Those familiar with the debate in the developed world about the virtues of private versus public education will not be surprised to find the same arguments raging when it comes to providing education to some of the poorest children in the world. On the one hand, there is the principle enshrined in public schools of providing universal access to the same quality of education. On the other, there are private schools that offer quick fixes to the endemic failures of the public system. They go by a variety of names in the West—charter schools in the United States, Free Schools in Sweden—but they aim to leverage private money, whether in fees or in contributions from business, to give children a good education sooner rather than later.

In the developing world, some of the distinctions between public and private are blurred: in places like Kenya, private schools that cost up to $5 a day, just in reach of the poor, may not offer an education that is very different from what the public system offers. That is, a single classroom that a local woman runs in her house. Meanwhile, the public system is rarely free. Parents are assessed school fees, some legitimate, others bribes to the teachers.

As the United Nations contemplates its next targets for global education attainment following on the Millennium Development Goals, there has been a renewed interest in how to provide better quality education to the poor in developing countries. A number of fascinating public and private initiatives have been launched or expanded over the last couple of years. They include the Bridge private school system in Kenya, funded by the Pearson Group to provide a systematized curriculum in return for a small fee, and BRAC in Bangladesh, a nongovernmental organization (NGO) which partners with governments to provide free education to the hardest to reach.

The result has been to reopen the debate on where the international community should focus as it seeks to expand access to education for the world’s poorest. In a departure from Solutions’ usual format, we are presenting the views of two experts who advocate for different sides of the private versus public debate. They are Justin Sandefur of the Center for Global Development and Kevin Watkins of the Brookings Institute. Given the scale of the challenge—over 60 million children have no access to education, and some 1 billion children living in poverty (below $2.50 dollar a day) receive a limited or poor one—we hope you agree that all solutions are welcome

Justin Sandefur

By one count, over half of countries were on track to meet the Millennium Development Goal (MDG) for primary education as of 2011. But going to school is not an end in and of itself. And the push for universal primary school enrollment has been an abject failure in terms of what really matters—learning.

Kenya is a good example, where enrollment and learning diverge. Behold the following reading passage from Kenya’s public school curriculum.

Look at my hair.

It is long and black.

I like my hair.

I wash it on Sunday.

According to a national survey by Uwezo Kenya, an assessment organization, only half of children in third grade can read this type of paragraph, although it’s on the second-grade syllabus. So while it’s great that only 3 to 4 percent of primary-school-age children in Kenya are not in school, it’d be nice if they could also read.

But the real scandal here, as highlighted by my colleague Lant Pritchett and coauthor Amanda Beatty, is not that half of third graders effectively can’t read, it’s that staying in school doesn’t help much. This is what Pritchett and Beatty call the “flatness of the learning curve.” Of Kenyan children who couldn’t read the paragraph above in grade three, only a third will learn to do so in grade four—in India, Pakistan, and Tanzania only one in five will learn to