the education budget low. With regard to changes in enrollments. private finance initiatives, the major argu- Vouchers, subsidies, and private opera- ment in their favor is cost savings. The tions, in theory, can have significant effects cost savings generated by the other types on education outcomes as discussed fur- of contracts depend on the specifics of the ther in the next section. In contrast, private contract (for example, the face value of the finance initiatives can only influence edu- voucher) and the private sector's effective- cation outcomes to a limited extent because ness in delivering the service. the link between infrastructure inputs and education outcomes is weak: changing Intermediate effects only infrastructure--without changing and final outcomes the pedagogic methods and teaching--will PPP programs affect school outcomes in have little or no effect on final outcomes two different ways. First, PPP programs can (Hanushek 2003). be expected to affect how schools function Equity is an important consideration in internally and, specifically, how they allo- the design of PPPs. There are those who fear cate their resources. Second, students and that increased choice will benefit only better- their families are likely to react to the new off and better-informed families, even if the incentives that are inherent in, for example, program is ostensibly targeted to the poor. voucher programs, leading to a reallocation Better-informed families, it is argued, know of students among schools. which schools have the best outcomes and The theoretical literature on the topic facilities and are, therefore, the best option suggests that there are four ways in which for their children. In other words, school the private provision of public services choice may result in students from more affects educational outcomes (see LaRocque privileged homes becoming segregated in and Patrinos 2006; Savas 2000; Nechyba, the best schools, thereby further improving 2000; Epple and Romano 1998). Each their own outcomes, while other students study analyzed certain critical variables to are left behind in ever-deteriorating schools assess the actual effect of a PPP program (Fiske and Ladd 2000). Nonetheless, sev- on education outcomes. The following four eral programs reviewed in chapters 1 and 2 conclusions may apply slightly differently explicitly target low-income students, fami- to each of the four kinds of PPP contract-- lies, and communities, and all contracts can private management, vouchers, subsidies, have a clear redistributive objective as long or private finance initiatives: as targeting is part of the agreement between the public and private sectors. Clearly, this 1. PPP contracts give schools more flex- feature of these contracts has to be carefully ibility in how they manage and pro- monitored by the public sector to avoid the vide education services than the pub- segregation effect. lic sector alone does. Generally, the Some evidence suggests that the private public sector gives schools very little sector delivers high-quality education at flexibility in hiring teachers and orga- low costs around the world. Indeed, the nizing schools, so a flexible PPP con- correlation between the private provision tract can make it possible for schools of education and high values for indicators to create a better fit between supply of education quality is positive. Using data and demand. from the OECD's Programme for Interna- Two of the school's management tional Student Assessment (PISA), Woess- decisions are critical--how teachers mann (2005) showed that public schools are hired and how the budget is al- produce lower test scores than privately located. In general, schools operat- managed but publicly funded schools do. As ing under a PPP contract have more a result, partnerships between the private freedom in teacher hiring and firing 34 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION than public schools do. Also, private ratio, revenue-to-student ratio, and schools can give their administrators cost-to-student ratio can be proxies more budgetary freedom, which may to measure risk-sharing. In short, to lead to a more efficient allocation of ascertain different risk-sharing op- resources. Finally, schools operating tions, it is first necessary to examine under PPPs have more flexibility in quantifiable financial indicators. determining such matters as the length 4. PPPs can promote competition in of the school day and the length of the the market for education. The private academic year. sector can compete with the public 2. Private providers in PPP contracts are sector for students. In turn, the public usually chosen through an open bid- sector can react to that competition ding process based on quality and cost by improving the quality of the edu- criteria. Furthermore, the contracts cation that it provides in its schools. often require contractors to produce The argument in favor of competi- certain outcomes, such as increases tion is that if that option is available in test scores. Thus the contracting (for example, in a voucher system), process and the resulting contract can students and families will shop for the yield a higher quality of education. schools that provide the best quality This argument is especially rel- of education (Friedman 1955). For evant for the private management of competition to thrive, a market for public schools. The process by which schools must exist and information beneficiary schools are chosen can be on the quality of schools must be transparent and, thus, can be directly freely and widely available. However, observed. Some PPP programs set it is clearly a challenge to measure quality requirements for their contrac- competition. tors. For instance, concession schools Competition implies that a con- in Bogota, Colombia, are based on a tracting program will also affect non- bidding process in which the appli- beneficiary schools because some stu- cant must have previous experience dents will leave those schools. Ideally, in the education sector (Barrera-Os- the impact of such a program should orio 2007). On top of this, part of the be measured by comparing two very assessment of the applicant's bid in- similar groups of individuals, one cludes examining its performance in group that benefits from the program its other schools based on a national (the treatment group), and one group standardized examination. In short, that does not participate in the pro- outcomes such as test scores and gram, (the control group). These two dropout rates are critical variables to groups should be similar in terms of measure in awarding these contracts. both their observable and unobserv- As shown below, most studies con- able characteristics. Since competition centrate on measuring the impact of implies that the program affects the PPPs on these kinds of outcomes. control group as well as the treatment 3. A PPP contract can achieve an opti- group, this makes it more difficult mal level of risk-sharing between the to make a meaningful and accurate government and the private sector. comparison between beneficiaries This risk-sharing may increase effi- and nonbeneficiaries. Data from be- ciency in the delivery of services and, fore and after the start of the program consequently, may increase amount can help to analyze the flow of stu- of resources and improved provision dents between schools. This makes it in the education sector. possible to track students who switch Measuring the optimal level of risk- from nonbeneficiary schools to ben- sharing is not straightforward. Finan- eficiary schools and consequently to cial indicators such as revenue-to-cost control for these effects. What Do We Know About Public-Private Partnerships in Education? 35 In general, students who receive vouch- Table 3.2 The effects of different types of public-private partnership contracts on education ers will choose to spend them at a better outcomes school than their current one. In the new Private school, they will interact with students management Private finance Factor of schools Vouchers Subsidies initiative who on average perform better academi- Flexibility Significant Moderate Moderate Low cally than their previous classmates. This interaction can improve the education out- Quality criteria Significant if in Significant if Moderate but Low the contract parent and significant if in comes of the voucher recipients through student driven the contract peer effects (Nechyba 2000), but this makes Risk-sharing Low Low Moderate Significant it very difficult to disentangle the effects of the voucher itself from the effects of inter- Competition Low Significant Low Low acting with better students. Source: Authors' compilation. In short, there are four factors that can improve education outcomes--flexibility in provision, incorporating quality criteria Finally, in the case of private finance ini- into the contract, optimal risk-sharing, and tiative contracts, the most important factor competition. Table 3.2 presents the relative is the risk-sharing agreement between the strengths of these factors in the four differ- government and the provider. ent types of contracts that we discuss in this Overall, some PPP programs are com- chapter. plex interventions that create incentives In the case of the private management that modify several aspects of students' of schools, flexibility, quality criteria, and behavior and of school operations. Identi- higher provider standards play a significant fying the ways in which PPPs affect educa- role in ensuring success. Quality criteria tion outcomes is extremely difficult because seem to be particularly important in these it requires analysts to disentangle each of kinds of contracts, which usually give a these incentives. This is an area in which great deal of flexibility to the private pro- more study is needed. vider. By contrast, risk-sharing plays a mini- mal role in this kind of contract because the Empirical evidence government guarantees a certain amount of The main challenge in evaluating PPPs is to resources and the demand for places is usu- overcome the problem of endogeneity, which ally very strong so the private provider does typically arises because of self-selection. not face much risk. The challenge is to build the right con- In the case of vouchers, the main fac- trol group with which to compare the tors that positively influence education outcomes of the beneficiaries of program. outcomes are quality criteria and competi- This challenge exists in all impact evalua- tion, as explained previously. Quality con- tions, but in education it is exacerbated by trol in this case is exercised when parents the fact that self-selection comes from two take their children out of the worst schools sources, schools and students. For instance, and enroll them in the best ones. As long in the case of subsidies, schools decide first as the parent picks a private school for his whether to apply for the subsidy and then or her child, then flexibility seems to play a students decide which school to attend, significant role in improving higher educa- based partly on whether the school receives tion outcomes. Risk-sharing is not a critical the subsidy. Clearly, schools and students component of this kind of contract. who decide to apply for the subsidy have In the case of subsidies, the most signifi- different characteristics than the ones who cant factor is the quality criteria. In general, choose not to apply. the government establishes certain require- This section presents empirical evi- ments that must be met by the private dence of the impact of PPPs along two schools that receive subsidies. Risk-sharing dimensions: first, by the type of empirical and competition are of secondary impor- strategy used to tackle the problem of endo- tance for this kind of contract. geneity, and second, by the type of contract 36 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION involved in the PPP. Our analysis is limited Colombia's voucher program had posi- to those evaluation studies that address self- tive effects on several outcomes over both selection through one of six strategies-- the short and long term. Recipients of the randomization, regression discontinuity vouchers attended private schools 15 per- analysis, instrumental variables, Heckman cent more than nonrecipients did. Benefi- correction models, difference in difference ciaries had 0.1 more years of education than estimators, and propensity score matching nonrecipients did as well as lower repetition (see appendix B for detailed description of rates. They were also more likely than non- these evaluation methods). recipients to complete the eighth grade, and Besides the manner in which endogene- they scored 0.2 standard deviations higher ity is addressed, our analysis also takes into on standardized tests than nonrecipients account the type of PPP contract used. To did--a significant finding. The evaluations this end, this chapter evaluates nine studies did not identify any short-term effects on of vouchers, three studies of subsidies, four enrollment. studies of private management contracts, In a complementary paper, Bettinger, and one study of private finance initiatives. Kremer, and Saavedra (2008) present evi- Because we discussed the details of these dence in opposition to the hypothesis that programs in chapter 2, this chapter focuses vouchers succeed only through peer effects. only on the results of the studies. Moreover, they show that private vocational institutions teach more relevant skills than Vouchers public voucher establishments do, which There are numerous studies of the educa- confirms the theory that flexibility in tion effects of vouchers, especially in the school management is a key to better edu- United States (Gill et al. 2007) but also in cation outcomes. other parts of the world (Barrera-Osorio The evidence on vouchers in Chile is and Patrinos 2009). Table 3.3 briefly sum- mixed and controversial. Chile's experience marizes these studies. dates from the 1980s, and any control group Colombia's Programa de Ampliación is likely to be subject to competition effects de Cobertura de la Educación Secundaria and thus would contaminate the effects of was a voucher program launched in large the voucher system (Bellei 2005). Disen- cities in 1991 by the national government. tangling these effects is difficult, especially Its main objective was to increase access to because of the lack of randomized assign- secondary education for low-income fami- ment and limited baseline information lies, and it assisted 125,000 such students. It (Hoxby 2003). Presumably, this is the rea- targeted neighborhoods classified as falling son why different studies have yielded such into the two lowest socioeconomic strata different results. and children who attended public primary Examples of early studies of the voucher schools, offering the families of these chil- system in Chile include Rodriguez (1988), dren a voucher worth approximately $190 Aedo and Larranaga (1994), and Aedo to use at the school of their choice. Families (1997). In general, these studies compared could supplement the value of the voucher the outcomes of beneficiaries with those if their chosen school charged more than its of nonbeneficiaries and are thus prone to value, but not all private schools accepted bias in their estimates. A second genera- the vouchers. The majority of cities and tion of studies used better data and more towns allocated vouchers through a lot- sophisticated estimation methods (Bravo, tery when demand exceeded supply, which Contreras, and Santhueza 1999; Carnoy enabled Angrist et al. (2002); Angrist, Bet- and McEwan 2000; Mizala and Romaguera tinger, and Kremer (2006); and Bettinger, 2000; Vegas 2002). Nevertheless, these stud- Kremer, and Saavedra (2008) to evaluate the ies were still prone to selection bias. Our program using randomization techniques. review of the empirical literature includes Angrist et al. (2002) and Angrist, Bet- the most recent studies of the program's tinger, and Kremer (2006) found that effects on education outcomes, including What Do We Know About Public-Private Partnerships in Education? 37 Table 3.3 Studies of vouchers Empirical strategy Country and study Data: type and year Outcome variables Results Colombia (Angrist, 1999-2001 ICFES Standardized Math and reading scores generate a voucher effect of 0.2 Bettinger, and (National exam, test scores standard deviations. Improves test scores for both average Kremer 2006) grade 11) records (math, reading), students and those over the 90th percentile. student level completion, cost Colombia (Angrist 1995; Cross-section, Standardized test Voucher recipients scored 0.2 standard deviations higher et al. 2002) student level scores (math, than nonrecipients. Voucher recipients' 8th grade completion reading, writing) rates were 10 percentage points higher, and their private Randomization school attendance rates were 15 percentage points higher. Republic of Korea 1995 TIMSS Standardized test 1 standard deviation increase in mean quality of peers (Kang 2007) (International test in scores (math), 7th enhances math scores at the 0.25 and 0.5 quantiles by 0.47 math and science); and 8th grades and 0.42 standard deviations. Cross-section, Students above the 0.75 quantile are not affected by the mean student level quality of their peers, but weak and median students around and below the 0.5 quantile are strongly affected by it. Chile (Hsieh and Cross-sections, Standardized test Increase in 1 standard deviation of private enrollment Urquiola 2006) different sources, scores (language, decreases change in sorting measure (language) by 1.21­0.19 1982, 1970, 1999 math), 4th and 8th standard deviations without pre-trend control for different TIMSS; student grades; repetition instrumental variables. When controlling for pre-trend level rate; years of changes in standard deviation, goes in same direction. Same schooling; sorting pattern observed when sorting measure is mathematics. measure of scores Chile (Hsieh and 1983­96 SIMCE 1 increase in standard deviation of private enrollment Urquiola 2006) (National exam, increases standard deviation in sorting measure of repetition different grades); rate by 0.50­1.62 using different instruments without pre- student level trend and with pre-trend goes from 0.47­1.71 standard deviations change. Chile (Gallegos Cross-section, Average of the 1 standard deviation in private enrollment generates about 2004) 1994­97, student math and Spanish 0.20 standard deviation in test scores and 0.24 in productivity. level portions of test Instrumental 1 standard deviation of number of priests per person boosts scores in 4th and variable private enrollment by 8 percentage points. 8th grades Chile (Contreras, Cross-section, 2005, Selection of After controlling for self-selection, no differences between Bettinger, and student level schools (parents' public and private schools. Sepulveda 2008) questionnaire), standardized test, math reading and science, 4th grade Sweden (Sandström National No failing grades Greater competition improves the standards of public and Bergström achievement test, schools. 2004) 1997­98, student and municipal level Netherlands National data, Secondary school Positive link between intensity of competition and academic (Himmler 2007) 2002­03, student grades, per student achievement in secondary school. level spending, grade inflation Chile (McEwan Cross-section, 1997, Standardized test Adjusted differences in test scores between public 2001) student level scores (language, corporations, Catholic voucher, Protestant voucher, math) 8th grade nonreligious voucher, private nonvoucher schools and public schools show differences of ­0.16, 0.35, ­0.18, 0.002, and 0.62. Similar results emerge for Spanish. Heckman Chile (Sapelli and Cross-section, 1998 Standardized test Large positive effects; 0.5 standard deviation. Effects not due correction Vial 2004) and 1999, student scores, language to sorting or peers. model level Chile (Elacqua, Cross-section, 2002 Standardized test Franchise schools scores were between 0.20 and 0.50 Contreras, and student level scores, language standard deviation higher than private independent schools. Salazar 2008) and math, 4th, 8th, No differences between private independent and public and 10th grades schools. Sources: Authors' compilation; Barrera-Osorio and Patrinos 2009. 38 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION McEwan (2001); Gallegos (2004); Sapelli the vouchers to attend private institutions. and Vial (2004); Hsieh and Urquiola (2006); (Sapelli and Vial 2004, however, presented Elacqua, Contreras, and Salazar (2008); evidence against the sorting hypothesis.) Patrinos and Sakellariou (2008); and Con- Overall, the evidence from the Chilean treras, Bustos, and Sepulveda (2008). experience with vouchers presents a mixed Sapelli and Vial (2004) used a Heckman picture, with strong empirical support for estimation to model parents' decisions to sorting of students by private subsidized participate in Chile's voucher program. schools. They found that the program had positive Studies of education systems in Den- effects (a 0.5 standard deviation) on ben- mark (Andersen 2005), the Netherlands eficiaries' test scores. By contrast, McEwan (Himmler 2007), and Sweden (Sandström (2001) used a similar estimation and found and Bergström 2004) that allow for a high that adjusted differences in math test scores degree of school choice, suggest that vouch- between public schools and different types ers have increased competition among of private schools--Catholic voucher, Prot- schools. Furthermore, in the Netherlands estant voucher, and nonreligious voucher and Sweden, this competition has had posi- schools--ranged from 0.02 to 0.31 stan- tive effects on student achievement. dard deviations. A recent estimation (Elac- qua, Contreras, and Salazar 2008), using Private management of schools the same type of technique and analyzing The literature on evaluating charter schools differences between public schools and two in the United States has grown substantially types of private schools (franchises and in recent years (see Carnoy et al. 2005 for a independent private schools), found no comprehensive review and Gill et al. 2007 differences between private independent for a recent review). Table 3.4 summarizes schools and public institutions and a dif- the recent literature. Studies by Booker ference of 0.20 standard deviation between et al. (2008), Hanushek et al. (2007), Sass franchises and public schools. Patrinos and (2005), and Solmon (2004) use micro-data Sakellariou (2008) found that the main ben- from Arizona, Florida, and Texas and apply eficiaries of the 1981 reform were students fixed effects (difference in difference) tech- who were just entering primary school or niques to reach similar conclusions. Ini- who were already in basic education. tially, students in charter schools fare worse The issue of sorting--private schools on standardized tests than their peers in choosing the best students and the best public schools, but after a period of time public students choosing private schools-- (usually three years), the scores of char- has often been explored in the context of ter school students catch up with those of the Chilean voucher program. Contreras, their public school peers. Bettinger (2005), Bustos, and Sepulveda (2008) found that using longitudinal data at the school level, student selection is a widespread practice and Bifulco and Ladd (2006), using panel among private subsidized schools. After data at the student level, found that charter controlling for a series of selection criteria school students have lower academic scores and the segmentation effects that they pro- than public school students. duce, there were no differences in results Hoxby and Rockoff (2004) and Hoxby between public and private subsidized edu- and Murarka (2007) presented evidence cation. A student attending a school that based on randomized interventions. In used selection criteria performed 6 to 14 Chicago, Hoxby and Rockoff (2005) found percent better on standardized mathemat- improvements of 10­11 percentage points ics tests than a student from a school that in the early grades of charter schools. By did not use selection criteria. Hsieh and contrast, in New York, Hoxby and Murarka Urquiola (2006) used several data sets and (2007) found that the charter school effect instrumental variable approaches to argue was evident in grades 3­8, and was between that the positive effects of the voucher pro- 0.04 and 0.09 standard deviation. gram are due to sorting. They argued that Barrera-Osorio (2007) and Allcott and the best students in public schools used Ortega (2007) reached similar conclusions What Do We Know About Public-Private Partnerships in Education? 39 Table 3.4 Studies of private management Empirical strategy Country and study Data: type and year Outcome variables Results United States, Administrative data, Standardized test Positive effects for lower grades: reading, 11 percentile points Charter schools, cross-section, 2000­ (math, reading), and math, 10 percentile points. None or a negative impact (­4 Chicago (Hoxby 2002, student level 1st­8th grades percentile points in math) for higher grades (6­8). and Rockoff 2004) Randomization United States, Administrative data, Standardized test Positive effects on math (0.09 standard deviation), reading Charter schools, cross-section, 2000­ (math, reading), (0.04 standard deviation), 3rd­8th grades, achievement New York (Hoxby 2005, student level 1st­12th grades positively correlated with the length of academic year. and Murarka 2007) United States, Administrative data, Standardized test Initially (one year), students perform worse in charter schools Charter schools, longitudinal 1995­ (math, reading), than in public ones; after three years, students have similar Texas (Booker et al. 2002, student level 3rd­8th and 10th scores to those in public schools. 2008) grades United States, Administrative data, Test scores No statistical differences between (nearby) public schools Charter schools, panel, school level and charters. Michigan (Bettinger 2005) Difference in difference United States, Longitudinal data, Standardized test Charter students score 0.1 (reading) and 0.16 (math) standard Charter schools, 1996-2002 scores (math, deviation lower than public students. North Carolina reading), 4th­8th (Bifulco and Ladd grades 2006) United States, Administrative data, Standardized test Initially, charter students score lower than regular public Charter schools, longitudinal panel, scores (math, students (0.17 standard deviation), but this depends on how Texas (Hanushek 1996­2002, student reading), 4th­7th long the school has been operating; after three years of et al. 2007) level grades operation there are no differences between them. Colombia, Panel at school level, Dropout rates, Positive effects on math (0.19 standard deviation), reading Concession schools two years, 1999­2003, standardized test (0.27 standard deviation). (Barrera-Osorio student level scores (math, Positive effects on dropout rates (1.7 percentage points). 2007) reading), 11th grade Propensity Some evidence of completion effects on nearby public and matching schools. Venezuela, Fe y Cross-section, 2003, Standardized test Positive effect on math scores (0.08 standard deviation), Alegría (Allcott and student and school scores (math, verbal (0.1 standard deviation). Ortega 2007) level reading), 11th grade Sources: Authors' compilations; Gill et al. 2007. for Colombia and Venezuela. They found For this reason, the authorities decide to that the private management of schools run a lottery to decide who should benefit. yielded higher test scores for students at Kim, Alderman, and Orazem (1999) found the end of their basic education (grade 11) that the program had had a positive impact than public institutions did. These two on girls' enrollment rates. However, because studies used propensity score matching the baseline treatment and control groups estimators with limited data and, there- had important differences, it is unclear fore, their results should be viewed with whether the differences in enrollment caution. between beneficiaries and nonbeneficiaries can be attributed to the program or to other Subsidies differences in their characteristics. There is only limited robust empirical Uribe et al. (2006) investigated the dif- evidence on subsidies (table 3.5). Kim, ferences between public and private schools Alderman, and Orazem (1999) evaluated a that receive subsidies, especially those subsidy program in Pakistan's Balochistan concerned with the use of school inputs. province. The budget allocation for the They reached several conclusions. First, program was smaller than the resources after controlling for such factors as fam- needed to cover all the target population. ily background and teacher characteristics, 40 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION Table 3.5 Studies of subsidies Empirical Country Data: type Outcome strategy and study and year variables Results Randomization Balochistan, Panel: baseline Enrollment rate Positive impact on girls' enrollment: Pakistan (Kim, and follow-up 22 percentage points (baseline: 56 Alderman and data, 1994 and percent enrollment). Orazem 1999) 1995, student level Difference Bogota, School-level Standardized Private and public schools yield in difference Colombia (Uribe panel data, test, math, 5th the same achievements, after et al. 2006) 1999 and grade controlling for individual and school 2000, student, characteristics. Private schools teacher, and have wider dispersion; public school level schools have teachers with higher level of education. Strong evidence of presence of peer effects. Public schools have larger classes. Sources: Authors' compilation. they found that students from private and funded by a private finance initiative and public schools had similar test scores. Their the other by the public budget--there is a second finding was that peer effects were 90 percent chance that the school whose one important explanation of higher test reconstruction was funded by a private scores. Indeed, students with classmates finance initiative will have a faster rate whose mothers had more education had of academic improvement. Moreover, an higher test scores. Third, class size was analysis of the first private finance initiative an important determinant of test scores. school in the United Kingdom concluded Fourth, private schools were more flexible that the overall impact of the partnership in contracting teachers, and teachers in is positive as measured by improvements in public schools have more education than the quality of infrastructure, academic per- those in private schools. Fifth, the authors formance, students' attitudes and behavior, found that the dispersion in test scores in and attendance indicators (Gibson and the private sector is wider and that the com- Davies 2008). Nonetheless, whether these bination of inputs is more diverse than in results will be applicable elsewhere remains public schools. to be seen because schools funded by private finance initiatives tend to be characterized Private finance initiatives by exceptional circumstances such as posi- As discussed earlier, it is not yet clear how tive attitudes toward the partnership on the upgrading infrastructure affects educa- part of both the school and the private tion outcomes (see table 3.1). Moreover, an contractor that influence the behavior of important line of research casts doubts on students and teachers (Gibson and Davies whether increasing inputs such as infra- 2008). structure influences education outcomes The main rationale for private finance (Hanushek 2003). The few studies that have initiatives is cost savings. The private sec- assessed the impact of private finance initia- tor, with a clear interest in the financial tives on education outcomes are case stud- outcomes of its own investments, is more ies; for example, Audit Office of New South efficient than the public sector in using Wales (2006) in Australia and Gibson and resources and in maintaining the infra- Davies (2008) in the United Kingdom. structure that it builds (Latham 2005). In terms of academic achievement, Nevertheless, there is little actual evi- KPMG (2008) found a significant correlation dence that private finance initiatives lead between private finance initiative schools to cost savings, and, because the model is and improvements in test scores. Essen- relatively new, it is difficult to reach firm tially, between two rebuilt schools--one conclusions. What Do We Know About Public-Private Partnerships in Education? 41 Channels of impact schools. However, all in all, more research It is critical to understand the internal is needed on the impact of PPPs on educa- work of schools and how interventions tion outcomes. such as PPPs can modify how schools func- tion. Unfortunately, there is little rigorous Conclusions empirical research on this topic. Uribe et There is an extensive empirical literature al. (2006) is one of the few attempts to on vouchers that is technically sound. In quantify the ways in which PPPs affect contrast, the literature on subsidies, private education outcomes. Hoxby and Mura- management, and private finance initiatives rka (2007) present some evidence of the is less abundant, with most of the evidence relationship between certain traits of on private management consisting of stud- charter schools and education outcomes. ies of the charter school experiment in the Also, Bettinger, Kremer, and Saavedra United States. It is critical for researchers (2008) present evidence on how stu- to explore further the relationship between dents acquire skills in private vocational PPPs and education outcomes. spotlight Spotlight on Education for All in Africa In Africa, a region in which the challenge to fulfill the Millennium Development Goals (MDGs) is dire, a dynamic private sector has expanded the access to and quality of education through a variety of public-private partnerships. n many countries, the private sec- I cent with completion rates of only 79 Africa, 10 percent of students attend tor plays an important role in the percent. This excess demand for high- nonstate schools and that between provision of basic education. Tradi- quality secondary education resulted 1991 and 2003, the number of pri- tionally, most nongovernment schools in a growth in the creation of low- vate primary schools increased by 113 have been faith-based and communi- cost private schools in many countries percent. ty-managed schools catering to chil- including Benin and Nigeria. In Lagos, Governments are increasingly dren from low-income backgrounds, it is estimated that in 2006, 36 percent acknowledging the role that private and in some cases, these schools have of total enrollments were in private education plays in improving educa- received public funds. Such is the case schools, and these figures may even be tion quality and widening access. Many in Burkina Faso, where the govern- an underestimate given the increase of them are devising ways to strengthen ment funds Catholic and secular pri- in the number of private unregistered the capacity of private schools to deliver vate schools, which enroll 35 percent schools. In Benin, enrollment rates in high-quality education (for example, of all secondary school students. Some private primary and secondary schools by facilitating loans to private schools countries use contracts to outline the increased from 8 to 25 percent between to improve school infrastructure as in respective responsibilities of the pub- 1990 and 2005. Mauritius and Ghana) and to ensure lic and private sectors in the provision Increased private sector partici- greater coverage for poor students and of education, as in Côte d'Ivoire and pation in the provision of education developing appropriate regulations to Uganda. helps governments to absorb student govern the operation of private schools. A worldwide commitment to ensur- demand. Public perceptions of poor However, governments still face the ing universal coverage of basic educa- quality education at public schools at challenge of putting in place favorable tion has led to public resources being all levels are driving the expansion of regulatory environments. It will be cru- concentrated at the primary level. The private schooling. Many of these new cial for governments to establish clear increasing flows of students completing private schools cater to low-income guidelines for the creation of private primary education resulted in competi- families, are for-profit, and charge schools; set up quality assurance and tion for limited places in high-quality low fees. The new private providers monitoring processes, and incentive public or private-aided secondary are aiming to offer a more conducive structures; and ensure the free flow of schools in many countries, thus giving learning environment than the public information to parents about their edu- an advantage to academically superior schools provide, sometimes with lower cation options. students. In Kenya, for instance, the pupil-teacher ratios, better infrastruc- Sources: IFC 2006; World Bank 2008; UNESCO transition rate from the primary level ture, and more motivated and account- 2007; Fielden and LaRocque 2008; Verspoor to public secondary schools was 57 per- able teachers. It is estimated that across 2008; Lewin and Sayed 2005. Designing a Conducive Environment for Education Contracting 4 For public-private partnerships to live up ensuring the success of PPPs in education is to their potential of bringing many ben- to put in place an enabling policy and regu- efits to the education sector, they must be latory framework that creates the conditions well designed. Poorly designed PPPs can under which private schools can operate c h a p t e r expose governments to significant finan- effectively and efficiently. This framework cial and policy risks, such as cost increases should also ensure that the sector as a whole and unmet objectives. Governments can do delivers high-quality education and that the several things to create an environment that wider public interest is protected. is conducive to the establishment of well- A key weakness in many countries is gov- designed and successful PPPs in education, ernment resistance to accepting the private and these are summarized in box 4.1 and sector as its partner in the social sectors. discussed in detail in this chapter. Thus, while these governments might allow private schools to exist, they do not fully recognize their contribution to achieving Common weaknesses in education important economic, social, and education policy and regulations goals. In addition, many governments limit Positive economic outcomes depend upon the the number of private schools that can be right policies and sound institutions. Well- established and discourage private invest- designed policies are also vital for achieving ment in the education sector. There are positive outcomes in the education sector in many examples of such inappropriate reg- general and in private education specifically. ulation. The most severe are laws that ban Box 4.2 presents key principles for effective private schools outright or fail to recognize design and implementation of public-pri- their existence. Other weaknesses include vate partnerships in education. The key to unclear and subjective school registration BOX 4.1 Summary of actions that governments can take to promote public-private partnerships in education · Provide a sound basis for the establish- · Ensure that PPP contracts give private · Establish appropriate performance ment of the private school sector providers considerable flexibility measures and include performance · Allow private schools to set their own · Establish quality assurance processes and incentives and sanctions for inad- tuition and other fees provide families with information to help equate performance in PPP contracts · Allow both not-for-profit and for-profit them to choose schools for their children · Develop an effective communications schools to operate · Use a transparent, competitive, and strategy to inform parents about · Promote and facilitate foreign direct multi-stage process for selecting pri- school characteristics, and the public investment in education vate partners in PPPs about the benefits and objectives of PPPs · Establish clear, objective, and stream- · Separate the purchaser and provider lined criteria and processes for estab- roles within the education administra- · Introduce a framework for evaluating lishing and registering private schools tive agency the outcomes of contracts · Provide subsidies to the private school · Ensure that the PPP contracting agency · Involve international organizations in sector has adequate capacity encouraging the growth of PPPs 43 44 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION BOX 4.2 Principles for designing and implementing public-private partnerships Defining public-private partnerships in education ability of all partners and should establish proce- dures to deter, detect, and sanction corruption. Principle 1: The nature and extent of PPPs should be based on a government's assessment of its appropriate Principle 8: Education authorities and private organizations role in education and the relative costs and bene- should agree on the output- or performance- fits of private involvement in the sector, whether based specifications to be included in the con- this involves education delivery, financing, or tract as well as sanctions for nonperformance. regulation. Principle 9: The process for awarding PPP contracts should Principle 2: The equity impact of PPPs should be a key be competitive and should guarantee procedural consideration in determining the nature and fairness, no discrimination, and transparency. extent of public and private involvement in education. Principle 10: Governments should ensure that the public agencies responsible for forming and oversee- Promoting public-private partnerships in education ing PPPs have the resources, information, and skills required to design, develop, and manage Principle 3: A sound general policy and regulatory environ- the complex contracting processes. They should ment, including high standards of public and cor- ensure that the purchaser and provider roles of porate governance, flexible labor markets, trans- the agency are separate; the government can parency, and the rule of law, including protection assign responsibility for PPPs to specialized agen- of property and contractual rights, are essential cies on partnerships and contracting education for attracting the participation of the private services if necessary. sector in all sectors of the economy, including education. Principle 11: Education authorities should have the capacity to identify fraud, track payments, and ensure Principle 4: Authorities can promote private involvement that subsidies and payment claims are legiti- by putting in place an enabling regulatory envi- mate and accurate. They should also ensure that ronment, including recognizing the role of the their private sector partners are paid in a timely private education, providing clear and stream- fashion. lined registration processes, setting up effective quality assurance systems, avoiding regulation of Principle 12: Public authorities can increase the popularity of private school fees, and providing incentives for PPPs by encouraging informed debate on the private participation. role and impact of these partnerships, consult- ing stakeholders and the public about the use of Principle 5: Access to capital markets is an essential factor PPPs, putting in place an effective communica- in increasing private participation in education. tions and awareness strategy, and creating a rig- Restrictions on access to markets and obstacles orous evaluation program. to international capital movements should be phased out. International organizations can Encouraging responsible business conduct help to promote private sector involvement in education by widening access to capital Principle 13: Private partners should observe the principles markets. and standards for responsible business conduct that have been agreed on with the government Principle 6: Public authorities can promote foreign invest- and should participate in such projects in good ment in education by treating local and foreign faith. They should not resort to bribery and other providers equally, providing investment incen- irregular practices to obtain contracts, nor should tives, and ensuring a supportive and efficient they agree to be party to such practices in the environment for investors. Investment promotion course of their infrastructure operations. agencies can support investment in education by promoting education as a priority investment Principle 14: Private partners should participate in the govern- sector. ment's strategies for communicating and con- sulting with the public. Implementing public-private partnerships Principle 15: Private providers need to be mindful of the Principle 7: PPP processes should be free of corruption and consequences of their actions for communities subject to appropriate levels of accountability, and to work together with public authorities to while public authorities should take effective avoid and mitigate any socially unacceptable measures to ensure the integrity and account- outcomes. Designing a Conducive Environment for Education Contracting 45 criteria and standards, which result in providers, but they must be well designed. A inconsistent and nontransparent application policy framework that supports the private of rules; limits on private schools' ability to sector and assures education quality is also set tuition and other fees, or to operate as key to ensuring the economic and politi- for-profit entities; foreign investment con- cal sustainability of the private education trols; lengthy and complex school registra- sector in developing countries. Public per- tion processes (box 4.3); limits on private ceptions of the quality of private education schools' ability to offer alternative curricula are crucial and can easily be influenced by and qualifications; enrollment restrictions any adverse publicity about low-quality pri- on private schools; restrictive teacher regis- vate providers. This can lead governments tration requirements; land-use limits; and to bow to public opinion and reverse their onerous requirements on private schools' enabling policies, which would adversely registration, such as financial prerequisites affect all providers. and ownership of land and infrastructure. Furthermore, governments generally Improving education favor public provision in their funding policy and regulations policies, and this does little to create an environment that enables growth in private There are several different aspects to a pol- education. Over the longer term, this is likely icy framework that encourages the growth to reduce both the quality and sustainabil- of private schools in developing countries. ity of the private school sector in developing The principle behind the framework should countries. There are hardly any quantitative be the creation of conditions under which measures of the extent to which appropri- private schools can operate effectively and ate regulations foster private education, but efficiently, while ensuring that the edu- one exception is the African private schools cation that they provide is still of high investment index, which ranks 36 African quality. countries by how attractive an environment they have created for private investment in Provide a sound basis for the operation education. The index shows that there is of the private school sector much progress to be made in developing Governments can encourage the expansion more enabling regulatory frameworks, with of the private school sector by recognizing index scores ranging from only 29 to 68 out the important role that the sector plays in of 100 (School Ventures 2008). providing education. China, Côte d'Ivoire, Regulations can be an important tool the Philippines, and Senegal have done this for protecting students from low-quality by explicitly recognizing the private sector BOX 4.3 Registering a private school in Kenya Registering a private school in Kenya is a meeting in which the application pro- · If approved, the Committee forwards long and complex process. The key steps posal was discussed, the certification the application to the Ministry of Edu- are as follows: of registration of the school's business cation for authorization. · The organization or individual that name, the names of the school manag- · The Minister of Education authorizes wishes to open a new school submits ers and their education certificates, and the school to operate. a registration application to the regis- proof of land ownership. · The registrar issues a certificate after a trar through the district or municipal · Once the registrar receives the applica- final inspection of the school has been education officer along with inspection tion, he or she presents it to the Minis- conducted by public officials. reports from the public health officer terial Committee on the Registration of and the inspector of schools, the min- Schools for evaluation. Source: Verspoor 2008. utes of the district education board 46 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION in legislation (LaRocque 1999, 2002; Borja families, and poor families. Examples of 2003). This recognition can be the founda- for-profit school chains include the Beacon- tion for building political and public sup- house Group in Pakistan, the Scholastica port for the private sector's involvement Group in Bangladesh, and international in education and for minimizing investor providers such as Global Education Man- uncertainty. This is particularly important agement Systems and SABIS. In Pakistan, given that education is often seen as a social close to 10 percent of children from fami- rather than commercial endeavor. lies in the poorest socioeconomic deciles were studying in private schools at the end Allow private schools to set tuition of the 1990s. A recent report by the educa- and other fees tion NGO Pratham found that rural private Governments can promote private involve- schools in India enrolled around 20 percent ment in education by allowing private of all students in India in 2007 (Andrabi, schools to set their own tuition and other Das, and Khwaja 2006; Srivastava 2007). fees. The governments of Ghana, India, the Governments often regulate for-profit Philippines, and Vietnam limit the level of schools to ensure that they make quality a tuition and other fees charged by indepen- higher priority than profit. However, that dent private schools (private schools that concern should be weighed against the ben- do not receive government subsidies). They efits of allowing for-profit schools to oper- also regulate the distribution of tuition and ate freely. These include increasing access to other fees among school owners or require education for both poor and nonpoor fami- schools to consult the government about lies, encouraging innovation, and attracting any fee increases. While such controls are new capital investment and new manage- often aimed at making private education rial, pedagogical, and technical skills. The affordable for the poor or preventing price prevalence of private for-profit education gouging, they can also have negative effects worldwide, including private for-profit edu- such as causing the quality of education cation that serves the poor, suggests that in to deteriorate and limiting the profitabil- practice it has become a valuable alternative ity of education investments. Even when to public provision. tuition and other fee limits exist but are not enforced, they can reduce investments Promote and facilitate foreign direct by increasing investors' uncertainty. One investment in education possible exception is when such limits are Foreign direct investment in education is agreed as part of a contractual arrangement limited but growing in developed countries, between the government and a private pro- developing economies, and transition econ- vider, for example, when the government omies. In 2005, foreign direct investment in enters into an education purchase arrange- education was nearly $3.5 billion, up from ment with a private school for the delivery just $86 million in 1990 and $401 million of education services. in 2002, and most of it is invested in devel- oped countries (UNCTAD). Foreign direct Allow both not-for-profit and for-profit investment in education remains smaller schools to operate than in other sectors of the economy. In Governments can promote investment in 2007, it accounted for less than 0.1 percent private education by allowing for-profit of foreign direct investment in the service schools to operate or to receive govern- sector (UNCTAD). ment subsidies. Several countries ban for- Several large-scale private providers profit providers from the education sector operate internationally, including the Aca- or limit government funding to for-profit demic Colleges Group, the Beaconhouse private schools. However, this bias against Group, the Delhi Public School Society, for-profit provision is not universal. Private Global Education Management Systems, for-profit schools come in various forms and SABIS. Religious orders, including and serve the full range of communities, the Catholic Church and the Seventh Day including elite families, middle-income Adventist Church, have significant global Designing a Conducive Environment for Education Contracting 47 networks of private schools. These exam- The governments of developing coun- ples show that foreign investment does not tries can also attract foreign direct invest- flow only from developed to developing ment by providing tax incentives such as countries but in fact much of it is between exemptions from customs duties on educa- developing countries or from developing tion inputs (books, teaching aids, and infor- countries to developed countries. This trend mation technology equipment) to those is likely to continue given the expansion of companies that invest in the sector. How- education provision in, and the increased ever, while these tax incentives are common globalization of, China and India. In recent in developing countries, the evidence sug- years, the governments of both China and gests that they have not been particularly Vietnam have encouraged foreign invest- successful in attracting investment. This is ment to help to meet the growing demand probably because foreign companies make for education in their countries (Borja 2003; their investment decisions based on a range VietNamNet Bridge 2006). of factors including the country's political Promoting foreign direct investment in and macroeconomic stability, the avail- education can yield great benefits for the ability of human and natural resources, the domestic education sector. Foreign private state of its infrastructure, and the trans- schools can provide families with a wider parency of its regulatory framework (Tanzi range of education options, increase com- and Zee 2001). Another problem with tax petition among schools, and foster innova- incentives is that they cost governments tion. They can also bring in much needed a significant amount of revenue and, if skills, technology, capital investment, and these costs exceed the benefits, then this is knowledge. By increasing the stock of an expensive way to achieve public policy skilled labor resulting from well-function- goals. The OECD has prepared a checklist ing school and higher education sectors, for countries to assess their incentive poli- foreign direct investment may improve the cies for attracting foreign direct investment investment climate for subsequent foreign (OECD 2003). investment. There are several steps that governments Establish clear, objective criteria can take to promote foreign direct invest- for establishing and registering ment in education, including establishing private schools an enabling policy framework within which Many countries limit the number of new foreigners can operate schools for both local providers who can set up in the educa- students and expatriates and providing for- tion marketplace. The objective of many of eign investors with investment incentives these regulations is to protect consumers such as tax holidays, subsidies, and land. from substandard education services, and Governments can also support potential this is a laudable goal. However, ensuring investors by the quality and safety of private schools and · providing them with information on protecting consumers from unscrupulous investment opportunities in educa- operators must be balanced against the neg- tion, the regulatory framework, and the ative impact of overly restrictive entry cri- broader investment environment; teria, especially in situations where demand for education exceeds what the public sec- · facilitating and simplifying the process- tor is able to supply. If the process for reg- ing of foreign investment applications; istering private schools is convoluted and · setting up an agency to promote educa- onerous, then this often has the opposite tion as a target investment sector; impact of what the government intended. · attending and sponsoring education Rather than promoting increased access, fairs, exhibitions, and conferences to better quality, and safer schools, overly promote private education investment restrictive registration criteria often deter opportunities; potential providers or increase their costs · proactively seeking to form partnerships so much that the newly created schools with potential investors. become unaffordable. Alternatively, these 48 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION restrictions may prompt some schools to the applicants of the criteria that they failed operate outside the law as unregistered or to meet and, where appropriate, give them clandestine providers, meaning the govern- a second opportunity to meet them within ment has no way of protecting the affected a reasonable timeframe. Regulators should consumers. The costs of this lack of protec- not be required to provide provisional reg- tion invariably fall disproportionately on istration when the applicant fails to meet a the poor, who have fewer education options large proportion of the criteria. Instead, these than others. applicants could be required to submit a To encourage the creation of new pri- new, revised application. The process should vate schools and to promote private invest- include an appeal procedure that specifies ment in education, registration criteria for clear and objective grounds on which those schools should be applicants who feel that their application has not been fairly considered may appeal. · realistic and achievable, so that they do Some governments may choose to intro- not unduly restrict the establishment of duce a graduated registration system for new schools; private schools, with provisional registra- · objective and measurable, to limit the tion followed by full registration after a set scope for corruption; period. Governments may also choose to · open to all prospective private school grant private organizations, such as private entrants; school associations, the power to register · output-focused, to allow schools to private schools or at least to play a greater change how they deliver their education role in school registration. See box 4.4 for services; an example of this in Cameroon. · applied consistently across different gov- ernment levels and departments. The registration process should not be BOX 4.4 Registering a private too long. To avoid unnecessarily long delays, school in Cameroon the government could establish perfor- mance targets for the regulatory authority Cameroon has a significant private school sector. Private schools are required to be and impose time limits on its decision- members of whichever private school making. For example, schools could be reg- association is relevant to their school (for istered once a certain amount of time had example, lay schools or Catholic schools). elapsed, irrespective of whether the pro- These associations have several functions, spective operator had received official noti- including representing the private sector fication from the regulatory authority. The in policy discussions with the govern- government could also establish one-stop ment. In addition, private school associa- tions play a key role in the private school shops (centralized PPP managing agencies) registration process in that they to coordinate the process. The government could also provide potential investors with · work with the prospective private school operator to prepare the applica- guidance and information (both on paper tion to open a school; and on the Internet) about how to register, · carry out initial reviews of the applica- including the registration criteria, a detailed tion to open a school (including site description of the process, the registration visits to the school) and recommend timelines, and relevant forms. any changes needed to improve the The government should inform applicants school; of its decisions in a timely fashion and should · lodge the application with the relevant include the grounds on which it accepted or provincial delegate once complete, rejected the application. There should be a along with the private school associa- tion's decision whether or not to sup- provision in the regulations for provisional port the application. registration when certain applicants meet the bulk of the registration requirements. In Source: LaRocque and Jacobsen 2000. these cases, the government should inform Designing a Conducive Environment for Education Contracting 49 Subsidize private schools to encourage · take advantage of the significant net- investment in education work of private schools in many coun- In addition to providing general investment tries to increase access; incentives, governments can encourage pri- · use funding to encourage competition vate investment in education by offering among schools and promote improve- monetary or in-kind subsidies to private ments in the quality of education, espe- schools. These subsidies can be given at the cially among schools serving low-income outset in the form of, for example, free or families. discounted land, establishment grants, and Funding-based PPPs can also be a cata- education infrastructure. Land can be espe- lyst for the expansion of the private school cially important in urban areas where land sector. Kim, Alderman, and Orazem (1999) is expensive. Another way in which govern- found evidence that subsidies led to a sig- ments can encourage private investment nificant increase in overall enrollments would be to facilitate work visas for foreign in private schools in poor urban areas teachers, management, and technical staff. (though not in poor rural areas). Simi- Ongoing support can be provided larly, Filer and Münich (2000) found that through funding-based PPPs, such as con- private schools tended to be established in tract schools and charter schools in the areas where there was excess demand and United States, concession schools in Colom- where the quality of the state schools was bia, and private school subsidy and voucher low. In Pakistan, the Punjab Education programs in both developed and develop- Foundation's Foundation Assisted Schools ing countries. Governments can also offer Program has expanded rapidly from just tax credits to parents to cover private school 20,000 students in 54 schools in late 2005 tuition and other fees as an alternative to to more than 500,000 students in 1,157 providing subsidies or give tax benefits to schools today (box 2.2 in chapter 2). Also individuals and firms that donate to schools in Pakistan, The Educators, a school fran- or education trust funds. chise model operated by the Beaconhouse These funding-based PPP models all Group, has grown to 75,000 students in 230 combine government funding with the pri- schools in 130 cities across the country, and vate delivery of education services. In this 95 low-fee private schools have been estab- respect, they differ fundamentally from lished under the World Bank's Balochistan both the traditional model of organizing Education Support Project in the first year schooling, in which the public sector both following the introduction of a voucher- funds and delivers education services, and type program. from other forms of PPPs such as Adopt-a- Funding systems for private schools need School models, in which the government to be well designed to ensure that they oper- and the private sector both provide funding ate effectively and to minimize corruption. and the public sector delivers the education Broadly speaking, governments' school service. Funding-based PPPs support the funding systems should be neutral to pro- growth of private education by making it vide equal treatment to public and private more affordable to families. They are also schools, responsive to avoid unnecessary more effective than alternative funding and delays in school registration and contract- delivery models--even fully public and fully ing processes, and targeted to underserved private models--in rapidly increasing access students. While there are many options and to high-quality education because they designs available for funding-based PPPs, · benefit from the much more flexible oper- there are several characteristics that they ating environment in the private sector; all need to have: · harness the full range of available public · Public and private schools should be and private resources; funded in a similar manner, with access · provide families with the funding that to funding based on the quality of the they need to be able to afford private education that the school provides rather schooling; than on who owns it. 50 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION · The amount of funding provided should the government, will select, employ, and be based on student numbers rather than pay school staff. Operational contracts are on inputs such as teachers' salaries. superior to management contracts because · The funding should be aimed at over- they give the private sector greater flex- coming the barriers that poor students ibility to reorganize work schedules, select face in accessing education (for example, appropriately skilled staff, pay the level of funding could be targeted by a student's salaries required to attract good staff, and income or socioeconomic status). dismiss nonperforming staff. Management · The funding criteria should be trans- contracts that put government restrictions parent, publicly available, and easily on how the contractor operates the school understood. (beyond the minimum standards required to assure safety) can significantly ham- Governments should make their fund- per the contractor's ability to determine ing for private schools conditional on the appropriate resource allocations, introduce school's satisfactory performance or to management and pedagogical innovations, its registration and accreditation status and improve the quality of education that to ensure that the funds are allocated to it delivers. schools with a proven performance record. In operational contracts, the govern- However, governments should not make the ment simply pays the private provider a funding conditional on extensive regulation management fee and an amount per stu- of the schools' inputs and operations as this dent to operate the school and then allows would limit their ability to run the school the provider to make all operational deci- in a flexible and responsive manner. It is sions, including those related to staffing. also important for governments to ensure The provider hires all staff, which is par- that these funding programs are well man- ticularly important when private providers aged and monitored and that they make are expected to improve the performance payments to private schools on a timely of failing schools where poor teaching is basis. This is not the case in many existing often a factor. Forcing private providers to programs, for example, in the educational operate within the same restrictive regula- contracting program in the Philippines, tory framework that hobbles public schools where there are long lags between when the would significantly restrict the gains from students enroll in the school and when the adopting a contracting model and limit the government pays the school its subsidy. positive impact of competition. Indeed, one recent study found that more than Ensure that private providers have the two-thirds of U.S. school district superin- flexibility to deliver services effectively tendents surveyed believed that reducing For PPPs to be implemented successfully, bureaucracy and increasing flexibility were private partners need to be given consider- very important ways to improve public edu- able flexibility in how they deliver the ser- cation (Belfield and Wooten 2003). vice for which they have been contracted. Contracts should also reflect the nature The government should spell out the desired of the service provided, encourage private outputs and performance standards and set sector investment, and ensure that all risks penalties for failure to achieve and rewards for nonperformance are covered. Contracts for success, but thereafter they should leave should be contestable--meaning that they it to the providers to decide how best to are awarded competitively, thus allowing deliver the required outputs to the speci- public authorities to compare different fied standard. offers and select the best provider. Many Providers must have as much manage- PPPs involve relatively long-term contracts. ment freedom as possible, especially in For example, private finance initiative con- staffing and employment as well as in cur- tracts are generally for 25­30 years, Bogota riculum and budget allocation. To achieve concession school contracts are for 15 years, this, governments should adopt operational and charter school contracts are for three to contracts in which it is specified that the five years. Long-term contracts are helpful managers of the private school, rather than for giving private partners greater certainty Designing a Conducive Environment for Education Contracting 51 about work stability and thus generate and their staff qualifications, and their reg- increased interest in education contracts ulatory authorities gather little information from the private sector. This is especially on the size and nature of the private school important given that some governments and tertiary education sectors. Some coun- may easily be persuaded to reverse their tries publish exam scores on a school by policies that favor PPPs given that private school basis (for example, the Philippines education remains controversial. Longer- and Uganda), while others have adopted term contracts also allow contractors more innovative ways to provide consumers with time to achieve their objectives, such as information on the performance of schools improving school performance. and tertiary education institutions. Less welcome outcomes of longer- Well-informed consumers and regulators term contracts are that they limit some are vital for the successful operation of a mar- of the benefits of competition, such as the ket in education. One way to keep consumers entry and exit of providers in response to informed is to put more stringent require- changes in demand, and lock in any poorly ments on education providers to disclose designed features of contracts for long peri- information about their schools. Among the ods of time. However, these costs need to various ways to make this happen are be weighed against the benefits of increased · requiring schools to disclose information interest from the private sector and reduced to regulators and the general public as a uncertainty for contracted providers. Also, condition of registration; to offset some of these negative effects, some contracts include clauses that require · collecting and disseminating informa- ongoing performance evaluations and the tion by education authorities on schools reauthorization of contracts at intermediate according to a number of indicators, points during the contract. including the quality of their infra- structure, facilities, and curriculum, the Improve information flows qualifications that they offer, and their and establish an effective quality class sizes, fee levels, teacher qualifica- assurance system tions, and exam scores; An important weakness in many countries · introducing independent school review is the lack of available consumer informa- systems to provide information on tion on the private education market despite school performance, such as the Educa- the rapid growth of private education and tion Review Office in New Zealand (box the wide variations in their price and qual- 4.5) and the Office of Standards in Edu- ity. Many governments collect only limited cation in the United Kingdom; amounts of information on the fees charged · introducing independent accrediting agen- by schools, the programs that they offer, cies that focus on school performance. BOX 4.5 New Zealand's Education Review Office The Education Review Office is a New provides contract evaluation services, and office's reports on individual schools and Zealand government department respon- evaluates nationwide education issues. early childhood centers are freely available sible for evaluating and reporting to The office publishes national reports that to the public and can be obtained from the public on schools, early childhood evaluate specific education issues using the individual school or center or from the centers, and other forms of pre-tertiary its inspection evidence. Education Review Office itself. education in New Zealand. The Education Review Office schedules The creation of the office played an The office disseminates useful infor- reviews of schools and centers based important part in supporting the intro- mation relevant to parents, educators, on their prior performance, current risk duction of school choice in New Zealand managers, and others involved in schools appraisal, and the amount of time since by providing information on the perfor- and early childhood education as well as their last review. Schools are usually mance of every school. to government policymakers. It reviews reviewed every three to four years, but Source: New Zealand Education Review individual schools and groups of schools, this can be more frequent if necessary. The Office Web site (www.ero.govt.nz) 52 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION Governments might choose to use com- While making information freely avail- panies in the private sector that offer test- able is an important way to improve the ing services as well as school evaluation quality of schooling, formal independent and review services. For example, the CfBT quality assurance and monitoring mecha- Education Trust, a U.K.-based not-for- nisms that evaluate the performance of pro- profit education company, reviews schools viders and their outcomes are also needed. in Oman under contract with the govern- This would ensure independent, unbiased ment in a role similar to that played by the assessments of the performance of PPPs. Education Review Office in New Zealand Well-designed quality assurance mecha- (see box 4.5). In the United States, Stan- nisms can provide consumers, providers, dard and Poor's provides school evaluation and government officials with valuable services to school districts, analyzing aca- information on the performance of pri- demic, financial, and demographic indica- vate schools and ensure that providers are tors and trends; establishing benchmarks; meeting quality standards. Several mecha- and presenting findings on school perfor- nisms have been used around the world to mance. In addition, a number of organiza- assure quality in both the private and pub- tions, such as www.SchoolResults.org (a lic sectors: public-private effort), have developed tools · Private school associations in the Phil- that enable parents to compare the perfor- ippines operate a formal accreditation mance of schools or school districts. system for private schools and higher Private sector organizations such as the education institutions. Educational Testing Service, Pearson Educa- tional, and Kaplan in the United States, and · The De La Salle Supervised Schools Pro- the Center for Educational Measurement in gram in the Philippines provides adminis- the Philippines, provide testing and assess- trative, academic, and spiritual assistance ment services that track the education perfor- to private schools that cater predomi- mance of schools and governments. Private nantly to students from low- and middle- school associations in the Philippines operate income backgrounds. a voluntary accreditation scheme for private · Various public and private organizations schools and higher education institutions (for example, the U.K.-based Worldwide (box 4.6), and other organizations provide Education Service of the CfBT Educa- information and rankings to inform stu- tion Trust; the Education Review Office dents' education decisions, including provin- in New Zealand; and the Office for Stan- cial school report cards published annually dards in Education, Children's Services, by the Fraser Institute (www.fraserinstitute. and Skills in the UK) provide school org/reportcards/schoolperformance) and the inspection and review services. Montreal Economic Institute (www.iedm. · The Sindh Education Foundation in org/main/reportcards_en.php). Pakistan operates two programs that BOX 4.6 Private school accreditation in the Philippines The private sector can play significant for schools and higher education institu- ing this service. The federation comprises role in regulating economic activity. In tions. The accreditation scheme provides several private accreditation associations, the United States, many regulations are for four levels of accreditation that confer each linked to private school associa- produced and enforced by independent benefits on institutions in the form of tions and all of which are recognized by parties and trade associations (Yilmaz increased operational freedom or eligibil- the government. In 2002­03, there were 1998). Thus, there is scope for making ity for government assistance. The private some 1,200 accredited programs in the greater use of the private sector in regu- accreditation scheme is managed and Philippines. lating a number of aspects of private overseen by the Federation of Accredit- education. The Philippines operates a ing Agencies of the Philippines, which private voluntary accreditation system charges fees to cover the costs of provid- Source: LaRocque 2002. Designing a Conducive Environment for Education Contracting 53 aim to improve the quality of education Designing public-private in low-fee private schools. partnerships Foreign organizations can play an Good design, while important, is not suf- important role in helping developing coun- ficient to ensure the success of a PPP in tries to improve the quality of education, education. It must also be effectively and particularly those countries where corrup- efficiently implemented and governments tion in the education sector is endemic (in can take several actions to improve the way the areas of testing, school licensing pro- in which PPPs are carried out. This sec- cesses, and school reviews). For example, tion presents several broad principles and foreign organizations such as Cambridge guidelines for implementing education International Exams and the International PPPs. Baccalaureate provide independently administered and internationally recog- Employ a transparent, competitive, nized qualifications. International school and multi-stage process for selecting chains such as SABIS and the Global Edu- private partners in PPPs cation Management System bring a world- A key element of effective contracting is a class curriculum to the countries in which transparent and competitive bidding pro- they operate. International organizations cess. Bidding for service delivery contracts can also help to ensure that education such as school management initiatives or standards in particular countries reach private finance initiative contracts should international benchmarks. For example, be open to all private organizations, includ- at the tertiary education level, the Inter- ing both for-profit and not-for-profit pro- national Maritime Organization is criti- viders. Contracts should be open to any cal in enforcing international standards local, national, and international organiza- in seafarer education. Governments can tions that may wish to bid to operate a pub- also require schools to be accredited by lic school, and the bidding process should international organizations or affiliated be competitive whenever possible. with foreign schools as a condition of their Schools whose management or con- registration and operation. This model is struction will be contracted out should widely used at the tertiary education level, be identified well in advance, and the list but foreign accreditation or affiliation is should be made widely available, perhaps expensive for education institutions. through an easily accessible public regis- Public and private schools should ide- ter. The bidding process should also be set ally be subject to the same quality assur- out clearly and in advance. The education ance system, but governments too often authority should send out a request for pro- impose quality assurance requirements posals to all potential bidders and should and systems on private schools that they do publicize its request widely to encourage not apply to public schools. This restricts as many bidders as possible. The result of private providers' ability to compete. The the bidding process should be advertised purpose of quality assurance mechanisms to ensure that all market participants are should be to improve the quality of educa- made aware of the identity of the success- tion delivered and to yield better education ful provider. outcomes. Too often, much of what passes A transparent and competitive bidding as school supervision involves compliance, process is likely to have positive effects in red tape, and the enforcement of rules that both the short and long term. In the short add little to a student's education experi- term, competitive bidding is most likely ence. Unnecessary rules and regulations to yield bids that deliver value for money foster an environment that is conducive to (that is, the lowest price for a given level of corruption. There are other ways to assure desired quality) and to minimize the poten- quality in private schools, including requir- tial for corruption in the awarding of the ing private schools to display their quality contracts. Over the longer term, a competi- ratings determined by independent or pub- tive process is likely to build market confi- lic quality assurance institutions. dence in both the bidding process and the 54 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION contracting agency, thereby encouraging is responsible for both purchasing and the the growth of the market in private educa- provision (and regulation) of education, tion services over time. then there is a risk that it will be biased The contracting agency should use a in favor of public schools because private multi-stage process to select providers of sector competition can threaten the viabil- education services, and these stages should ity of some struggling public schools. As include Eggers (1998, 28) argues, "Splitting policy functions from service delivery creates · clarifying requirements, including devel- incentives for governments to become more opment of contract objectives and speci- discriminating consumers, looking beyond fication of desired services and expected government monopoly providers to a wide outcomes; range of public and private providers." · developing a procurement strategy and In the United States, some states go fur- hiring a procurement team; ther in their effort to split the purchaser and · writing the request for proposals; provider functions in education by allow- · inviting expressions of interest; ing groups seeking to open and operate a · conducting contract prequalification charter school to be approved by the local checks in which bids are assessed against school district, a university, or other body requirements and a shortlist of bidders is such as education contracting agencies. selected; Build the capacity of the · interviewing the shortlist of bidders, contracting agency assessing proposals in greater depth, and An important factor in the successful negotiating contractual issues with the design and implementation of PPPs is the shortlist of bidders; need to ensure that the government agency · selecting the preferred bidder and award- responsible for these partnerships has the ing the contract; resources, information, and skills needed to · advertising the result of the selection design, develop, and manage the complex process; contracting processes that underlie PPP · commencing service (International programs. Financial Services London 2001). First, the contracting agency should have access to reliable and accurate financial and Savas (2000) presents a comprehensive administrative information. Also, updated discussion of the steps involved in con- and accurate baseline information on price tracting for the delivery of public services. and outputs is essential for the contract- Among the issues highlighted by the author ing agency to be able to make an informed are the need for a feasibility study to assess assessment of the bids submitted by orga- whether it is appropriate to contract out the nizations seeking to deliver education service; the need to foster competition in services. For example, to be able to assess the process; the mechanics and importance whether the bidding process is generating of a fair bidding process (the expression of value for money, the contracting agency interest, the bid specifications and process, has to have reliable information on the unit and the evaluation of the bids); and the costs associated with existing or alternative need to monitor, evaluate, and enforce the providers in both the public and private implementation of the contract. sectors. The contracting agency must also have access to baseline information on the Split the purchaser and provider education outcomes yielded by the sector in roles within the education general and by the schools to be contracted administrative agency out in order to be able to specify appropri- PPPs function best when the education ate performance benchmarks for the private department's policy and regulatory func- sector contractors. tions are kept separate and distinct from Second, it is vital that the contracting its service delivery and compliance func- agency employ people with the skills needed tions. If the same government department to manage the complex task of contracting Designing a Conducive Environment for Education Contracting 55 with private sector partners. The skills that pays schools in a timely fashion. The expe- are needed correspond to the wide range of rience with education service contracting in functions that must be undertaken by regu- the Philippines is instructive in this regard, lators, including designing, developing, and as recent audits have discovered fraud in the managing payment systems; accrediting form of some "ghost" schools that received and registering schools; carrying out qual- funding but existed in name only. In addi- ity assurance functions; and running pri- tion, payments to schools under the scheme vate sector incentive programs. While the were often delayed several months, which move toward PPPs may seem to signal the discouraged many potential providers from withdrawal of governments from their role bidding for contracts to operate schools. An in providing education, it does not. Rather, effective audit procedure is a vital compo- the role of government is simply changing nent of any payment or fraud monitoring from being the exclusive provider of a ser- system, and NGOs can often be successfully vice to being the facilitator and regulator for employed in such roles. a range of different providers. This means Governments that need to build their that the skill set required by the public sec- capacity to implement PPPs in education tor is also changing and now encompasses can take advantage of case studies, good skills that are very different from the skills practice guidelines, and lessons learned that used to be needed. disseminated by other countries with In particular, the shift from input con- more experience with education contract- trols to output-based contracting means ing. These come in the form of manu- that government agencies must develop als, checklists, toolkits, and standardized their capacity to contracts. Examples include the United Kingdom's Schools Private Finance Ini- · assess the various services that are tiative website (www.teachernet.gov.uk), provided in the education sector to which provides guidance and standardized determine when and under what circum- contracts for school infrastructure PPPs, stances contracting, rather than direct and the National Association of Charter public provision, could be used; School Authorizers' website (www.quali- · design, negotiate, implement, and moni- tycharters.org), which provides guidance tor education service contracts; for organizations that authorize the estab- · develop legislation that sets up a com- lishment of charter schools in the United petitive and transparent contracting States. system; In many countries, governments have · develop appropriate quality assurance chosen to set up either a dedicated, cross- mechanisms. sectoral unit to oversee the implementation of PPPs or specialized PPP teams within The move toward PPPs in education sector ministries. Establishing these units also requires public officials to adopt a new is the best way to overcome capacity weak- administrative culture. As Harding (2002) nesses such as a lack of knowledge about noted (in relation to the health sector but contracting, a dearth of the skills required it is equally applicable to education), pub- to implement PPPs, high transaction costs, lic officials need to stop thinking of them- and poor procurement incentives that can selves as administrators and managers of lead to corruption. public employees and other inputs, and In most of these countries, the units have start thinking of themselves as contract been given responsibility only for infra- managers with the ultimate responsibility structure PPPs (including schools) rather for delivering services (but not necessarily than for the entire range of PPPs. However, delivering those services themselves). in principle, governments could extend the The contracting authority must also have remit of these PPP units to include policy the capacity to identify fraud, track pay- formulation and coordination, technical ments, and ensure that claims for payment assistance, quality control, the standard- from participating schools are legitimate ization and dissemination of information, and accurate. It should also ensure that it and the promotion and marketing of PPP 56 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION initiatives. PPP units in the social sectors · A strong focus on international student have the potential to play a key role in pro- assessments may give schools an incen- viding education authorities with technical tive to refuse entry to any students who assistance in designing and implement- are unlikely to be strong performers ing contracts and in standardizing PPP academically. processes in countries with decentralized This is not to argue that performance education systems. They could also play an measures should not be set or that they important role in promoting and market- should not be backed up by financial incen- ing the concept of public-private partner- tives. Performance measures and financial ships and of specific PPP initiatives, which incentives can help align the interests of the tend to be more controversial in education school with those of students and the gov- than in other sectors. ernment. Appropriate incentives can also Establish appropriate performance help to ensure that schools remain focused measures, incentives, and sanctions for on students' needs and keep abreast of failing to perform in PPP contracts changing demands in the marketplace. The specification of requirements at an early Establishing appropriate performance phase can be crucial to the eventual success measures is critical in the design of any of the contract and needs to be carried out contract. Performance measures are neces- carefully by a multi-disciplinary team to sary for determining whether the service ensure that all aspects that influence edu- provider has met the agreed terms and con- cation quality are considered. Also, these ditions of the contract and are even more targets and expectations should be realistic important when they are prerequisites for and achievable. determining the compensation to be paid The degree to which performance indi- to the contractor. The selected performance cators can be specified will vary with the measures must be appropriate and must nature of the contract. It will be easier to reflect the outcomes required by the con- specify these measures when the services tracting authority because the contractor's being purchased are narrow in scope and behavior will be driven largely by what will simple to measure (for example, remedial be measured and rewarded under the terms instruction and literacy programs) than of the contract. Performance indicators when the services being purchased are should be specified as much as possible in broader in scope and harder to measure (for terms of measurable outcomes (for exam- example, whole school management). ple, learning improvements as measured by Performance indicators can be mea- test scores, reading levels, reduced dropout sured both qualitatively and quantitatively rates, and reduced teacher-student absen- and can be reported at different intervals. teeism) rather than inputs (for example, For example, quantitative indicators such hiring additional staff or spending more on as standardized test scores, attendance particular activities). rates, and dropout rates can be supple- These performance measures must mented by qualitative methods of assessing be selected with care because, if badly performance, such as surveys of parents designed, they can produce perverse incen- and teachers and site visits by third parties tives and lead to undesirable outcomes. For to assess progress in areas such as leader- example: ship development, the arts, and character · A heavy emphasis on academic outcomes development. Education service contracts in contracts may cause contractors to should also include performance incentives ignore the development of "softer" skills and should make payment conditional on such as teamwork. the contractor achieving the performance · An overly rigid focus on measurable out- measures. There are many examples of PPPs comes may lead to the contractor paying (including private finance initiatives such too little attention to desirable outcomes as the Bogota Concession Schools program that are more difficult to measure, such and the Punjab Education Foundation's as student self-esteem. Foundation Assisted Schools Program) Designing a Conducive Environment for Education Contracting 57 that include performance measures in their the general public about the academic ben- contracts and that make the contractors' efits that can accrue from involving private compensation (or continued participation partners in education, and promote best in programs) conditional on their satisfac- practices in developing and applying PPPs. tory performance. The World Bank's 2004 toolkit for public Of course, these performance incentives communications programs on privatiza- and sanctions will be utterly ineffective if tion is a useful resource even though PPPs the education authority lacks the ability or do not involve outright privatization. capacity to monitor contractors' perfor- mance. This monitoring should aim both Introduce a framework for to prevent fraud and to ensure that the evaluating program outcomes objectives and targets of the contract are Each PPP should be accompanied by a well- met, especially in complex PPPs such as pri- designed, rigorous evaluation. Although vate finance initiatives and funding-based a wide range of PPPs exists around the initiatives (for example, school manage- world, there is a lack of rigorous evidence ment and school subsidy programs). A par- on the impact of these programs (World ticular risk in PPPs that receive per student Bank 2006). This is especially true for PPPs funding is the potential for unscrupulous operating outside the United States and for contractors to inflate enrollment figures or nonvoucher programs. As noted in Patrinos to claim funds for schools that only exist on (2005), the best evaluations of programs paper. Various PPP programs have adopted involve experiments that randomly assign strategies to address this risk, including benefits and include a true control group. school accreditation schemes, requiring In the absence of a random design or some contractors to allow open access to school form of natural experiment, it is preferable enrollment data, and third-party validation to use such rigorous techniques as propen- of enrollment figures. sity score matching, local average treatment effects, and regression discontinuities. Develop an effective Education PPPs are highly amenable to communications strategy proper impact evaluations because many of Efforts to involve the private sector in edu- the interventions are output-driven. Hav- cation often face concerted opposition from ing more rigorous impact evaluations is rival political parties, labor unions, the important because this would increase the media, the public at large, and specific inter- amount of information available to poli- est groups. Therefore, a crucial component cymakers when they make decisions about of any PPP in education is an effective strate- program design as well as expanding the gic (as opposed to piecemeal or ad hoc) com- international knowledge about the circum- munication plan as this can substantially stances under which particular types of on reduce political risk and be an effective way education PPPs work best. of promoting a PPP initiative. A strategic communication plan needs to be built on Involve international organizations ongoing opinion research that assesses how in fostering PPPs various stakeholders are affected by the ini- International organizations can play sev- tiative. The results of this research will help eral roles in promoting PPPs. A key one the government determine what steps to is providing "early stage" equity and loan take to build support for, promote partici- capital to promote investments in private pation in, and mitigate social opposition to, education. Schools find it difficult to access the private participation initiative. investment capital with a sufficiently long The strategy may include featuring spe- time horizon. Private equity companies are cific PPPs at the school level, stressing the generally not interested because they expect desirable objectives and the solid experience short-term returns on their investments. of the private contractors. This can educate International lenders can raise the profile stakeholders about the potential advan- of private education as a legitimate sector tages and disadvantages of PPPs, inform for investment, and they can also work with 58 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION banks to mitigate some investment risks in provide education entrepreneurs with the sector. International organizations can access to capital and has recently launched also build the capacity of both banks and a microcredit program in Kenya, which tar- the education sector and help countries to gets private school operators and includes a create enabling regulatory frameworks for technical assistance facility (box 4.7). private education. The focus of most international orga- nization projects is on improving public sector schools and tertiary education insti- BOX 4.7 Microcredit facilities tutions. As Sosale (2000) shows, World for education Bank lending for education projects totaled Private schools in Sub-Saharan Africa have $4.9 billion in 1995­97, but only 11 of 70 limited access to medium- and long-term projects (about 15 percent) included a pri- investment capital. Few local banks lend vate sector component and only about half to private schools, and most loans are for of those were at the primary or secondary very short periods. Many schools also need level. However, this has changed in the last technical assistance to build their financial, decade. In 2007, 57 percent of World Bank managerial, and administrative capabilities education projects had a PPP component and to operate more efficiently. The investment component will sup- (Baksh, forthcoming). port school loans from partner banks. In addition, the World Bank provides These loans will be used to finance the policy support to governments that are construction of facilities, the purchase of looking for effective ways to involve the educational materials, and other capital private sector in providing education. It expenditures. To be eligible for financ- has also created a unit to conduct more and ing, schools will need to meet the partner better quality evaluations of the impact bank's underwriting criteria. The program will focus initially on 10 countries that have of PPPs on education. The International high enrollment rates in private schools. Finance Corporation (IFC) also supports private education projects. It has funded Source: IFC 2007. operations, including one in Ghana, that Conclusions Despite recent increases, enrollment rates enough rigorous research on the effects of remain low in several developing regions, contracting in education to be able to draw and the quality of education lags consider- many definite conclusions at this time. ably behind that in developed countries. Given market failures and equity consid- A framework for understanding erations in many countries, the public sec- public-private partnerships tor continues to be an important player in education in providing education services. However, For education services to be provided suc- increasing access, equity, and achievement cessfully, all participants--citizens, service in education in developing countries will providers, and governments--should be require innovative programs and initia- held accountable. Contracting in educa- tives from the private sector as well as pub- tion can improve service delivery by clearly lic resources and leadership. One form of assigning responsibilities among these public-private partnership (PPP) that has actors, identifying objectives and outputs, been tried in education and other sectors gathering information on the performance involves the government contracting with of the partnership in order to assess its private organizations to provide a specified progress, and ensuring enforceability of service of a defined quantity and quality at the contracts. an agreed price for a specific period of time. Many forms of contracting are currently When central and local governments pro- used in education in developing countries. vide the finance for education services but Some governments buy the services involved contract out the actual provision of those in producing education (inputs), such as services to the private sector, this can help teacher training, management, curriculum to improve the quality of education and rap- design, or the use of a school facility from idly expand access to schooling, especially private organizations. Other governments for under-served parts of the population. contract with private organizations to man- Nevertheless, PPPs are a controversial age and operate public schools (processes), subject. Some studies suggest that this including all of the activities involved in the arrangement can lead to students being education process. Some other governments segregated by income level and academic contract with private organizations to pro- achievement, with no improvement in aver- vide education to specific students (out- age academic achievement. Other studies puts). The challenges and potential benefits suggest that, in large-scale voucher pro- of contracting for services that are inputs, grams, the positive effects of competition processes, or outputs are very different. benefit only high-achieving students and There are seven main forms of contracts: that not all parents choose their children's schools based only on academic criteria. · Management services. Weak school man- While private participation in primary and agement is a common constraint to secondary education has increased signifi- improving public school performance. cantly over the last two decades in various To address it, some governments have forms of contracting models, there is not brought in private organizations to 59 60 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION manage public schools. Management lic school. The aims of these initiatives contracts can entail the private organiza- are usually to free schools from public tion managing a single school or an entire service constraints, give them autonomy, school district. Its responsibilities usually and to harness the interest and knowl- fall into four categories: financial man- edge of parents and other community agement, staff management, long-term members to improve the oversight of the planning, and leadership. Nonmanage- school. In many cases, the local com- rial personnel usually remain as public munity also contributes to the construc- sector employees. tion, upkeep, or improvement of school · Support services. Noninstructional facilities. activities, including maintenance, stu- · Education services. Instead of engag- dent transportation, and school meals, ing a private entity to operate a public are often very costly for public schools. school, some governments pay for stu- Policymakers in many countries have dents to enroll in private schools, thus, contracted out these kinds of support in essence, buying outputs. By enroll- services to increase cost-effectiveness ing students in existing private schools, and free up resources and time so that governments can quickly expand access school staff can focus on the learning without having to spend the money to process. Usually, governments tender build and equip new schools. Other contracts that cover multiple schools so governments pay for students to access that contract management expertise can specialized services in the private sector, be developed in a single place and so that such as alternative education not avail- the contracts are large enough to attract able in the public sector. When govern- many bidders. ments contract for education services, · Professional services. Contracting out they are underwriting individual stu- professional services such as teacher dent enrollment by means of vouchers, training, the provision of textbooks, scholarships, or per pupil subsidies, all curriculum design, and quality certifi- of which make it possible to target ben- cation of schools is straightforward and efits to specific students and groups. usually effective. Its main advantage is · Facility availability. Governments have that it brings private providers' expertise tried to mobilize private investment in to bear on public education. The content needed capital stock in many different and oversight of contracts are critical sectors, including education. Contract- when buying inputs. Simple input ser- ing for the provision of school facilities vices are relatively easy to specify in con- is appealing because it relieves govern- tractual terms, and the performance of ments of having to provide capital up the contractors can also be conveniently front and all at once. Contracting for monitored. Also, because there are the private finance and construction of almost always many potential providers, facilities allows the government to pay contractors must be competitive to be for these capital investments over the awarded a contract, and the government term of the contract instead of all at can credibly threaten cancelation if the once. provider's performance is not up to par. · Facility availability and education ser- Another advantage is that economies of vices. Sometimes, governments contract scale can often be achieved because one with the same private firm not only to organization can deliver these input ser- build the facility but also to undertake vices to multiple schools under many all of the activities associated with deliv- contracts. ering education and related services. In · Operational services. In some countries, these cases, the government simultane- the government contracts with private ously implements two forms of contract organizations to operate public schools. with the same operator--a contract for In these operational contracts, private facility financing, development, and agencies both manage and staff the pub- availability and a long-term contract Conclusions 61 for providing education services. The to operate within a centrally determined rationale cited most often for this form regulatory framework but provide them of dual contracting is to obtain neces- with no funding from the public budget sary capital investment while giving the can be described as having a "nascent" contractor a big incentive to organize PPP environment. Countries where the and deliver services as efficiently as pos- government subsidizes private schools can sible. The efficiency gains that the pri- be described as having an "emerging" PPP vate organization can capture from both environment. A "moderate" PPP environ- constructing and operating the schools ment is evident in those countries where may make up for the potentially high the government enters into contracts with costs of borrowing. private schools that requires them (and pays them per pupil) to educate a specified International experiences number of students for a specified length of of using public-private partnerships time. In countries with an "engaged" PPP to fund existing private schools environment, private organizations sign an agreement with the government to man- Many governments around the world have age and operate public schools in exchange been exploring different ways to involve for payment from the public budget. In the the private sector in providing education, strongest or "integral" PPP environment, including vouchers, subsidies, capitation the public sector funds private schools by grants, stipends, and contracts. In addi- providing students with vouchers that will tion, demand-side mechanisms such as pay for their education at whatever school vouchers have the advantage of promot- they choose to attend, thus encouraging ing parental choice, school competition, student choice and school competition (see and accountability. The idea is that parents figure 2.1 in chapter 2). choose the best school for their children on Some governments have used univer- the grounds of quality, which in turn puts sal voucher programs to increase access to pressure on schools to compete to attract high-quality schooling and to make schools students and to achieve better academic more diverse. Several high-income coun- results at a lower cost. tries have school financing systems that use The most common type of partnership vouchers or similar mechanisms, including is where the government funds existing Belgium, the Czech Republic, Denmark, private schools, mainly to increase access Ireland, the Netherlands, and Sweden. to education but also to enhance quality Prominent features of voucher systems by enabling poor students to attend better- include the following characteristics: performing private schools and to increase school competition to promote efficiency. · Funding is based on expressed demand Governments are increasingly recogniz- by parents. ing that PPPS have a useful role to play in · All private schools share the risk that education and are developing institutions, without students they will have to close. funding mechanisms, and regulatory · Private schools are diverse and innovative. frameworks to leverage private capacity and · Parents and students can freely choose expertise to enhance public education. between public and private schools. Countries lie on a continuum in the · Finance and provision are separate. extent to which they are using PPPs. This continuum ranges from those countries in · All schools must comply with educa- which education is provided only by the tion standards defined by the central public sector to those in which it is largely government. publicly funded but privately provided. Developing countries have begun to Countries in which the government is recognize the important role that private fully responsible for education and related schools can play in increasing access and services and assumes all regulatory and improving the quality of education through financing functions have no PPP environ- competition. Several countries subsi- ment. Countries that allow private schools dize private schools, mostly faith-based 62 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION nonprofit organizations, either by fund- contracts differ from voucher-like pro- ing school inputs (such as teachers' salaries grams in that they introduce a risk-sharing and textbooks) or through per pupil grants. element. The public and private sectors face The governments of The Gambia, Mauri- the same risk of financial loss for noncom- tius, Tanzania, and Uganda have formed pliance and share the same incentives to alliances with private schools to deliver improve their performance. education. Recently, as a result of the drive Governments can contract with NGOs towards universal primary education, there to provide professional and support ser- has been more demand for education than vices to public and private schools that the public education systems in many coun- cater to low-income students. For example, tries can handle. This problem, coupled in some situations, the capacity of the pub- with limited public funding, has resulted in lic sector to deliver high-quality education a growth in the number of private low-cost is compromised by a lack of knowledge schools that cater to low-income students. about effective pedagogical practices. PPPs Experience with PPPs across the world enable governments to introduce into pub- has shown the importance of: (i) strength- lic schools education methods that have ening the capacity of public education agen- proven to be effective in private schools by cies to regulate, monitor, and contract with contracting with private agencies to pro- private schools; (ii) building the capacity vide teacher training, curriculum design, of private providers to deliver high-qual- textbooks, and supplemental services. In ity education by giving them more access Colombia, the government contracts with to capital and technical assistance to help the Escuela Nueva Foundation to train them to improve their educational and rural schoolteachers, distribute textbooks, management practices; and (iii) creat- update curricula, and provide technical ing institutions to implement PPPs and to assistance to rural schools. guarantee access to information about edu- Governments can also contract with pri- cational outcomes of schools. vate organizations to take over the opera- Targeting voucher programs to under- tion of entire schools, including teaching, served populations (such as girls and dis- management, finance and staffing, support advantaged, hard to reach, and minority services, and building maintenance. Schools students) can increase equity in access that are publicly funded but privately man- to schooling and in eventual educational aged have the potential to improve quality achievement. A program in Bangladesh and increase efficiency because they have that gave stipends to girls substantially more autonomy than traditional public increased girls' enrollment. A similar pro- schools, which means that they are subject gram in Pakistan helped to solve the under- to fewer constraints such as bureaucratic supply of education services in urban requirements and pressure from teachers' areas by encouraging new private schools unions. In addition, in schools that are pub- to open. Another way to target is to use licly funded but privately managed, deci- funding formulas that favor students from sions about school management are made lower-income families. For instance, in at a level that is closer to the beneficiary South Africa, the government categorizes than in other schools. When governments public and private schools on the basis of make such operation contracts with private their relative poverty level and provides organizations, they are leveraging not only them with subsidies based on the level of the organization's expertise, but also its tuition and other fees that they require innovative instructional and management their students to pay. As a result, the poor- practices. Publicly funded private schools est schools receive the highest subsidies. can transform the education system from Contracts to provide education services the outset, simply by providing a wider are another kind of PPP in which the pub- range of schooling alternatives. Moreover, lic sector contracts with private providers because they must offer free education, they to educate a specific number of students in provide additional places for students who exchange for a per pupil payment. These are traditionally underserved. Conclusions 63 The United States has a highly decentral- implementation of their educational strate- ized education system and an active capital gies. Also, contracts for operational service market that invests in for-profit education attract a wide range of different kinds of management organizations and institu- private providers, which means that the tions that channel funds to education busi- supply of education becomes more diver- nesses. Consequently, the United States is sified. Colombia, Qatar, and the United the country with the most experience with States have explicitly created incentives to contracting for the private operation of attract high-performing or specialized edu- public schools. There are two types of pri- cation organizations to drive up the overall vate management of schools in the United quality of the education provided in those States--education management organi- countries. In Colombia, bidders to run zations and charter schools. Other coun- concession schools had to show that they tries are following suit. The government of already operated education institutions that Qatar introduced the independent school had scored above the average on national program in 2004 as part of an overarching examinations. Qatar allowed international decentralization reform; the management operators to bid to run schools and allowed of all public schools will be transferred to providers to make a reasonable profit as an independent operators by 2011 in order to incentive to attract bidders. promote accountability and improve aca- In many of the PPP models, decision- demic performance. Latin America has making power over school management two examples of privately managed pub- is transferred to the school itself, which lic schools. The first one is Venezuela's Fe makes the provider much more immedi- y Alegría Network, which provides free ately accountable to the user of the service education to poor communities in under- (parents, students, and local communities) served areas and receives funding from the and which tends to lead swiftly to increased government through an agreement between efficiencies in inputs and improvements in the Ministry of Education and the Venezu- service. Although privately operated pub- elan Association of Catholic Education. Fe lic schools spend less money per pupil than y Alegría schools account for 8 percent of public schools do, they are more successful total enrollments in Venezuela. In Colom- in raising their students' academic achieve- bia, the concession model was created in ment. One reason for this is that they have 1999 to provide high-quality education to more autonomy than public schools to low-income students. Concession school make decisions about pedagogical methods operators are private schools or universities and the management of their financial and that have excellent academic performance human resources. Although the concept records. of a charter school requires open enroll- Contracting out school operations can ment and free education, these schools are replicate and scale up successful practices to allowed to tailor their curricula to target bring them within reach of more students. specific populations, such as likely drop- The World Bank's World Development outs or students with a particular academic Report 2004 identifies a lack of systematic interest. learning from innovation and insufficient The United Kingdom's private finance replication of successful practices as prob- initiative model allows private consortiums lems at the basic education level. Two addi- and public authorities to become partners tional ways to give schools an incentive with the government in the construction to improve their outputs are to allow the and maintenance of education facilities. most competent operators to manage more This kind of initiative has been accompa- schools and to standardize good practices nied by a substantial growth in the global based on either on local research or proven pool of capital available for investment in examples of success on the ground. infrastructure. Infrastructure funds man- Publicly funded private schools lead to age an estimated $133 billion, 77 percent of innovation and experimentation because which was raised between 2006 and 2007 they have autonomy over the selection and (Palter, Walder, and Westlake 2008). This 64 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION model has now spread from the United endogeneity, which typically arises because Kingdom to several European countries as of self-selection. The challenge is to build well as Australia, Canada, and Egypt. the right comparison group, whose data can be compared with those involved in What do we know about public- the contracts to judge the program's effec- private partnerships in education? tiveness. This challenge exists in the case of Increasing the private sector's role in educa- all impact evaluations of any kind, but in tion through PPPs can have several benefits education it is exacerbated by the fact that over traditional public delivery of educa- self-selection comes from two sources-- tion, including greater efficiency, increased schools and students. For instance, in sub- choice, and wider access to government ser- sidy programs, schools first decide whether vices, particularly for people who are poorly to apply for the subsidy and then students served by traditional schools. Increased decide which school to attend, based partly private involvement in education, through on whether the school receives a subsidy or contracting or vouchers, may also increase not. There are six empirical strategies that the expertise and capacity of the education can be used to overcome endogeneity-- sector and has the advantage of avoiding randomization, regression discontinuity the operating restrictions faced by tradi- analysis, instrumental variables, Heckman tional public schools, such as inflexible sal- correction models, difference in difference ary scales and work rules. estimators, and propensity score-matching The main goals that governments hope (see appendix B). to achieve by contracting with the private Although only very few empirical stud- sector in education are to increase enroll- ies of the impact of PPPs exist, it is possible ment, improve educational outcomes (such to draw some useful lessons about the feasi- as standardized tests scores and dropout bility of certain contracts. It seems that the rates), and widen access to education for private management of public schools has low-income families. They also hope to had a positive impact on student test scores. reduce the costs of providing education Less is known, however, about what exactly while increasing its cost-effectiveness. it is about charter and concession schools This book has assessed the strengths of that make them perform better than other four types of contracts--vouchers, sub- schools. sidies, private management, and private Most studies have shown that the private finance initiatives--in the context of four management of public schools is effective main objectives--increasing enrollment, in a range of respects. The body of evalu- improving education outcomes, reducing ation evidence on charter schools in the inequality, and reducing costs. In terms United States has grown substantially in of enrollment, vouchers and subsidies can recent years. This research has found that, deliver very strong results, as long as the initially, students in charter schools seem private supply is adequate. However, these to score lower than their peers in public contracts may cause students to desert pub- schools on standardized tests, but after a lic schools for better-performing private period of time (usually three years), their schools. Private management and private scores increase to levels similar to those of finance initiatives presumably require large their public school peers. Evidence from initial capital investment in the construc- randomized interventions from Chicago tion of schools, which in turn may limit has shown that the positive effects of a their ability to produce substantial changes charter school education on test scores are in enrollments. Vouchers, subsidies, and concentrated in the early grades. Studies private contracts can have strong links of Colombia and Venezuela similarly con- with education outcomes. In contrast, pri- cluded that privately managed schools tend vate finance initiatives' power to influence to yield higher test scores than public insti- education outcomes is small. tutions for students at the end of their basic The main challenge involved in evaluat- education. These two studies used propen- ing contracting programs is the problem of sity score matching estimators with only Conclusions 65 limited data, and, therefore, their results The empirical literature on vouchers is should only be used with care. large and technically strong. The evidence Vouchers are associated with much con- on the other three types of contracts-- troversy. Several countries allow parents to subsidies, private management, and pri- choose to send their children to any school, vate finance initiatives--is less abundant, provide public funding for private and reli- with the evidence on the impact of private gious schools, and allocate resources to management mainly consisting of the char- schools based on their enrollment rates--in ter school literature in the United States. short, voucher-like systems. Some of these Therefore, more research on the relation- systems are more than 100 years old, such ship between PPPs and education outcomes as those in Denmark and the Netherlands, is urgently needed. Future evaluations of while others are more recent, such as those PPP models need to be rigorously designed in Chile and Sweden. Colombia's targeted from the outset. program has been the subject of extensive analysis because of its randomized design. Improving education policy Colombia's program is well targeted, effec- and regulatory frameworks tive, and efficient. It provided quality edu- Some policy changes can provide an cation to more than 125,000 students at a enabling policy and regulatory frameworks lower cost than public schools did, and for private schools in developing countries. much of this positive effect has been shown Such a framework would create the condi- to be a result of competition. On the other tions under which private schools can oper- hand, the evaluation evidence of the voucher ate effectively and efficiently, while ensuring reform of 1981 in Chile is mixed. While that education is still of high quality. some studies found the reform to have had Provide a sound basis for the private positive effects, others have challenged these school sector. In many countries, the current findings as having problems of selection climate in the education sector is hostile to bias and a lack of adequate instruments. private providers of education, particularly Furthermore, for many years following the those that are for-profit. Some governments voucher reform, overall school quality in do not allow any for-profit schools to be Chile did not improve. Things have been opened at all, while others try to limit or tax changing more recently, as there have been any surpluses that they may make. However, rapid increases in test scores. In general, in once governments recognize the benefits most universal voucher programs in Europe, that private education can yield to the sec- the availability of school choice has led to a tor as a whole, they can start by adopting a more competitive schools market, and in policy that clearly welcomes private provid- most cases this competition has led to better ers and encourages them to establish new outcomes overall, as would be predicted by schools or universities. Ideally, this policy theory. Nevertheless, there is much to learn statement would define the place of private about school choice and vouchers. providers in the national long-term educa- Two types of PPPs on which much more tion strategy to give potential investors and research effort is needed are subsidies partners the confidence to invest. (public funds given to private schools) and Consider allowing private schools to set private finance initiatives (long-term gov- their own tuition and other fees. Many coun- ernment contracts with private partners to tries and jurisdictions, including Ghana, provide school infrastructure). However, India, the Sindh province of Pakistan, the neither the lack of evidence in one area nor Philippines, and Vietnam, limit the tuition the positive results in another are reasons and fees that private schools can charge, to ignore PPPs or to embark on a large- require that they consult the governments scale expansion. Such programs should be about any increases, or regulate the distri- piloted and rigorously evaluated in differ- bution of tuition fees. These restrictions ent settings, and this study provides guid- do not encourage private providers to get ance on how to conduct better evaluations involved in increasing the supply of edu- in these important areas. cation. If, instead, governments allowed 66 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION private schools to set their own fees, this schools often have the opposite impact of would give private providers an incentive what is intended. Rather than increasing to invest in the education sector. access, improving quality, and making Consider allowing both not-for-profit and schools safer, overly restrictive registration for-profit schools to operate. Several govern- criteria, long and convoluted school regis- ments restrict the extent to which for-profit tration processes, and onerous mandatory providers can operate in the education sec- regulations can deter potential providers tor or limit the funding for not-for-profit or increase their costs so much that the private schools. However, this bias against schools become unaffordable. Alterna- for-profit provision is not universal. Pri- tively, such regulation may push schools vate for-profit schools are growing in many to operate outside the law as unregistered countries. While private schools are often or clandestine providers, meaning that the seen as catering solely to the wealthy, the government would have fewer ways to pro- reality is that for-profit schools provide a tect consumers. This can impose costs on significant number of places to the poor. consumers, and invariably these costs will Private for-profit schools come in a vari- fall disproportionately on the poor, who ety of forms, including single owner- have fewer education options. In particular, operated schools, chains that operate a governments can ensure that school regis- large number of schools, and education tration criteria are management organizations, such as Edi- realistic and achievable, while meeting son Schools. For-profit schools serve the · policy goals efficiently and effectively; full range of communities, including elite families, middle-income families, and · objective and measurable, to mini- the poor. mize discretion and limit scope for Promote and facilitate foreign direct corruption; investment in education. Foreign direct · transparent and available to prospective investment in education is small but grow- private school entrants; ing in developed countries, developing · output-focused to allow for flexible and economies, and transition economies. In diverse delivery approaches; 2005, foreign direct investment in educa- · applied consistently across various levels tion globally was nearly $3.5 billion, up of government. from just $86 million in 1990 and $401 mil- lion in 2002, and most of this investment Give subsidies to the private school sector. has been in developed countries. However, In addition to providing general invest- foreign direct investment in education ment incentives, governments can encour- remains smaller than in other sectors of the age private investment in education by economy. In 2007, it accounted for less than offering monetary or in-kind subsidies 0.1 percent of foreign direct investment in to private schools. These subsidies can the service sector. Therefore, there is scope be given up front, for example, as free or for governments of developing countries to discounted land, establishment grants, or promote and facilitate foreign direct invest- education infrastructure. Land can be espe- ment in their education systems. cially important in urban areas where it is Establish clear and objective establishment expensive. Governments can also encour- criteria and streamline processes for register- age private investment by facilitating work ing private schools. Many countries limit the visas for foreign teachers, management, and scope for new providers to enter the edu- technical staff. It is important to ensure cation marketplace. Many of these regula- that private schools have sustainable fund- tions are aimed at protecting consumers, ing to underwrite their effective operation which is a laudable objective. Establishing and to minimize corruption in the delivery minimum standards can help to ensure the of services. Broadly speaking, governments quality and safety of private sector provi- should preside over school funding systems sion while still protecting consumers from that are integrated, neutral, responsive to unscrupulous operators. However, poorly the changing needs of schools, and tar- designed registration criteria for private geted to low-income families. Ideally, the Conclusions 67 funding system should have the following for governments to put stringent require- characteristics: ments on education providers to disclose information about their operations. This · Public and private schools should be could be done by funded within the same system. · Demand-side financing techniques · requiring schools to disclose information should be used where necessary. to regulators and the general public, · Funding for schools should be targeted · introducing a system for collecting and to factors that pose barriers. disseminating information from schools on a number of specific indicators, · The criteria for receiving funding need introducing school reviews to collect to be transparent, publicly available, and · information, and easily understood. · creating independent review and accred- Ensure that PPP contracts are flexible iting agencies. enough for private providers. The key to implementing successful PPPs is ensuring Implementing education public- that the private partners are given consider- private partnerships in developed able flexibility in terms of how they deliver and developing countries the service for which they are being con- tracted. The government should spell out Good design, while important, is not suf- its required outputs and performance stan- ficient to ensure the success of a PPP in dards and set penalties for failing to achieve education. It must also be effectively and them and rewards for achieving them, but efficiently implemented. Weak implemen- thereafter, they should leave providers to tation of PPPs can expose governments decide for themselves how best to deliver the to substantial financial and policy risks, required outputs to the specified standard. but governments can take several actions Providers must have as much management to improve the way in which PPPs are freedom as possible, especially in staffing implemented. and employment and budget allocations as Employ a transparent, competitive, and well as over the curriculum. multi-stage process for selecting private Improve information flows and establish partners in PPPs. A key element of effec- an effective quality assurance system. A key tive contracting is a transparent and com- weakness in many countries is the lack of petitive bidding process. Bidding for service available information on the private educa- delivery contracts such as school manage- tion market. This is especially important ment initiatives or private finance initia- given the growth of private education in tive contracts should be open to all private many countries and the wide variations in organizations, including for-profit and price and quality. Many countries have only not-for-profit providers. Contracts should limited information on the fees that they be open to any local, national, and inter- charge, the programs that they offer, and national organizations that may wish to the qualifications of their staff. Even the bid for the opportunity to operate a pub- regulatory authorities have little informa- lic school. The contracting agency should tion on either the size or nature of the pri- select providers by means of a multi-stage vate school and tertiary education sectors. process, and these stages should consist of Some countries publish exam scores on a · clarifying requirements, including objec- school-by-school basis (for example, the tives, services, and outcomes; Philippines and Uganda), while others have · developing a procurement strategy and found innovative ways to provide consum- identifying a technically strong procure- ers with information on the performance of ment team; schools and tertiary education institutions. writing the request for proposals; Well-informed consumers and regulators · are an important component of any regula- · inviting expressions of interest; tory framework for education. One way to · conducting contract prequalification ensure that consumers are kept informed is checks; 68 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION · interviewing bidders, assessing propos- conditions of the contract. Performance als, and negotiating with a shortlist of measures are even more important when bidders; they are prerequisites in the contract to the · selecting the preferred bidder and award- provider's compensation. The performance ing the contract; measures and standards in each case must · advertising the result of the selection be carefully designed because they can process; introduce perverse incentives and lead to undesirable outcomes. For example, it is · commencing the service. usually desirable to bear in mind the fol- Split the purchaser and provider roles lowing issues: within the education administrative agency. · A heavy emphasis on test scores may PPPs function better when the education crowd out focus on other skills. department separates its purchaser role from its provider role. In this situation, · Setting measurable outcomes may lead the ministry's policy and regulatory func- providers to pay too little attention to tions are kept separate and distinct from its achieving other desirable outcomes. service delivery and compliance functions. · High scores in external tests may pro- If the same government department is vide schools with an incentive to select responsible for both purchasing and provi- only strong students. sion (and regulation) of education, there is Develop an effective communications a risk that it will be biased in favor of public strategy. Efforts to involve the private sec- schools because private sector competition tor in education often face concerted oppo- can threaten the viability of some public sition from rival political parties, labor schools. unions, the media, the public at large, and Build the capacity of the contracting specific interest groups. Therefore, a crucial agency. The key to successfully designing component of any PPP in education is an and implementing PPPs is to ensure that effective, strategic (as opposed to piecemeal the government agency responsible for PPPs or ad hoc) communications strategy, as this has the resources, information, and skills can substantially reduce political risk and needed to design, develop, and manage the be an effective way of promoting a PPP complex contracting processes that under- initiative. lie PPP programs. In particular, the recent Introduce a framework for evaluating shift away from input controls to output- program outcomes. PPPs should be accom- based contracting means that government panied by a well-designed, rigorous evalu- agencies must develop their capacity to ation. Although a wide range of PPPs exists · assess services to determine when con- around the world, there is a lack of rigor- tracting is appropriate; ous evidence on the impact of these part- · design, negotiate, implement, and moni- nerships on academic outcomes and other tor education service contracts; education indicators. This is especially the true for PPPs outside the United States and · develop legislation that supports a com- for nonvoucher programs. petitive and transparent system; · develop appropriate quality assurance Recommendations mechanisms. Where appropriate, PPPs can increase access Establish appropriate performance mea- and improve quality in education by giving sures and include performance incentives students choices and by putting competi- and sanctions for failing to perform in PPP tive pressures on schools. Public funding contracts. Establishing appropriate perfor- of private schools is justified because dis- mance measures is critical in the design of advantaged students will benefit from the all contracts. Performance measures are opportunity to enroll in schools appropri- necessary for determining whether the ser- ate for their needs. Nevertheless, ensuring vice provider has met the agreed terms and academic quality in this kind of education Conclusions 69 system is a persistent challenge. Experience tal and preferential loans to improve with PPPs in various countries yields the infrastructure and other critical inputs; following recommendations: and providing technical assistance and quality certification to enhance financial · Include output specifications that define management, instructional delivery, and performance standards and facilitate the school leadership. measurement and tracking of quality and · Establish a specialized group of authori- school efficiency. Performance indicators ties to manage PPP programs and the flow can be quantitative, such as standardized of funds from the government to private tests or enrollment figures, or qualita- schools, and to enforce qualifying criteria tive, such as school and parent surveys and regulations. or school inspections. It is particularly important to include quality indicators The role of the World Bank Group that will encourage improvements in International organizations can be vital in the performance of private schools and, promoting high-quality private education equally important, to reinforce them in several ways. They can provide "early with appropriate supervision. stage" equity and loan capital to finance · Define operating requirements and perfor- investments in the private education sector. mance standards that private schools and Private schools sometimes find it difficult operators should follow. Private schools to access short-term (five to seven years) should meet eligibility criteria to receive investment capital because private equity is public funding (such as infrastructure generally not interested in such a short time and staff requirements), follow national horizon. International lenders can help to curricula, and meet performance increase the attractiveness of the private benchmarks. education sector as an investment target. · Reward innovation and quality improve- They can also work with private sector ments. One way to reward schools is banks to mitigate some of their investment to provide monetary awards for good risks. Also, they can increase the capacity of performance. Conversely, sanctions for both private banks and the education sec- poor performance should include the tor and help countries to create regulatory revoking of any subsidies. frameworks that enable the development of · Help private schools to deliver high-qual- private education in developing countries. ity education and accompany voucher The main focus of most education proj- programs with capacity-building inter- ects supported by international organiza- ventions. Some private schools lack the tions is on improving public sector schools capacity to improve the quality of the and tertiary education institutions. How- education that they provide because ever, the International Finance Corporation their teachers lack qualifications, the (IFC) also provides financial and techni- schools lack the resources to buy mate- cal support to private education projects, rials and textbooks, and school manage- including operations that provide educa- ment is not aware of the most effective tion entrepreneurs in Ghana with access teaching techniques and management to capital, and it has recently launched a processes. Some ways to help build microcredit program in Kenya, which tar- this kind of capacity in private schools gets private school providers and includes a include increasing their access to capi- technical assistance facility component. a p p e n d i x Public-Private Partnerships: A Program and Policy Descriptions by Contract Type Country Program Program Description Objective Partners Educational Services 1 Argentina Public funding for Private school coverage at the primary and secondary level Increase access Local education private schools was approximately 26.4% in 2005. Local education authorities authorities, public provide public resources to support teacher salaries, totally or schools partially, in 65% of private schools. In 2005, the total amount of public resources transferred to private schools represented 13% of local education budgets, which is equivalent to 0.47% of GDP. The Law of Education Finance does not establish a concrete formula for transferring public resources to private schools. 2 Australia Public funding Independent schools receive both recurrent and capital Increase access Private, religious, for independent funding, with the former being the larger of the two and improve quality and public schools, schools components. Recurrent funding is provided both in the form government of of general per-student grants and specific funding aimed at Australia targeted groups of students. Payments to individual schools are based on a sliding scale that depends on a school's socioeconomic status. In 2007, enrollments in nongovernment schools represented over 33% of total enrollments. 3 Bangladesh Female Secondary Scholarships cover the direct cost of girls' secondary Increase access Government of School Assistance education. Once girls have satisfied a set of requirements, Bangladesh, Asian Project the corresponding schools are paid the entire tuition amount. Development Bank, Additionally, girls receive a stipend expected to cover 50% the World Bank, of school fees. Other components of the project include public and private curriculum reform, instructional materials development, schools teacher training, the improvement of school infrastructure, and institutional capacity building. 4 Bangladesh Subsidies The government subsidizes at least 9 teachers at 90% of the Increase access Nongovernment government base teachers' salaries at community-managed, schools, not-for-profit, nongovernment schools. Government subsidizes government of increases in enrollment by paying for additional teachers as Bangladesh long as the school meets the state criteria. Subsidies work like a quasi-voucher because they are tied to increased enrollment. 5 Belgium Voucher scheme The vast majority of approved private schools receive grants Increase access Public, private from the government on the same basis as public authority and improve quality schools, and the schools. Staff salaries are paid directly by the government and state funding is provided for other operating expenses. Assistance for the development of capital infrastructure is provided via grants, loan guarantees, and favorable interest rates. Grant- aided schools must meet minimum academic and operating standards but have the freedom to choose their educational plan. In 2004, more than 50% of enrollments at the basic and secondary levels were private. Most private schools have a religious association. 6 Canada Public funding for Accredited independent schools receive basic grants equal to Increase access Provincial (Alberta) nongovernment 35% of public school costs. Subsidized schools must comply and improve quality government schools with operating requirements, use the Alberta Program of of Alberta, Studies, and follow the same accountability requirements as independent public schools. However, they have the freedom to establish schools tuition fees and admission policies. continued 71 72 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION Country Program Program Description Objective Partners Educational Services 7 Chile Voucher scheme This voucher scheme involves the government paying a Increase access, Central monthly fixed fee to subsidized private schools according to improve quality, government, their enrollment numbers. There is price discrimination among and support the subsidized private subsidized schools depending on their location and decentralization of private schools, level of education. Families have no restrictions on school education municipalities selection, but private subsidized schools are not compelled to accept any student. Subsidized schools must meet minimum requirements but enjoy management flexibility. Vouchers are paid directly to private schools. In addition, the government gives nonportable subsidies to public schools in low-income areas. 8 Colombia Plan de Ampliacion Vouchers are made available to students from low-income Increase access Secretaries of de la Cobertura families who had been attending public schools but who to secondary Education at the de la Educacion had been accepted into a private school. Vouchers were schooling for departmental and Secundaria renewable subject to satisfactory academic performance. children from poor municipal level and families private schools 9 Colombia Contract schools Local governments contract private schools to deliver public Increase access to Territorial education services for a determined number of students in an quality education entities, private academic year. The private contractors assume all or some of services for low- schools, Ministry the costs involved in educating a student and the government income students of Education, reimburses them on a per-pupil basis in accordance with Secretaries of the pre-agreed contract. In 2004, 11.2% of the students in Education eight cities in Colombia were enrolled in contract schools. Secretaries of Education establish the number of places needed for public school students, develop a pool of bidders for the selection of education service providers, and process contracts. A list of eligible students is presented to each private school selected. Only in one city, Cali, are parents allowed to select their children's school. 10 Côte d'Ivoire Contracts for The government gives a payment to lower and upper Increase supply of Government of Côte education services secondary private schools for each public student that they education to meet d'Ivoire, private enroll. Schools must be "chartered" to take on additional student demand schools students, and placement depends in part on the educational performance of the school. Subsidies vary with school location and are loosely tied to the number of students enrolled. The number of students in the private school sponsorship program was 223,000 in 2001 (an increase from 116,000 in 1993). 11 Czech Voucher scheme All schools, public and private, receive public funding based Increase access Ministry of Republic on the number of students enrolled. The funding consists of and improve quality Education, nonstate (1) base funding equal to 50% of state school funding, and (2) schools supplementary funding based on quality, assessed on the basis of explicit criteria. 12 Denmark Voucher scheme Private schools, some of them organized by parents, receive Increase access Government of grants from the central government corresponding to and improve quality Denmark, private approximately 80% of their total expenditures. Schools must schools, parent meet centrally determined standards, and teachers must boards possess the required qualifications. The grants take into account property-related and operating costs and vary across schools depending on their size, the age distribution of their students, and the seniority of their teachers. 13 Gambia, The Scholarships The government funds full scholarships that cover the costs Increase access, Private schools, of tuition, books, and examination fees to one-third of girls retention rates, and Boards of in upper basic and secondary private schools with low girls' enrollment Governors, and the enrollment in the most deprived regions. They also cover rates government of The tuition and examination fees for 10% of girls who excel Gambia in science, math, and technology at the upper basic and secondary school level in less deprived regions. Public-Private Partnerships: Program and Policy Descriptions by Contract Type 73 Country Program Program Description Objective Partners Educational Services 14 Guatemala Scholarships/ This program gave vouchers to girls from low-income Increase access Public schools, Eduque a la Nina communities to induce them to enroll in the first, second, and retention rates parent committees, and third grades. It was implemented in 12 rural communities school teachers, and involved approximately 800 girls over 2 years. Target Asociacion communities were chosen because they had the largest Eduquemos a la differentials between male and female school attendance and Nina graduation rates. The voucher was renewed provided the girl was promoted to the next grade. 15 Haiti Haiti Education To increase access, the project will fund per student subsidies Increase access, Nonpublic For All Adaptable disbursed to school management committees of accredited improve quality, and primary schools, Program Grant nonpublic schools to subsidize the tuition fees and educational reduce inequity Department of materials for out-of-school children. The project will also Administrative train more new teachers and contract qualified NGOs and Affairs, school school inspectors to strengthen the capacity of school management management committees. To improve quality, the project committees, the will pilot student-centered learning programs adapted for National Education multi-grade classrooms, support student health and nutrition Partnership Office, programs, and strengthen private schools' capacity to deliver the National School early indicators (grade 2) of learning outcomes, particularly Feeding Program, basic literacy. the Department of Private Education, the Fast Track Initiative, the World Bank 16 Hungary Voucher scheme The guiding principles of the Hungarian system are high levels Improve quality, Private schools, of local control, school self-management, and acceptance of increase efficiency, autonomous local school competition. Private institutions are entitled to receive and support governments per-pupil grants from municipalities. Funding for private decentralization schools is formula-based and dependent on the number of students enrolled. 17 India System of Grants to aided schools account for a substantial proportion Increase access Private schools, government grant- of the education budget. Any recognized private school can central government in-aid to privately apply for government grant-in-aid, and once granted aided managed schools status, it receives block grants in the form of the payment of all teacher salaries. Teachers are paid out of school revenues and are accountable to fee-paying parents and school managers. Teachers are managed at the school level. There is a high inter-state variation in grant-in-state aid. 18 Italy Voucher scheme In 9 (of 20) regions, the government subsidizes tuition fees Improve quality and National at private primary and secondary schools. There are (1) increase choice and regional income-targeted vouchers that offer partial reimbursement of governments, private school tuition, and (2) voucher programs that provide private schools a fixed payment, conditional on school performance and family income. Schools receiving vouchers must be legally recognized by the government. 19 Korea, Subsidies Under the "leveling policy for secondary education," Improve quality and Central Republic of elementary school graduates are randomly assigned to either reduce inequity government, private public or private middle schools in their district of residence. in the school secondary schools Private schools are required to meet government requirements environment and are not allowed to charge tuition fees. As a result, 95% of private school costs are subsidized by government, including teachers and operations. 20 Mauritius Subsidies Most private schools in Mauritius are funded by the Increase access The Mauritius government and are non-fee-paying (some have a religious and improve quality Educational orientation). The government pays the wages of the staff of Development private schools and their running expenses based on a grant Company, the formula. Other indirect costs are met by students and their Development Bank families (such as books, transport, and uniforms). In 2005, 66% of Mauritius, the of students in the secondary mainstream and 58% of those in Private Secondary prevocational education were enrolled in private schools. The Schools Authority, program is supplemented by incentive grants such as a loan private schools scheme for private schools with preferential interest rates to upgrade their infrastructure facilities. continued 74 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION Country Program Program Description Objective Partners Educational Services 21 Netherlands Voucher scheme The government funds public and private schools on an equal Increase access The Dutch footing. Institutions are given considerable freedom to decide and improve quality government, private how to allocate their resources, although they must meet the schools government's performance requirements. School choice is promoted in order to increase competition between schools, and most students attend private schools (by 2004, 69% and 83% of enrollments at the primary and secondary level). 22 New Zealand Targeted Individual The Targeted Individual Entitlement program sought to assist Increase access Participating Entitlement children from low-income families to attend private schools, and improve quality schools, the to give more choice to parents with limited options, and to government of New increase educational attainment among low-income families. Zealand Low-income students received a subsidy (110% of the average cost of education at a state school) to attend private schools. This was a pilot program that was abolished in 2000. 23 New Zealand Public subsidies Independent schools receive subsidies of about 25% to 35% Increase access Independent for independent of the average per pupil cost in public schools. Subsidies and improve quality schools, the schools are enrollment-based and vary by grade level. Schools must government of New be registered. Subsidized schools can be for-profit, they do Zealand not have to use the national teacher's contract, and are not required to teach the national curriculum. 24 New Zealand Alternative The programs seeks to provide alternative education programs Increase access Single private education/per-pupil to students alienated from the education system and to and retention rates schools, funding prepare students to return to mainstream secondary education consortiums of or move onto tertiary education once they reach 16 years private schools, of age. The program contracts out the delivery of education not-for-profit in nonformal settings to not-for-profit, community-based community-based organizations or for-profit educational providers. organizations, for- profit educational providers, the government 25 New Zealand Integrated Schools Former private schools, mainly owned by private organizations, Increase access Integrated schools have been integrated into the public system and receive and the government recurrent funding equivalent to that received by public of New Zealand schools. Integrated schools are subject to state regulations and are not allowed to charge tuition but are allowed to charge fees to cover infrastructure expenditures. Integrated schools represented 11% of enrollments in New Zealand in 2007. 26 Pakistan Urban Girls In this pilot program launched in 1995, the government paid a Increase girls' Private schools, (Balochistan) Fellowship declining subsidy to private schools over a three-year period to enrollment in the Government enroll girls from low-income families, in addition to a tuition fee schools of Balochistan, per girl per year. The subsidy was paid directly to the school parent education and was limited to 100 girls. committees 27 Pakistan Basic Education Program that supports the establishment of new private Increase access to Private schools, the (Balochistan) Support Project schools by providing per-student subsidies to Private School low-fee high-quality World Bank, rural Implementation Partners (PIPs) for up to four year. Schools private education community schools, are able to charge top up fees of up to PRs300 per month. the Government of Additionally, PIPs receive per-student subsidies for facilities Balochistan and material costs. New schools will participate in the program if they have over 50 students and there is no public school in a radius of 20 kilometers. 28 Pakistan Financial The Punjab Education Foundation pays a subsidy to Improve quality Private eligible (Punjab) Assistance per participating private schools on a per pupil basis. The schools and increase schools, Punjab Child Enrolled Basis cannot charge fees on top of the per-student subsidy paid. productivity Education Program (FAS) Participating schools must meet eligibility criteria (in terms Foundation in enrollment, student composition, physical infrastructure, geographical location, the capacity to deliver quality education, and management) and register with the district government. Subsidies are paid directly to the school. The Punjab Education Foundation provides professional development support for the FAS schools. Public-Private Partnerships: Program and Policy Descriptions by Contract Type 75 Country Program Program Description Objective Partners Educational Services 29 Pakistan Pilot Education The Pilot Education Voucher Scheme will give education Improve quality Punjab Education (Punjab) Voucher Scheme vouchers to children from urban slums in Lahore. Its design will of education and Foundation, eligible include strategies to support school improvement and quality encourage girls' schools education. The vouchers will be nontradable. enrollment in schools 30 Philippines Educational Service The government contracts with private high schools to enroll Increase access Certified private Contracting (ESC) students in areas where there is a shortage of places in public and improve quality schools, Fund schools. Assistance is given only to students at institutions that for Assistance to charge low fees and serve low-income families. The program Private Education, also includes a certification aspect. In 2008­09, around 476,776 and the Department students were subsidized under the ESC program. of Education 31 Qatar Voucher scheme The voucher system allows parents to choose among Improve quality The Supreme independent, private accredited, and international schools. and increase Education Council, The schools remain autonomous and are held accountable educational independent or for student learning. The variety of schooling options will, choices to accredited voucher over time, give parents a growing range of different kinds of raise academic schools schools to choose from when selecting the best school for achievement their children. 32 Senegal Scholarships Communaute Urbaine de Dakar is an agglomeration of five Increase access Municipalities, municipalities that offers scholarships to students studying in and improve quality private schools both private and public schools inside or outside of Senegal. (national or international) 33 South Africa Subsidies State funding to public and private schools is organized Increase access Public and on a quintile system, in which schools are divided into five and reduce inequity private providers categories according to the poverty levels that prevail of education, in the areas that they serve. Schools in the lowest two provincial quintiles receive full funding from the government. Private education schools requesting funding must provide evidence of sound departments management and financial records and allow unannounced inspections by the provincial education department. 34 Sweden Voucher scheme Municipalities give capitation grants to private and public Increase access The National schools on an equal footing. They have more authority over and improve quality Agency for their own (public) schools than over private schools but Education, have full financial responsibility for the whole school system. municipalities, Independent schools often have a particular academic private primary and focus such as religion, art, sports, or music. Schools must secondary schools be approved by the National Agency for Education and meet certain regulatory requirements in order to be eligible for government funding. 35 Thailand Subsidies The government of Thailand provides monthly subsidies to Increase access Private and public private schools on a cost-per-student basis. This is a major and improve quality schools, the central source of income for most schools. Private schools are government, allowed to charge fees similar to tuition fees to improve the foundations, quality of education; additional fees are allowed for meals, and parents' transportation, health inspection, and other extras. Private associations schools providing basic education can qualify for state- subsidized loans to build new school buildings or to renovate old ones. The government also has a revolving fund for private schools, which offers 4 % interest loans with a repayment period of 10­15 years to schools that can offer collateral. 36 Uganda Universal The government of Uganda subsidizes 430 private secondary Increase access Private secondary Secondary schools serving approximately 56,000 students (as of 2008) schools, the Education Program in order to attain universal secondary education. The Ministry of Ministry of Education chooses the participating schools and Education and a memorandum of understanding is signed with individual Sports in Uganda private schools to ensure that they comply with the policy's implementation guidelines. continued 76 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION Country Program Program Description Objective Partners Educational Services 37 United Assisted Places The Assisted Places Scheme was introduced in 1980 and Increase access The UK Kingdom Scheme provided financial support for poor students with high and choice government, public academic achievement to attend private schools. The program and private schools served about 30,000 students in 1993/94 and was abolished in 1997. Additionally, a demand-driven funding system for public schools was introduced in 1998. Seventy-five percent of school funding is allocated based on age-weighted student numbers. Public schools cannot charge tuition fees. Schools are granted greater autonomy in management and administration and parents have free choice, although in practice competition is limited due to a rule that prevents the establishment of new schools as long as there are available places in public schools in the area. Only 5% of primary enrollments were in private schools in 2004. 38 United States Voucher scheme The Targeted Voucher System in Milwaukee gave vouchers Increase access School districts and (Milwaukee to kindergarten through 12th grade students from low-income and improve quality private schools and Florida) families to enable them to attend accredited secular or religious private schools. Private schools must administer nationally recognized tests and cannot charge fees higher than the voucher amount, though they may charge for extra- curricular activities. The number of vouchers was capped at 22,500 in 2007 (up from 15,000 the previous year). The McKay Scholarships Program in Florida offers parents of special needs students who are dissatisfied with their children's existing schools the chance to transfer them to another public school. During the 2007­08 school year, the program provided 18,919 special needs students in Florida the opportunity to attend a participating private school. $119.1 million was paid to scholarship program participants in 2006­07. 39 United States Voucher scheme This voucher program, which ended in 1995, was targeted to Increase access The government of (Puerto Rico) low-income families who could freely choose any school in any and improve quality Puerto Rico, public school district. The program covered 2,000 students in 1993 and and private schools over 14,000 in 1994. To qualify for a voucher, the student's family income had to be less than $18,000 per year. Any licensed or accredited school was entitled to receive vouchers. 40 Venezuela, Venezuelan The Ministry of Education and Culture gives subsidies to Increase access The Ministry of R. B. de Association of private schools located in low-income urban areas and and improve quality Education, Culture Catholic Schools indigenous communities, and to vocational schools and and Sports, AVEC, (AVEC) schools unable to meet more than 85% of their operational private schools costs. While subsidies are nonconditional, schools must under the AVEC provide financial and management reports on an annual basis network (some to government. are Fe y Alegria schools) Supplemental and Support Services 41 Australia Tutorial Voucher Parents and caregivers with children who score below a Improve quality and Brokers, private Initiative national reading benchmarking were eligible to receive a literacy skills tutors, the federal tutorial voucher valued up to $700. The voucher paid for a government pre- and post-tuition assessment and a number of hours of reading tuition delivered one-to-one outside school hours. The initiative is administered by brokers who are responsible for contracting tutors, confirming student eligibility, providing parents/caregivers with a choice of tutors, and managing the initiative's administration. The initiative accounted for $20 million of government funding. 42 Brazil Pitagoras/ The Pitagoras Network of Schools (PSN) works with a range Improve quality and Independent Corporate of independent schools, most of which are private and charge increase efficiency schools, Pitagoras, Sponsored Schools tuition. Schools enter into a yearly contract with PSN, which in management corporations then provides textbooks for all grades and students, teacher training services, and professional support. In return, schools have access to available services and are expected to participate in network activities. Public-Private Partnerships: Program and Policy Descriptions by Contract Type 77 Country Program Program Description Objective Partners Supplemental and Support Services 43 Colombia Escuela Nueva The Escuela Nueva Foundation delivers teacher training; Improve quality and The Ministry Foundation designs curricula, textbooks, and educational materials; provide technical of Education, conducts research on pedagogical approaches; and advises assistance to Fundacion Volvamos governments on how to adapt the Escuela Nueva model in schools and a la Gente, public schools. Escuela Nueva is a multi-grade rural school governments external funders, model that promotes leadership and cooperation between the other national administrative body, teachers, community, parents, and students. governments 44 India Computer The NIIT, a global IT corporation, works with the state Improve quality NIIT, four state education in governments of Tamil Nadu, Karnataka, West Bengal, and of computer governments, government Andhra Pradesh in infrastructure creation, systems integration, education and government schools facilities management, education delivery, and teacher computer-aided schools training, thereby providing quality computer education and education computer-aided education to thousands of schools. Many and increase of the classrooms have become NIIT centers, open to the operational school children and teachers during the day, then used by the efficiency franchise holder in the evenings. 45 Pakistan Quality The Aga Khan Education Foundation provides a wide variety Improve quality and Aga Khan Education Advancement of school improvement programs including student-centered increase efficiency Services, the and Institutional education, computers in the classroom, and preschool government, the Development in education. The objective of the program is to strengthen the Directorate of private schools capacity of low-cost private schools to improve the quality of Private Education the education delivered to poor communities. and Private Schools, private schools 46 Pakistan Cluster-based The Cluster-based Training of Teachers program provides Improve quality 23 organizations, (Punjab) Training of professional development for private school teachers with including 13 NGOs Teachers a focus on primary education. Training programs focus on and private teacher developing teachers' knowledge of content rather than on training institutions, pedagogical approaches. The training is provided to clusters of public schools, the approximately 7 to10 schools and 30 to 35 teachers. Teachers Punjab Education are paid an allowance to attend the training that covers Foundation transportation and other costs. Training can be contracted out to the Punjab Education Foundation or to another private provider. 47 Pakistan Quality Assurance The Sindh Education Foundation developed a quality Improve quality The Department (Sindh) Resource Center assurance certification program to categorize schools of Education in terms of their quality as a means of informing parents' & Literacy, the schooling decisions. The program also provides tailored quality government of enhancement support for public, private, and community/NGO Sindh, public, schools, including the training of teachers and school staff. private, and community/NGO schools, the Sindh Education Foundation 48 United States Supplemental Under the No Child Left Behind Act, private providers can Improve quality State educational educational be contracted to provide additional academic instruction and the academic agencies, public services in schools that have not made adequate yearly progress achievement of and private in increasing student achievement for three years. State low-performing schools, school education agencies identify organizations, whether public or students districts, public and private, that qualify to provide these services. private providers Operational and Management Services 49 Argentina Fe y Alegria Jesuit-controlled NGO that operates formal preschool, primary, Improve quality of Ministries of Network secondary, and technical education (primarily formal primary education provided Education, education). to poor people foundations, international agencies, civil society, communities 50 Bolivia Fe y Alegria Jesuit-controlled NGO that operates formal preschool, primary, Improve quality of Ministries of Network secondary, and technical education (primarily formal primary education provided Education, education). to poor people foundations, international agencies, civil society, communities continued 78 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION Country Program Program Description Objective Partners Operational and Management Services 51 Brazil Fe y Alegria Jesuit-controlled NGO that operates formal preschool, primary, Improve quality of Ministries of Network secondary, and technical education (primarily formal primary education provided Education, education). to poor people foundations, international agencies, civil society, communities. 52 Canada Charter schools A small number of charter schools (capped at 15) operate in Improve quality The Provincial (Alberta) the province of Alberta in a similar manner to charter schools and increase government of in the U.S. They have more management flexibility than public management Alberta, school schools. efficiency and boards, and private accountability operators 53 Colombia Concession schools The management of public schools is turned over to private Improve quality Secretaries schools with proven track records of delivering high-quality and increase of Education, education for a period of 15 years under performance-based management, associations of contracts. efficiency, and private educational accountability providers 54 Colombia Fe y Alegria Jesuit-controlled NGO that operates formal preschool, primary, Improve quality of Ministries of Network secondary, and technical education (primarily formal primary education provided Education, education). to poor people foundations, international agencies, civil society, communities 55 Dominican Fe y Alegria Jesuit-controlled NGO that operates formal preschool, primary, Improve quality of Ministries of Republic Network secondary, and technical education (primarily formal primary education provided Education, education). to poor people foundations, international agencies, civil society, communities 56 Ecuador Fe y Alegria Jesuit-controlled NGO that operates formal preschool, primary, Improve quality of Ministries of Network secondary, and technical education (primarily formal primary education provided Education, education). to poor people foundations, international agencies, civil society, communities 57 El Salvador Fe y Alegria Jesuit-controlled NGO that operates formal preschool, primary, Improve quality of Ministries of Network secondary, and technical education (primarily formal primary education provided Education, education). to poor people foundations, international agencies, civil society, communities 58 Guatemala Scholarships/ Vouchers were given to girls from low-income families to Increase access Public schools, Eduque a la Nina induce them to enroll in the first, second, and third grades of and retention rates parent committees, education. Implemented in 12 rural communities and involved teachers, and approximately 800 girls over two years of age. The target the Associacion communities were chosen because they had the greatest Eduquemos a la differences between male and female school attendance and Nina graduation rates. Each girl's voucher was renewed conditional on the girl's promotion to the next grade. 59 Honduras Fe y Alegria Jesuit-controlled NGO that operates formal pre-school, Improve quality of Ministries of Network primary, secondary, and technical education (primarily formal education provided Education, primary education). to poor people foundations, international agencies, civil society, communities Public-Private Partnerships: Program and Policy Descriptions by Contract Type 79 Country Program Program Description Objective Partners Operational and Management Services 60 Nicaragua Fe y Alegria Jesuit-controlled NGO that operates formal pre-school, Improve quality of Ministries of Network primary, secondary, and technical education (primarily formal education provided Education, primary education). to poor people foundations, international agencies, civil society, communities 61 Pakistan Adopt-a-School Governments hand over control of under-utilized/failed schools Improve quality of ITA and other Program to the NGO Itara-e-Taleem-o-Agahi (ITA), which takes them education civil society over and provides free schooling. ITA's role is to provide organizations, teacher training, formulate exercises, and make infrastructure government improvements. ITA formulates a school council that is held schools, the responsible for monitoring and maintaining school facilities. Education A memorandum of understanding is formulated between ITA Governorate and the Department of Education, Punjab. No teachers or staff members are removed from their positions. 62 Pakistan Pakistan Railways Pakistan Railways (PR) contracted Beaconhouse to operate Improve school PR, Beaconhouse Schools schools for the children of their employees. School fees were management Schools minimal (PRs25) with an option to enroll students for a fee. Staffing decisions and hiring remained in the control of PR. 63 Pakistan Management Cooperation for Advancement, Rehabilitation, and Education Improve CARE, public of Government (CARE)--a local NGO--takes over the management of management of schools, Schools in Lahore public schools by hiring internal, external, and academic schools and quality government head City and Sarghoda coordinators who work with school staff; supervise the of education teachers, academic performance of CARE and government teachers; and monitor coordinators, teacher attendance, performance, and test administration. and internal CARE employs and pays 1,000 teachers (one-third) and the and external government employs 2000 (two-thirds). CARE also improves coordinators to and provides additional infrastructure. monitor school performance 64 Pakistan Quality Education The National Rural Support Program, a semi-autonomous Improve quality National Rural (Punjab) for All not-for-profit agency, took over the management of 48 of education in Support Program, public schools through a 5-year contract with the district primary schools, public schools, government. NRSP is responsible for the operational reduce number private sector, budget and maintenance and has authority over staff. The of dropouts, and district government remains responsible for capital works. The PPP and increase governments is governed by a memorandum of understanding--a 5-year enrollments management contract--that sets out performance targets and accountabilities. 65 Panama Fe y Alegria Jesuit-controlled NGO that operates formal pre-school, Improve quality of Ministries of Network primary, secondary, and technical education (primarily formal education provided Education, primary education). to poor people foundations, international agencies, civil society, communities 66 Paraguay Fe y Alegria Jesuit-controlled NGO that operates formal pre-school, Improve quality of Ministries of Network primary, secondary, and technical education (primarily formal education provided Education, primary education). to poor people foundations, international agencies, civil society, communities 67 Peru Fe y Alegria Jesuit-controlled NGO that operates formal pre-school, Improve quality of Ministries of Network primary, secondary, and technical education (primarily formal education provided Education, primary education). to poor people foundations, international agencies, civil society, communities continued 80 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION Country Program Program Description Objective Partners Operational and Management Services 68 Qatar Independent ISs are government-funded (based on the number of students Improve Private operators, schools (IS) enrolled), privately managed schools. ISs can be newly management of the Supreme established or existing public schools. Operators hire teachers schools and quality Education Council, and establish their own personnel policies. Contracts are of education and the Ministry of for three years and are renewable conditional on school increase school Education of Qatar performance. Under the IS system, the government continues to independence assume the cost of education for eligible students and provides funds directly to the school. Operators can charge a fee to students who are not eligible for a subsidy. Operators are allowed to make a reasonable profit. Funding mechanisms include per pupil operating rate, start-up funding, and possible special grants. 69 United Education Action Local councils take bids from private organizations to run Improve quality, Schools, local Kingdom Zones failing schools and to manage Education Action Zones (local tackle social education clusters of 20 primary, secondary, and special schools). exclusion, and authorities and Education Action Zones were intended to run for an initial promote innovation other local period of 3 years with the possibility of extending to 5 and greater organizations, years, after which they should have been transformed into cooperation the business "Excellence Clusters." between schools community, higher education institutions 70 United Academies Independent schools sponsored by businesses, faith-based Improve quality and Private enterprises, Kingdom groups, or voluntary groups working in partnership with the increase efficiency charities, central government and local education partners. Funding philanthropists, comes from the Department for Education and Skills through a the Department parity of funding with school operators. Private organizations of Education and become sponsors of academies and contribute up to £2 million Skills towards their creation and are permitted to engage in trade in order to generate profits. 71 United States Contract schools Contract schools are privately managed but remain publicly Improve quality Local school owned and funded. Typically, private operators are brought in and increase boards, education to manage the worst-performing schools. Students do not pay management management fees to attend these schools. Private sector operators must efficiency and organizations, meet performance benchmarks and are paid a fixed amount accountability private educational per student, usually equivalent to the cost in the public sector, providers and a fixed management fee. Teaching and other staff continue to be employed by local authorities. 72 United States Charter schools Charter schools are publicly funded, privately run, secular Improve quality District school public schools of choice that operate free from the regulations and increase boards, universities that apply to public schools. Charters are granted for three management or other authorizing to five years. Schools must meet academic benchmarks and efficiency and agencies in charge standards on curriculum and management or the contracts can accountability of granting charters be revoked. In 2007­08, there were over 4,000 charter schools, (depending on with enrollment rates of some 1.2 million. local regulations). Managing agents include local communities, for- profit, and not-for- profit providers 73 Venezuela, Fe y Alegria Jesuit-controlled NGO that operates formal preschool, primary, Improve quality of Ministries of R. B. de Network secondary, and technical education (primarily formal primary education provided Education, education). Fe y Alegria establishes schools for marginalized to poor people foundations, populations in urban areas and in isolated rural settings. FyA international principals hire, train, and supervise teachers. The principal and agencies, the school council are at the center of local decision-making civil society, and the national government deals with strategic issues such communities as growth plans and fundraising. The government pays teacher and principal salaries, while external donors pay for land, construction, and maintenance of schools. Public-Private Partnerships: Program and Policy Descriptions by Contract Type 81 Country Program Program Description Objective Partners Infrastructure Services/Education Services 74 Australia New Schools The private sector finances, designs, and constructs public Outsource and Private operators, Project in New schools following standards established by the Department finance school the State South Wales of Education and also provides cleaning, maintenance, repair, construction and Department of security, safety, utility, and related services for buildings, maintenance of Education and furniture, and equipment until 2032. Private operators receive infrastructure Training performance-related monthly payments. At the end of the contract, the buildings will be transferred to the public sector. 75 Australia Education Works The private sector will deliver new school infrastructure. The Outsource and The government (South New Schools funding arrangement provides for the development of six new finance school of South Australia, Australia) schools in the Playford North, Inner North, and Inner West construction and the Department areas of metropolitan Adelaide. The project value is estimated maintenance of of Education and at $128 million ($A134 million) for the 2006­07 budget infrastructure Children 76 Australia Private finance The private sector will take responsibility for the construction Outsource and The Queensland (Queensland) initiatives and maintenance of seven schools over a 30-year contract, but finance school State Government, the education services will still be provided by the government. construction and private financial The tasks to be contracted out to the private sector include maintenance of partners building repairs, cleaning, janitorial duties, grounds-keeping, infrastructure and security. Handing over responsibility for these services to the private sector will allow teachers to focus on providing education. 77 Belgium Private finance The government selects a single consortium to be responsible Outsource and The Flemish (Flanders) initiatives for the design, construction, financing, and maintenance of finance school Agency for all school building projects. The consortium does not own the construction and Infrastructure in buildings but will receive financial compensation over 30 years maintenance of Education, private in exchange for making the buildings available to the schools. infrastructure financial partners Secondary advantages include a decrease in costs due to economies of scale and the fact that school boards can focus on providing education. 78 Canada PPPs for Education The government contracts with private providers, on a basis Outsource and The government Infrastructure Nova of competitive bidding, the design, construction, finance, and finance school of Nova Scotia, Scotia maintenance of schools for a period of 20 years. Incentives construction and private providers were built into contracts to ensure quality construction and maintenance of maintenance. infrastructure 79 Canada Private finance The Alberta Government is moving forward with the planning Outsource and The government (Alberta) initiatives and construction of 14 new schools that will feature innovative finance school of Alberta, design concepts for middle and senior high schools for the construction and construction firms, Calgary and Edmonton regions. These 14 new schools are part maintenance of school jurisdictions, of a 20-year plan to build new schools in areas of greatest infrastructure communities need. 80 Denmark Trehoje School A Design, Build, Maintain, and Operate contract for a period of Outsource and Private operators 24 years valued at $24.5 million (DKr116 million) finance school and the government construction and of Denmark maintenance of infrastructure 81 Egypt, Arab PPP for new The government provides land while its private sector Outsource and The government Rep. of schools partners design, construct, finance, and furnish public schools finance school of Egypt, private and provide noneducational services under 15­20 year construction and operators agreements. maintenance of infrastructure 82 Germany Offenbach schools The government contracts out the financing, refurbishment, Outsource and SKE, HOCHTIEF, and operation of government schools. Its private sector finance school the government partners operate the schools for 15 years. construction and of the County of maintenance of Offenbach infrastructure 83 Germany City of Cologne The government contracts out the refurbishment and operation Outsource and HOCHTIEF, the schools of government schools. Its private sector partners operate finance school government of the schools for 15 years. construction and City of Cologne maintenance of infrastructure continued 82 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION Country Program Program Description Objective Partners Infrastructure Services/Education Services 84 Greece Macedonia schools Private operators will build 51 new schools using a Design, Outsource and Private operators and Attica schools Build, Finance, Maintain, Operate mechanism valued at $424 finance school and the government million dollars (269 million). Contracts are for periods of 25 construction and of Greece years. The University of the Peloponnese is also being built maintenance of under a PPP scheme. infrastructure 85 India Jointly financed High school buildings are donated, built, and managed by Outsource and Local community in (Gujarat schools local communities. The running expenses, including teacher finance school Gujarat State and State) salaries, are met by the government. construction and government maintenance of infrastructure 86 Ireland Design-build- The private sector consortium (Jarvis Project Limited) was Outsource and The Department operate- finance contracted to design, build, operate, and finance five schools finance school of Education and over a 25-year period. The first school was officially opened in construction and Science, schools, 2002. maintenance of the Jarvis Project infrastructure Ltd 87 Korea, Build-Transfer- The private sector finances and builds social infrastructure Outsource and Private operators Republic of Lease Scheme facilities, transfers ownership of the facilities to the finance school and the government government upon completion of the construction, is granted construction and of Korea the right to operate the facilities for a specified period, and maintenance of leases the facilities back to government in order to recover the infrastructure project costs. Current plans are to use the BTL scheme for 973 schools and 51 universities with a value of $6.7 million dollars (W7 trillion). 88 Netherlands Ypenburg suburb Government contracts with private operator to build and Outsource and The TalentGroep of the Hague operate a new secondary school. The secondary school finance school consortium, the is expected to grow from 150 students at the beginning of construction and government of the the contract to 1,200 by 2009. The contract term is 30 years maintenance of Netherlands (1.5 years for construction and 28.5 years for maintenance, infrastructure including cleaning, furniture, information, ICT, and possibly catering). 89 Norway Persbraten and Private sector operators build, maintain, and operate two Outsource and Private operators Herbraten schools schools for a period of 24 years. The transaction is valued at finance school and the government $100 million dollars (64 million). construction and of Norway maintenance of infrastructure 90 Pakistan Leasing of public The government leases under-used and dilapidated Lease public school Private operators, school buildings to government school buildings to private schools. The private buildings to private the Punjab private operators sector is given the right to operate a school in the afternoon operators government shift, when the school building is closed. In exchange, the private operator must upgrade the building, pay the utility costs of both schools, contribute to the operating costs of both schools, and pay 10% of any profits to the public school council. More than 6,000 such schools are now operating in Punjab. 91 Scotland School estate The major capital investment in schools is made through Outsource and The Scottish strategy public-private partnerships. The Scottish Executive has made finance school Executive, local commitments to build and refurbish schools. construction and authorities, private maintenance of operators infrastructure 92 United Private finance A capital project is designed, built, financed, and managed Outsource and Private operators, Kingdom initiatives/ Building by a private sector consortium under a contract that typically finance school the Department of Schools for the lasts 30 years. The most common structure used is design- construction and Children, Schools, Future build-finance-operate. The private consortium is paid regularly maintenance of and Families from public money based on its performance throughout infrastructure the contract period. If a consortium misses its performance targets, the payment is reduced. At the end of the contract period, the school is returned to the government. a p p e n d i x Methods for Evaluation of B Public-Private Partnership Programs and Policies in Basic and Secondary Education Randomization and regression disconti- population close to the cutoff point but not nuity regressions show the real magnitude on the general population. In other words, of the effects of public-private partnership it is a local estimator. programs (the estimates are unbiased) Instrumental variable and Heck- under general assumptions. In general, man correction models produce correct, randomized studies randomly assign peo- unbiased estimates under more stringent ple to treatment groups. For example, in assumptions. Both methods require a vari- the secondary school voucher program in able with two traits. First, it must explain Colombia, the number of people applying the decision of the school or student to for the vouchers was larger than the num- participate in the program. Second, it can- ber of places available. Since the program's not be correlated with any unobservable budget allocation was not sufficient to cover characteristic that explains the outcome of the demand for vouchers, the recipients of interest, such as test scores. This variable the vouchers were selected using a lottery, makes it possible to model participation creating a treatment group (those selected in a program and, therefore, once self- in the lottery) and a control group (those selection is controlled for, it is possible to not selected in the lottery). The two groups assess a program's impact. The difficulty had, on average, similar observable and with these two methods is finding a valid unobservable characteristics. instrumental variable. Regression discontinuity analysis is typ- The difference in difference method ically applied when a program is allocated compares beneficiaries and nonbeneficia- using a continuous variable. For instance, ries before and after the program. Its key some programs use a means-tested index assumptions are that the trend in the out- to select the target population. In this way, come of interest before the intervention is the program specifies that households that equal for beneficiaries and nonbeneficia- score below a certain cutoff point are eli- ries, and that all nonobservable variables gible for the program and those above the that explain the outcome of interest are cutoff point are not. In this case, the pro- time-invariant. gram's impact can be assessed by dividing Propensity score-matching estimators individuals into a treatment group, con- take a slightly different approach. This taining individuals who score just below method assumes that program participa- the cutoff point, and a control group, tion can be fully explained by a large array containing individuals who score just of observable characteristics measured at above the cutoff point. The two groups are a baseline. Based on this information, the assumed to have very similar characteris- treatment and control groups are con- tics, with the only difference between them structed and their outcome measures com- being their inclusion or exclusion from the pared. The biggest challenge in using both program. Intuitively, for individuals, the difference in difference and propensity cutoff point is almost a random lottery. score-matching is obtaining the large array An important limitation of this method is of baseline data needed to ensure the statis- that it can assess a program's impact on the tical similarity of the two groups. 83 References Aedo, C. 1997. "Organización Industrial de la Schooling in Colombia: Evidence from a Ran- Prestación de Servicios Sociales." Working domized Natural Experiment." American Eco- Paper Series R-302. Inter-American Develop- nomic Review 92(5): 1535­58. ment Bank, Washington, DC. Angrist, J., E. Bettinger, and M. Kremer. 2006. Aedo, C., and O. Larrañaga. 1994. "Educación "Long-Term Educational Consequences of Privada vs. Pública en Chile: Calidad y Sesgo Secondary School Vouchers: Evidence from de Selección." Graduate Economics Program, Administrative Records in Colombia." Ameri- Santiago, and Instituto Latinoamericano can Economic Review 96(3): 847­62. de Doctrinas y Estudios Sociales Georgetown Aslam, M. 2007. "The Relative Effectiveness University, Washington, DC. of Government and Private Schools in Ahlin, A. 2003. "Does School Competition Pakistan: Are Girls Worse Off?" RECOUP Matter? Effects of a Large-Scale School Working Paper 4. University of Cambridge, Choice Reform on Student Performance." Cambridge. Department of Economics, Uppsala Univer- Audit Office of New South Wales. 2006. "The sity, Sweden. New Schools Privately Financed Project." Allcott, H., and D. E. Ortega. 2007. "The Per- Sydney. formance of Decentralized School Systems: Baksh, Shaista. Forthcoming. "Public-Private Evidence from Fe y Alegría in Venezuela" Pre- Partnerships in World Bank Education Lend- sented at the Conference on Public-Private ing, 1998­2007: Background Report for the Partnerships in Education, World Bank, Role and Impact of Public-Private Partnerships Washington DC, June 7­8. in Education." World Bank, Washington, DC. Andersen, S. C. 2005. "Selection and Competi- Barrera-Osorio, F. 2007. "The Impact of Private tion Effects in a Large-Scale School Voucher Provision of Public Education: Empirical System." For presentation at the conference Evidence from Bogota's Concession Schools." European Public Choice Society 1005, Uni- World Bank Policy Research Working Paper versity of Durham, United Kingdom, March No. 4121, Washington, DC. 31 to 3 April. Barrera-Osorio, F., and H. A. Patrinos. 2009. ------. "Private Schools and the Parents That "An International Perspective on School Choose Them: Empirical Evidence from the Vouchers" in Handbook of Research on School Danish School Voucher System." Scandina- Choice, Peabody College at Vanderbilt Univer- vian Political Studies 31(1): 69. sity, Nashville, Tennessee. Andrabi, T., J. Das, and A. I. Khwaja. 2006. Bedi, A. S., and A. Garg. 2000. "The Effectiveness "A Dime a Day: The Possibilities and Lim- of Private Versus Public Schools: The Case of its of Private Schooling in Pakistan." Policy Indonesia." Journal of Development Economics Research Working Paper 4066. World Bank, 69(2): 463­494. Washington, DC. Belfield, C. R., and A. L. Wooten. 2003. "Educa- Andrabi, T., J. Das, A. I. Khwaja, T. Vishwanath, tion Privatization: The Attitudes and Experi- and T. Zajonc. 2007. "Learning and Edu- ences of Superintendents." Occasional Paper cational Achievements in Punjab Schools 70. National Center for the Study of Priva- (Leaps): Insights to Inform the Education tization in Education, Columbia University, Policy Debate." World Bank, Washington, DC. New York. Angrist, J., E. Bettinger, E. Bloom, E. King, Bell, S. 1995 "Sharing the Wealth: Privatization and M. Kremer. 2002. "Vouchers for Private through Broad-Based Ownership Strategies." 85 86 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION Discussion Paper 285. World Bank, Washing- Up: Examining the Evidence on Enrollment ton, DC. and Achievement. Economic Policy Institute, Bellei, C. 2005. "The Private-Public School Con- Washington, DC. troversy: The Case of Chile." Presented at the Carnoy, M., and P. McEwan. 2000. "The Effec- Conference on Mobilizing the Private Sector tiveness and Efficiency of Private Schools in for Public Education, Harvard University, Chile's Voucher System." Educational Evalua- Cambridge, MA, October 5­6. tion and Policy Analysis 22(3): 213­39. Benveniste, L. A., and P. J. McEwan. 2000. "Con- Carnoy, M., and P. J. McEwan. 2001. "Privatiza- straints to Implementing Educational Inno- tion through Vouchers in Developing Coun- vations: The Case of Multigrade Schools." tries: The Cases of Chile and Colombia." In International Review of Education 46(1­2): Privatizing Education: Can the Marketplace 31­48. Deliver Choice, Efficiency, Equity and Social Bettinger, E. 2005. "The Effect of Charter Cohesion?, ed. H. M. Levin, 151­177, West- Schools on Charter Students and Public view Press, United States. Schools." Economics of Education Review 24 Center for Education Reform. 2007. "National (2): 133­147. Charter School Data 2007­2008 New School Bettinger, E., M. Kremer, and J. E. Saavedra. Estimates." (http://www.edreform.com/ 2008. "Are Vouchers Redistributive?" Mimeo _upload/CER_charter_numbers.pdf). Harvard, March 2008. Chakrabarti, R., and P. E. Peterson. 2008. School Bifulco, R., and H. F. Ladd. 2006. "The Impact Choice International: Exploring Public-Private of Charter Schools on Student Achievement: Partnerships. Cambridge, MA: MIT Press. Evidence from North Carolina." Journal of CIPPEC (Centro de Implementacion de Politi- Education Finance and Policy 1(1): 778­820. cas Publicas para la Equidad y el Crecimiento. Blomqvist, A., and E. Jimenez. 1989. "The Pub- 2007. Programa de Monitoreo de la Ley de lic Role in Private Post-Secondary Educa- Financiamiento Educativo. Presentación No. tion: A Review of Issues and Options." Policy 6. Aportes Financieros Estatales a la Edu- Research Working Paper 240. World Bank, cación Privada. Argentina. Washington, DC. Contreras, D., S. Bustos, and P Sepulveda. Booker, K., S. M. Gilpatric, T. Gronberg, and D. 2008. "When the Schools Are the Ones That Jansen. 2008. "The Effect of Charter Schools Choose: Policy Analysis of the Screening in on Traditional Public School Students in Chile." Department of Economy, University Texas: Are Children who Stay Behind Left of Chile, Santiago. Behind?" Journal of Urban Economics 64(1): Cox, D., and E. Jimenez. 1991. "The Relative 123­145. Effectiveness of Private and Public Schools." Borja, R. R. 2003. "China Seeking US Investment Journal of Development Economics 34(1­2): in Private School Sector." Education Week 24 99­121. September. Eggers, W. D. 1998. "Competitive Neutrality: Bravo, D., D. Contreras, and C. Sanhueza. 1999. Ensuring a Level Playing Field in Managed "Rendimiento Educacional, Desigualdad, y Competitions." How-to Guide No. 18, Reason Brecha de Desempeño Privado Público: Chile Public Policy Institute, Los Angeles, CA. 1982­1997." Department of Economy, Uni- Elacqua G., D. Contreras, and F. Salazar. 2008. versity of Chile, Santiago. "The Effectiveness of Private School Fran- Brewer, D., C. H. Augustine, G. L. Zellman, G. chises In Chile's National Voucher Program." Ryan, C. A. Goldman, C. Stasz and L. Con- Paper presented at the World Bank Confer- stant. 2007. "Education for a New Era. Design ence, "PrivatePublic Partnership," Washing- and Implementation of K­12 Education ton, DC, June 7­8, 2007. Reform in Qatar." RAND Corporation, Santa Epple, D., and R. E. Romano. 1998. "Competi- Monica, CA. tion between Private and Public Schools, Burch, P., M. Steinberg, and J. Donovan. 2007. Vouchers, and Peer-Group Effects" American "Supplemental Educational Services and Economic Review, 88(1): 33­62. NCLB: Policy Assumptions, Market Practices, European Commission. 2007. "School Auton- Emerging Issues." Educational Evaluation and omy in Europe Policies and Measures." Edu- Policy Analysis 29(2): 115­33. cation and Culture DG. Brussels, Eurydice. Carnoy, M., R. Jacobsen, L. Michel, and R. Fielden, J., and N. LaRocque. 2008. "The Evolv- Rothstein. 2005. The Charter School Dust ing Regulatory Context for Private Education References 87 in Emerging Economies". The World Bank Harding, A. L. 2002. "Introduction to the Private Group International Colloquium on Private Participation in Health Services Handbook." Education. Washington DC. In Private Participation in Health Services Filer, R. K., and D. Münich. 2000. "Responses of Handbook, eds. April L. Harding and Alexan- Private and Public Schools to Voucher Fund- der S. Preker, 46 p. Washington, DC: World ing: The Czech and Hungarian Experience." Bank. William Davidson Institute Working Paper Hatcher, R. 2003. "Privatization and the US No. 360, University of Michigan, Ann Arbor. School System: Voucher Programmes and Fiske, E. B., and H. F. Ladd. 2000. When Schools Education Management Organisations." Compete: A Cautionary Tale. Washington, DC: Paper presented at the ESRC Seminar, "Pri- Brookings Institution Press. vate Sector Participation in Public Sector Fitz, J., and B. Beers. 2002. "Education Manage- Education," Institute of Education, London, ment Organizations and the Privatization of 28 November, 2003. Public Education: A Cross-National Com- ------. 2006. "Privatization and Sponsorship: parison of the USA and Britain." Comparative The Re-Agenting of the School System in Education 38(2): 137­54. England." Journal of Education Policy 21(5): Foundation Center. 2007. Foundation Growth 599­619. and Giving Estimates: Current Outlook. New Hentschke, G. C. 2005. "Characteristics of York: Foundation Center. Growth in the Education Industry: Illustra- Friedman, M. 1955. "Role of Government in tions from U.S. Education Businesses." Dis- Education." In Economics and the Public Inter- cussion Draft, University of Bremen, Bremen. est, ed. Robert Solo, 123­44. New Brunswick, Himmler, O. 2007. "The Effects of School NJ: Rutgers University Press. Choice on Academic Achievement in the Gallegos, F. 2004. "School Choice, Incentives, Netherlands." Georg-August-Universität Göt- and Academic Outcomes: Evidence from tingen, Sweden. Chile." Boston: MIT Press. HM Treasury. 2008. "Infrastructure Procure- Gauri, V., and A. Vawda. 2004. "Vouchers for ment: Delivering Long-Term Value." London. Basic Education in Developing Economies: House of Commons Education and Skills Com- A Principal Agent Perspective." World Bank mittee. 2006­07. "Sustainable Schools: Are We Research Observer 19(2): 259­280. Building Schools for the Future?" London. Gibson, H., and B. Davies. 2008. "The Impact Hoxby, C. M. 2000. "Does Competition among of Public Private Partnerships on Education." Public Schools Benefit Students and Tax- International Journal of Education Manage- payers?" American Economic Review 90(5): ment 22(1): 74­89. 1209­38. Gill, B. P., M. Timpane, K. E. Ross, D. J. Brewer, ------, ed. 2003. The Economics of School and K. Booker. 2007. Rhetoric versus Reality: Choice. Chicago: University of Chicago Press. What We Know and What We Need to Know Hoxby, C. M., and S. Murarka. 2007. "Charter About Vouchers and Charter Schools. Santa Schools in New York City: Who Enrolls and Monica, CA: RAND Education. How They Affect Their Students' Achieve- Government Accountability Office. 2006. "Edu- ments." National Bureau of Economic cation Actions Needed to Improve Local Research, Cambridge, MA. Implementation and State Evaluation of Sup- Hoxby, C., and J. E. Rockoff. 2004. "The Impact plemental Educational Services." Report to of Charter Schools on Student Achievement." Congressional Requesters, Washington, DC. Taubman Center for State and Local Govern- Hanushek, E. A. 2003. "The Failure of Input- ment, Kennedy School of Government, Cam- Based Schooling Policies." Economic Journal bridge, MA. 113(485): F64­F98. Hsieh, C. T., and M. Urquiola. 2006. "The Effects Hanushek, E. A., J. F. Kain, S. G. Rivkin, and of Generalized School Choice on Achieve- G. F. Branch. 2007. "Charter School Quality ment and Stratification: Evidence from and Parental Decision Making with School Chile's Voucher Program." Journal of Public Choice" Journal of Public Economics 91(5­6): Economics 90(8­9): 1477­1503. 823­848. Hurst, C. 2004. "An Economic Analysis of Ire- Hanushek, E. A., and L. Woessmann. 2007. Edu- land's First Public Private Partnership." Inter- cation Quality and Economic Growth. World national Journal of Public Sector Management Bank, Washington, DC. 17(5): 379­388. 88 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION IFC (International Finance Corporation). 2007. tion: Evidence from the United States." Swed- "IFC Supports Private Education in Africa." ish Economic Policy Review 10: 67­76. Media Release, 17 July. Washington, DC. LaRocque, N. 1999. "The Private Education ------. 2006. "Proposed Investment in School in Sector in West Africa: Overview Report." Financing & Technical Assistance Facility. World Bank, Washington, DC. Kenya". IFC/R2006-0278. World Bank. Wash- ------. 2002. Private Education in the Phil- ington DC. ippines: A Market and Regulatory Survey. IMF (International Monetary Fund). 2004. Manila: Asian Development Bank. "Public-Private Partnerships." Fiscal Affairs ------. 2008. "Public-Private Partnerships in Department (In consultation with other Basic Education: An International Review" departments the World Bank and the Inter- CfBT Education Trust, Berkshire. American Development Bank). International LaRocque, N., and V. Jacobsen. 2000. "Private Monetary Fund, Washington, DC. Education in Cameroon." World Bank, Wash- International Financial Services London. 2001. ington, DC. "Public Private Partnerships: UK Expertise LaRocque, N. and H. Patrinos. 2006. "Choice for International Markets." London. and Contracting Mechanisms in the Educa- ------. 2008. "PFI in the UK and PPP in tion Sector." World Bank, Washington, DC. Europe 2008." IFSL Research. London. Latham, M. 2005. "The PFI Model and its Effect Jallade, J. P. 1973. "The Financing of Education: on Educational Outcomes." Prepared for the An Examination of Basic Issues." Working conference Mobilizing the Private Sector for Paper 157, World Bank, Washington, DC. Public Education, 5­6 October, World Bank James, E. 1984. "Benefits and Costs of Priva- and Kennedy School of Government, Har- tized Public Services: Lessons from the Dutch vard University. Educational System." Comparative Education Levin, H. M. 2000. "The Public-Private Nexus in Review 28(4): 605­664. Education." In Public-Private Policy Partner- ------ 1993. "Why Do Different Countries ships, ed. P. Vaillancourt Rosenau, 129­142. Choose a Different Public-Private Mix of Cambridge, MA: MIT Press. Educational Services?" Journal of Human Lewin, K. M., and Y. Sayed. 2005. "Non- Resources 28(3): 571­592. Government Secondary Schooling in Sub- Jones, D., B. Vann, and O. Hayford. 2004. "Pub- Saharan Africa. Exploring the Evidence in lic Private Partnerships--Managing the Chal- South Africa and Malawi." U.K. Department lenges." Clayton Utz, Sydney, Australia. for International Development, London. Justesen, M. K. 2002. Learning from Europe: The Lockheed, M., and W. van Eeghen. 1998. "Effec- Dutch and Danish School Systems. London: tive Financing of Education: Role of Public and Adam Smith Institute. Private Sectors." Paper presented at the 1998 Kang, C. 2007. "Classroom Peer Effects and Annual Seminar jointly organized by the Arab Academic Achievement: Quasi-Randomized Monetary Fund and the Arab Fund for Eco- Evidence from South Korea." Journal of Urban nomic and Social Development in collabora- Economics 61(3): 458­495. tion with the International Monetary Fund and Kim, J., H. Alderman, and P. F. Orazem. 1999. the World Bank, Abu Dhabi, February 21­22. "Can Private School Subsidies Increase Loevinsohn, B., and A. Harding. 2005. "Buying Enrollment for the Poor? The Quetta Urban Results? Contracting for Health Service Deliv- Fellowship Program." World Bank Economic ery in Developing Countries." The Lancet Review 13(3): 443­465. 366(9486): 676­681. Kingdon, G. 2007. "The Progress of School Edu- McEwan, P. 2001 "The Effectiveness of Public, cation in India." Oxford Review of Economic Catholic, and Non-Religious Private Schools Policy 23(2): 168­195. in Chile's Voucher System." Education Eco- nomics 9(2): 103­128. KPMG. 2008. "Investment in School Facilities and PFI: Do They Play a Role in Educational McEwan, P. J., M. Urquiola, and E. Vegas. 2007. Outcomes?" London. "School Choice, Stratification, and Student Performance: Lessons from Chile." Economia. Ladd, H. F. 2002. "School Vouchers: A Critical Journal of the Latin American and Caribbean View." Journal of Economic Perspectives Economic Association. 8(2): 1­27. 16(4): 3­24. Micklewright, J., and A. Wright. 2003. "Private ------. 2003. "Comment on Caroline M. Donations for International Development." Hoxby: School Choice and School Competi- References 89 Discussion Paper 2003/82. United Nations Orazem, P. F. 2000. "The Urban and Rural Fel- University, World Institute for Development lowship School Experiments in Pakistan. Economics Research, Helsinki. Design, Evaluation and Sustainability." The Minnesota House of Representatives. 2003. World Bank, Washington DC. "Charter Schools. Information Brief " Palter, R. N., J. Walder, and S. Westlake. 2008. Research Department. St. Paul. (http://www "How Investors Can Get More out of Infra- .house.leg.state.mn.us/hrd/pubs/chrtschl.pdf). structure." The McKinsey Quarterly (February Mizala, A., and P. Romaguera. 2000. "School 2008). Performance and Choice: The Chilean Expe- Patrinos, H. A. 2000. "Market Forces in Educa- rience." Journal of Human Resources 35(2): tion." European Journal of Education 35(1): 392­417. 61­80. Mohadeb, P., and D. Kulpoo. 2008. "The Provi- ------. 2002. "Private Education Provision sion and Financing of Quality Secondary and Public Finance: The Netherlands as a Education. Through PPP in Mauritius. Possible Model." Occasional Paper No. 59, A Success Story." Working Document. Draft National Center for the Study of Privatization for Beyond Primary Education: Challenges and in Education, Teachers College, Columbia Approaches to Expanding Learning Opportu- University. nities in Africa. Association for the Develop- ------. 2005. "Education Contracting: Scope of ment of Education in Africa. Future Research." Program on Education Pol- Molnar, A., D. R. Garcia, M. Bartlett, and icy and Governance Report 05-23. Harvard A. O'Neill. 2006. "Profiles of For-Profit Edu- University, Cambridge, MA. cation Management Organizations." Arizona Patrinos, H. A., and C. Sakellariou. 2008. "Qual- State University, Commercialism in Educa- ity of Schooling, Returns to Schooling and tion Research Unit, Tempe. the 1981 Vouchers Reform in Chile." World Muralidharan, K., and M. Kremer. 2006. "Public Bank Policy Research Working Paper No. and Private Schools in Rural India." Harvard 5617, Washington, DC. University, Department of Economics, Cam- Patrinos, H. A., and S. Sosale. 2007. Mobilizing bridge, MA. the Private Sector for Public Education: A View Nechyba, T. J. 2000. "Mobility, Targeting and from the Trenches. Directions in Develop- Private School Vouchers." American Economic ment: Human Development. Washington, Review 90(1): 130­146. DC: World Bank. The Netherlands Ministry of Education, Culture PricewaterhouseCoopers. 2005. "Delivering and Science. 2002. Education, Culture and the PPP Promise: A Review of PPP Issues and Science in the Netherlands, Facts and Figures Activity." Public Sector Research Centre. 2002. http://www.pwc.com/extweb/pwcpublications OECD (Organisation for Economic Co-operation .nsf/docid/5D37E0E325CF5D71852570DC000 and Development). 2003."Checklist for Foreign 9C39B/$File/promisereport.pdf. Direct Investment Incentive Policies." Paris. ------. 2008. "The Value of PFI: Hanging in http://www.oecd.org/dataoecd/45/21/2506900 the Balance Sheet?" Public Sector Research .pdf. Centre. http://www.ukmediacentre.pwc.com/ ------. 2004a. "Australia's First Public Private content/Detail.asp?ReleaseID=2617&NewsA Partnership School Project." The Journal of the reaID=2. OECD Programme on Educational Building. Raynor, J., and K. Wesson. 2006. "The Girls' 2004/2. No. 52: 23. Stipend Program in Bangladesh." Journal of ------. 2004b. "The United Kingdom's Part- Education for International Development Privately Funded Business Academy Bexley." 2:2. http://www.equip123.net/JEID/articles/3/ The Journal of the OECD Programme on Educa- Girls'StipendPrograminBangladesh.pdf. tional Building 52(2): 25. Rinne, R., J. Kivirauma, and H. Simola. 2002. ------. 2006. "Public-Private Partnerships in "Shoots of Revisionist Education Policy or Flanders." PEB Exchange 6 2006/6:1. Just Slow Readjustment? The Finnish Case of Educational Reconstruction." Journal of Edu- ------. 2007a. "PISA 2006: Science Competen- cation Policy 17(6): 643­658. cies for Tomorrow's World." Volume 1: Analy- sis, Paris. Rodriguez, A., and K. Hovde. 2002. "The Chal- lenge of School Autonomy: Supporting ------. 2007b. "Education at a Glance 2007." Principals." World Bank, Latin America and the Paris. Caribbean Regional Office, Washington, DC. 90 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION Rodriguez, J. 1988. "School Achievement and 27. International Monetary Fund, Washing- Decentralization Policy: The Chilean Case." ton, DC. Revista de Análisis Económico 3(1): 75­88. Taylor, R. J. 2003. "Contracting for Health Ser- Sadka, E. 2006. "Public-Private Partnerships: vices." In Private Participation in Health Ser- A Public Economics Perspective." IMF Work- vices Handbook, eds. A. Harding and A. Preker, ing Paper WP/06/77. Washington, DC. 195­204. Washington, DC: World Bank. Sakellariou, C., and H. A. Patrinos. 2008. "Inci- United Kingdom Audit Commission. 2003. "PFI dence Analysis of Public Support to the Pri- in Schools: The Quality and Cost of Build- vate Sector in Côte d'Ivoire." International ings and Services Provided by Early Private Journal of Educational Development. Finance Initiative Schemes." London. Sandström, F. M., and F. Bergström. 2004. United Nations Conference on Trade and Devel- "School Vouchers In Practice: Competition opment (UNCTAD). Various Issues. World Will Not Hurt You." Journal of Public Econom- Investment Report. Geneva. ics 89(2­3): 351­380. United Nations Educational, Scientific, and Sapelli, C., and B. Vial. 2004. "Peer Effects and Cultural Organization (UNESCO). 2007. Relative Performance of Voucher Schools in "Non-State Providers and Public-Private- Chile." Paper presented at the American Eco- Community Partnerships in Education." Back- nomic Association meetings, San Diego, CA, ground paper for the Education for All Global January 3. Monitoring Report 2008: Education for All by Sarwar, M. B. 2006. "Documenting Educational 2015: Will We Make It? A. K. F. Team. Paris. Innovations. Sharing Practices for Educa- United States Department of Education. 2007. tional Change." Sindh Education Foundation, "Improving Basic Programs Operated by Karachi. Local Education Agencies (Title I, Part a)." Sass, T. R. 2005. "Charter Schools and Student Office of Elementary and Secondary Educa- Achievement in Florida." Working Paper. tion, Washington, DC. Florida State University, Tallahassee. Uribe, C., R. Murnane, J. B. Willett, and M. Somers. Savas, E. S. 2000. Privatization and Public-Pri- 2006."Expanding School Enrollment by Sub- vate Partnerships, New York: Chatham House sidizing Private Schools: Lessons from Bogotá." Publishers. Comparative Education Review 50(2): 241­277. School Ventures. 2008. "The African Private Vegas, E. 2002. "School Choice, Student Per- Schools Investment Index (APSI) 2008." formance, and Teacher and Student Charac- Washington, DC. http://www.schoolventures teristics: The Chilean Case." Policy Research .com. Working Paper 2833, World Bank, Washing- Schütz, G., M. R. West and L. Woessmann. 2007. ton, DC. "School Accountability, Autonomy, Choice Verspoor, A. M. 2008. "The Power of Public- and the Equity of Student Achievement: Private Partnership: Coming Together for International Evidence from PISA." Educa- Secondary Education in Africa." Working tion Working Paper No. 14. Directorate for Document Draft, Association for the Devel- Education, OECD, Paris. opment of Education in Africa, Paris. Solmon, L. C. 2004. "Comparison of Traditional VietNamNet Bridge. 2006. "City's Educa- Public Schools and Charter Schools on Reten- tion Sector Gets Four More FDI Projects." tion, School Switching, and Achievement August 24. http://english.vietnamnet.vn/ Growth." Policy Report 192. Goldwater Insti- biz/2006/08/605411/. tute, Phoenix, AZ. Villa, L. and J. Duarte. 2005. "Concessionary Sosale, S. 2000. "Trends in Private Sector Devel- Public Schools in Bogota." In Private Educa- opment in World Bank Education Projects." tion and Public Policy in Latin America, eds. Policy Research Working Paper 2452, World L. Wolff, J. C. Navarro, and P. Gonzales, Bank, Washington, DC. 95­130. Washington, DC: PREAL. Srivastava, P. 2007. Neither Voice nor Loyalty: Wang, Y. 2000. Public-Private Partnerships in the School Choice and the Low-Fee Private Sector Social Sector. Issues and Country Experiences in India. Occasional Paper 134, Research Pub- in Asia and the Pacific. Tokyo: Asian Develop- lications Series, National Center for the Study ment Bank Institute. of Privatization in Education, New York. World Bank. 2003a. World Development Report Tanzi, Vito, and Howell Zee. 2001. "Tax Policy 2004: Making Services Work for Poor People. for Developing Countries." Economic Issues Washington, DC. References 91 ------. 2003b. Partnerships within the Public Sector, Latin America and Caribbean Region. Sector to Achieve Health Objectives. Washing- Report 38600, Washington, DC. ton, DC. ------. 2008. "Performance-Based Contracting ------. 2003c. "Project Performance Assess- for Health Services in Developing Countries. ment Report" Bangladesh, Female Secondary A Toolkit." World Bank Institute, Health, School Assistance Project. Sector and The- Nutrition and Population Series. Washington, matic Evaluation Group, Operations Evalua- DC. tion Department. Report 26226. Washington, World Bank Education Statistics Version DC. 5.3. 2008. Washington, DC. http://www ------. 2004. Public Communication Programs .worldbank.org/education/edstats. for Privatization Projects--A Toolkit for World Woessmann, L. 2005. "Public-Private Partner- Bank Task Team Leaders and Clients. Wash- ships in Schooling: Cross-Country Evidence ington, DC: Development Communications on their Effectiveness in Providing Cognitive Unit, External Affairs Vice Presidency. Skills." Prepared for the conference Mobiliz- ------. 2006. "Colombia Contracting Educa- ing the Private Sector for Public Education, tion Services." Report 31841-CO. Washing- World Bank and Kennedy School of Govern- ton, DC. ment, Harvard University, Cambridge, MA, ------. 2007a. What Is School-Based Manage- October 5­6. ment? Human Development Network, Wash- Yilmaz, Yesim. 1998. Private Regulation: A Real ington, DC. Alternative for Regulatory Reform. Cato Report ------. 2007b. "Haiti Education For All Adapt- No. 303, Cato Institute, Washington, DC. able Program Grant" Human Development Index Boxes, figures, and tables are indicated by b, f, and t, respectively. A CfBT Education Trust, United Kingdom, 52 academic achievement, 30, 40, 63 charter schools, 6, 17­18, 24, 80 Academic Colleges Group, 46 autonomy, 63 accountability, 11, 28, 44b, 59 Canada program, 78 in contracts, 17 evidence for, 38­39, 64­65 regulatory framework for, 18 national association for, 55 accounting standards, 27 vs. vouchers, 25b accreditation, 52b, 53 Chile, 19, 72 adopt-a-school program, 22­23, 79 vouchers, 6, 7, 30, 65, 72 Argentina, 17, 71, 77 studies, 36, 37t, 38 assessment services, 52. See also evaluation Cluster-Based Teacher Training Program, Assisted Places Scheme, United Kingdom, 76 Pakistan, 22, 77 Australia, PPP program descriptions, 71, 76, Cologne, City of, schools, 81 81 Colombia, 78. See also concession schools infrastructure, 26, 27 Escuela Nueva Foundation, 22, 77 autonomy, 25, 28, 63 Plan de Ampliación de Cobertura de la AVEC. See Venezuelan Association of Catholic Educación Secundaria, 20, 36, 72 Schools private management studies, 39t voucher program, 6, 7, 20, 65 B voucher study, 37t communications strategy development, 57, 68 Balochistan Education Support Project, 49 competition, 4, 8, 18, 34, 35 Bangladesh, 17, 20, 71 by PPP type, 35t Bangladesh Rural Advancement Committee comprehensive contracting, 13­14 (BRAC) Non-Formal Primary concession schools, Colombia, 6, 24­25 Education Program, 6 evidence for, 39t Bank Netherlands Partnership Program model, 17, 63 (BNPP), viii program description, 78 Basic Education Support Project, Pakistan, 74 providers, 28, 34 Beaconhouse Group, 46, 49 construction contracts, 32 Belgium, 26, 71, 81 contract schools, 72, 80 Benin, enrollment rates, 42 contracting agency, capacity building, 54­56, 68 bidding process, 4, 34, 67­68 contracting out, 1, 4, 5, 9, 9t. See also bidding selection of private partner, 53, 54 process Bolivia, Fe y Alegría Network, 77 benefits of expertise and innovation, 22 Brazil, 76, 78 benefits to schools, 63 build-operate-transfer agreements, 12 education services, 11 Burkina Faso, 42 facilities, 12­13 improves quality and access, 59, 61 C support services, 10­11 contracts, 14, 17, 32, 64 Cambridge International Exams, 53 cancellation guidelines, 25 Cameroon, registering private schools, 48b characteristics of, 50, 51 Canada, 26, 71, 78, 81 education services, 15, 20­21 capacity-building, 29, 69 flexibility of, 4, 33, 67 contracting agency, 54­56, 68 main forms of, 59­61 capital market access, 44b operations, 25, 28, 31­32 capital sources, 13f prompt payment, 55 Catholic Church, 46 risk-sharing, 28, 34 93 94 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION contracts (continued) empirical evidence, 35­36, 37t, 38­40 types, 2, 9t enrollment, 8, 32, 42 corruption, 44b, 55, 57 private sector, 2t, 15 cost savings, 13 underserved populations, 1, 3, 20 Côte d'Ivoire, 21, 72 vouchers, 19, 32­33 Czech Republic, PPP program description, 72 equity, PPP design, 33 Escuela Nueva Foundation, 22, 77 D evaluation. See also assessment evaluation methods for PPPs, 83 De La Salle Supervised Schools Program, the evaluation of education outcomes, 57, 68 Philippines, 52 decentralization of education, 8, 63 F Delhi Public School Society, 46 Denmark, 19, 72, 81 facilities contracting, 9t, 12­14, 60­61 design-build-finance-operate, 12, 82 faith-based organizations, 3 difference in difference, 39t faith-based schools, 8 Dominican Republic, Fe y Alegría Network, 78 Fe y Alegría Network, 6, 24, 39t, 63 PPP program description, 79, 80 E program descriptions, 77, 78 Female Secondary School Assistance Project, Ecuador, Fe y Alegría Network, 78 71 education, 3, 23, 28, 47, 55. See also schools Financial Assistance per Child Enrolled Basics agency specialization, 29 Program (FAS), Pakistan, 74 assessment, 5­6 financial reporting standards, 27 capital sources, 13b financing, 15, 18, 69 contracts, 9t Finland, 18 decentralization in United States, 63 flexibility, 35t, 50 demand, 62 foreign direct investment (FDI), 46­47, 66 efficiency, 30 foreign investment, 44b equity, 33 Foundation Assisted Schools Program, improvement factors, 35 Pakistan, 21, 21b, 29, 49 objectives, 32t fraud, 55, 57 outcomes, 33­34, 41 funding formulas to target low-income by PPP type, 35t families, 20, 62 evaluating, 57, 68 funding per pupil, 74 policy, 65­67 weaknesses, 43, 45 G private more effective than public, 29 quality and access improved through Gambia, The, 20, 72 contracting of services, 59, 61 Germany, 26, 81 quality effects of foreign organizations, 53 girls', enrollment of, 20, 62 rise of private sector involvement, 16 Global Education Management Systems, 46, 53 supplemental services, 23 government, 5, 43b, 47. See also contracting out underserved groups, 1, 3, 20 acknowledging private sector role in Education Action Zones, United Kingdom, 80 education, 42 Education for All, Africa, 42 contracting out, 9t Education for All, Haiti program grant, 73 education agency specialization, 21b, 29 Education Foundation, 49 education role, 1, 23, 55 education services' contracts, 9, 9t, 11, 15, 29, encouraging private involvement, 45­46, 60, 62 65 comprehensive contracting, 13­14 funding systems, 49­50 incentives, 56­57 levels of PPP engagement, 17­18 PPP program descriptions, 71­76, 80­82 PPP design and implementation, 44b quality output specifications, 20­21 resistance to private involvement, 43 Educational Service Contracting (ESC), 75 split of provider and purchaser functions, efficiency, 13 54 Egypt, Arab Republic of, PPP program usually favors public school over private, 45 description, 26, 81 Greece, PPP program description, 82 El Salvador, Fe y Alegría Network, 78 Guatemala, 73, 78 Index 95 H N Haiti, 18, 19b, 73 National Association of Charter School Heckman correction model, 37t, 38 Authorizers, United States, 55 Herbraten schools, Norway, 82 Netherlands, the, 8, 82 Honduras, Fe y Alegría Network, 78 voucher program, 19, 37t, 74 Hungary, voucher program description, 75 New Zealand, 51b, 74 Nicaragua, PPP program description, 79 I Norway, Persbraten and Herbraten schools, 82 Idara-e-Taleem-o-Agahi, Pakistan, 22 O impact evaluations, 57 incentives, 29, 56­57 Offenbach Schools, Germany, 81 and sanctions, 21b, 68 operating requirements, 29, 69 Independent School Program, Qatar, 24, 75, operational services' contracts, 9t, 17, 28, 79 31­32, 60 independent schools, funding programs, 71 cost savings, 13 India, 46, 73, 77, 80, 82 for entire school, 11 inequity, reducing through voucher programs, programs descriptions, 77­80 20 superior to management contracts, 50 information flows, 51­53, 67 Organisation for Economic Co-operation and infrastructure, 12, 12t, 17­18, 40 Development (OECD) countries, 3 construction quality, 27 outsourcing. See contracting funding for, 3 investment, 63­64 P private sector involvement, 25­28 infrastructure services, PPP program Pakistan, 6, 20, 46, 82 descriptions, 80­82 Cluster-Based Teacher Training Program, 22 innovation, 25, 29, 69 PPP program descriptions, 74, 75, 77, 79 instrumental variable, 37t Punjab Foundation Assisted Schools Integrated Schools, New Zealand PPP Program, 21, 21b, 29, 49 program, 74 Pakistan Education Foundation, 29 interest rates, 27 Pakistan Railways Schools, 79 International Baccalaureate, 53 Panama, PPP program description, 79 International Finance Corporation (IFC), 58, Paraguay, PPP program description, 79 69 payments, promptness of, 55 International Maritime Organization, 53 peer effects, 36 Ireland, 27, 82 per pupil funding, 74 Italy, voucher program description, 73 performance, 29, 56, 69. See also school performance K incentives and sanctions, 21b, 68 Persbraten schools, Norway, 82 Kenya, 42, 45b Peru, Fe y Alegría Network, 79 Korea, Republic of, 37t, 73, 82 Philippines, the, 52b, 75 Pilot Education Voucher Scheme, Pakistan, 75 L Pitagoras Network, Brazil, 23 policy and regulations, 45­58 Lagos, private enrollment, 42 policy environment, 44b loans, 58 PPPs. See public-private partnerships primary school, 2t, 3, 15 M private education providers, 25, 50 international, 46­47, 53 management services contracts, 10, 59­60 private education sector, 2­3, 45, 49 programs descriptions, 77­80 financing, 3f, 69 management support, 9t, 23 targeting low-income students, 16 market, creating for education, 23 private finance initiatives for infrastructure, 1, Mauritius, 20, 73 25­26, 27, 32t, 35t, 63­64 Mexico, 17 construction quality, 27 microcredit facilities, 58b empirical evidence for, 40 monitoring, 52, 57 impact on education outcomes, 33 96 THE ROLE AND IMPACT OF PUBLIC-PRIVATE PARTNERSHIPS IN EDUCATION private finance initiatives for partner selection, 53, 67­68 infrastructure (continued) programs by contract type, 71­82 key characteristic of, 28 promoting competition, 34 PPP program description, 81 pros and cons, 4­5 private management of schools, 32t, 35, 35t recommendations for, 69 empirical evidence for, 38­39, 39t research needed on education outcomes, 41 private partner selection, 34, 67­68 risk-sharing, 34 private schools, 29, 71 schools receive decision-making power, 28 enrollment, 15 self-selection problem, 31, 35, 36 establishing and registering, 45b, 47­48, separating purchaser and provider roles, 48b, 66 54 experiences of PPP funding, 61­63 teams, 55 government funding of, 18­21 types, 2­3, 35t more effective and higher quality, 29 Puerto Rico, PPP program description, 76 PPP financing options, 18 Punjab Education Foundation, 21b setting tuition and fees, 46, 65­66 subsidizing, 30, 49­50 Q targeting low-income students, 20 vs. public schools and use of subsidies, 39­40 Qatar, 28, 75 private sector, 3, 4, 32. See also bidding process Independent School Program, 24, 63, 80 infrastructure building, 25­28 quality, 21b, 35, 35t international involvement, 57­58 and low price, 5 operation of public schools, 24­25 incentives, 29, 69 private tutoring increase, 3 recommendations for, 29 professional services, 9t, 10 specifications, 20­21 and support services, 22­23 through private management of public contracts, 60, 61 schools, 62 Programme for International Student Quality Advancement and Institutional Assessment (PISA), 5, 30 Development, 77 Programa Ampliación de Cobertura de la Quality Assurance Certification Program, Educación Secundaria, 20, 36, 72 Pakistan, 22 propensity and matching, 39t Quality Assurance Resource Center, Pakistan, 77 public education sector, financing and service quality assurance system, 51­53, 67 provision, 3f Quality Education for All, Pakistan, 79 public-private partnerships (PPPs) for education, 1, 2, 5, 64 R access and quality improvement, 28­29 bringing in alternative operators, 22­28 randomization, 37t, 38, 39t continuum, 16­17, 16f, 61 randomized studies, 83 definition and objectives of, 31­32 regression discontinuity analysis, 83 designing, 33, 44b, 53­58 regulation weaknesses, 43, 45 education impacts, 33­34 regulatory environment, 42, 44b empirical evidence, 7, 35­36, 37t, 38­40 regulatory framework, 18, 65­67 endogeneity, 35, 36 religious schools, 8 evaluation methods, 83 risk allocation, 28 financing, 3f, 18 risk-sharing, 4, 17, 34 flexibility, 34 by PPP type, 35t framework for understanding, 59­61 funding of existing private schools, 15, 18­21 S funding-based characteristics, 49­50 government promotion of, 43b SABIS, 46, 53 impacts on education objectives, 32t scholarships, 72, 75 implementing, 44b school choice, 7, 8, 18­20, 30 incentives and sanctions, 56­57 school evaluation and review services, 52 international experiences, 61­63 school management, 33­34 international involvement in, 57­58 school management program, Pakistan, 79 levels of engagement, 17­18 school performance, 15, 52 management by specialized institution, 21b, school types, 2­3 29 school-based management, 19 Index 97 schools. See also charter schools; concession private finance initiatives for infrastructure, schools; education 25­26, 63­64, 82 accountability and student selection, 28 Schools Private Finance Initiative, 55 allowing profit and not-for-profit, 46, 66 United States, 3, 24, 63 contracting benefits, 63 charter schools, 39t, 80 publicly funded and privately managed, evidence on, 64­65 24­25, 35 split of purchaser and provider academic achievement, 63 functions, 54 improve efficiency and quality, 62 vs. vouchers, 25b subsidies and private vs. public, 39­40 contract schools, 80 Schools Private Finance Initiative, United supplemental education services, 23­24 Kingdom, 55 supplemental PPP programs, 77 Scotland, PPP program description, 82 voucher programs in Milwaukee and secondary school, 3, 8, 20, 21 Florida, and Puerto Rico, 76 self-selection, 31, 35, 64 Urban Girls Fellowship, Pakistan PPP Senegal, scholarship program, 75 program, 74 service purchasing contract, 14 Seventh Day Adventist Church, 46 V Sindh Education Foundation, 29, 52­53 socioeconomic segregation, 4, 18­19 Venezuela, Republic Bolivariana de, 24, 39t, sorting, 38 76, 80 South Africa, 62, 75 Venezuelan Association of Catholic Schools Standard and Poor, 52 (AVEC), 76 student selection, 28, 30 voucher programs, 19, 32t, 35t subsidies, 30, 32t, 35, 35t and controversy of, 65 Bangladesh, 71 Belgium, 71 empirical evidence for, 38­39, 40t characteristics of, 61­62 enrollment increase, 32 Chile, 6, 7, 30, 65, 72 for private schools, 66­67 Colombia, 36 Mauritius, 73 Czech Republic and Denmark, 72 New Zealand, 74 empirical evidence for, 36, 38 self-selection in programs, 64 endogeneity, 31 South Africa, Korea, and Thailand, 73, enrollment increase, 32 75 features, 19 supplemental services, 23 Foundation Assisted Schools Program, supplemental services' programs, 76­77 Pakistan, 21, 21b support services, 10, 22­23, 60, 62 Hungary, Italy, and the Netherlands, 73 support services' programs, 76­77 oil, 7 Sweden, 19­20, 37t, 75 Pakistan, 74 parent participation model, 38 T Puerto Rico, Qatar, and Sweden, 75 quality and competition, 35 tax credits, 49 reducing inequity, 20 tax incentives, 47 sorting, 38 teacher training, 22. See also support services targeting to the underserved, 62 tertiary education, accreditation, 53 vs. charter schools, 25b testing services, 52 Thailand, subsidy program, 75 W The Educators, 49 Trehoje School, Denmark, 81 World Bank, 1­2, 69, viii tuition, 11, 46 Balochistan Education Support Project, 49 Tutorial Voucher Initiative, Australia, 76 education lending, 58 Haiti education grant, 19b U World Development Report 2004, 9, 25, 63 Uganda, 21, 75 Z United Kingdom, 23, 24, 80 Assisted Places Scheme, 76 Zimbabwe, targeting low-income students, Education Action Zones, 80 20 ECO-AUDIT Environmental Benefits Statement The World Bank is committed to pre- Saved: serving endangered forests and natural · 12 trees resources. The Office of the Publisher has · 9 million British chosen to print the The Role and Impact of Thermal Units of Public-Private Partnerships in Education total energy on recycled paper with 10 percent post- consumer fiber in accordance with the · 1,085 pounds of recommended standards for paper usage CO2 equivalent set by the Green Press Initiative, a non- greenhouse gases profit program supporting publishers in · 4,502 gallons of using fiber that is not sourced from endan- waste water gered forests. For more information, visit · 578 pounds of solid www.greenpressinitiative.org. waste E nhancing the role of private sector partners in education can lead to signi cant Given the educational expansion in recent years, it has become improvements in education service delivery. However, increasingly clear that the public sector the realization of such bene ts depends in great part on alone cannot hope to deliver what is the design of the partnership between the public and needed for an e cient and equitable private sectors, on the overall regulatory framework school system. is book provides of the country, and on the governmental capacity to excellent arguments in favor of public- oversee and enforce its contracts with the private sector. private partnerships as a solution to Under the right terms, private sector participation in the problem. It should be required reading for those who, based on education can increase e ciency, choice, and access ideology rather than evidence, are still to education services, particularly for students who skeptical of private sector involvement tend to fail in traditional education settings. Private in education. for-pro t schools across the world are already serving George Psacharopoulos, formerly of the a vast range of users--from elite families to children London School of Economics and the in poor communities. rough balanced public- World Bank private partnerships (PPPs) in education, governments Children's well-being and the can leverage the specialized skills o ered by private quality of education are not organizations as well as overcome operating restrictions always the main criteria for policy such as salary scales and work rules that limit public decisions. is book will contribute sector responses. to a more responsible analysis of new ways of organizing education. e Role and Impact of Public-Private Partnerships in Enrique Peñalosa, Urban Strategist and former Mayor of Bogotá, Colombia Education presents a conceptualization of the issues related to PPPs in education, a detailed review of An authoritative summary of what rigorous evaluations, and guidelines on how to create is known about the extent and successful PPPs. e book shows how this approach can e ectiveness of the current growth in facilitate service delivery, lead to additional nancing, government scal support for privately expand equitable access, and improve learning managed schools. outcomes. e book also discusses the best way to set Paul E. Peterson, Shattuck Professor of Government and Director, Program up these arrangements in practice. is information on Education Policy and Governance, will be of particular interest to policymakers, teachers, Harvard University; Senior Fellow, Hoover Institution; Editor in Chief, researchers, and development practitioners. Education Next ISBN 9780821378663 SKU 17866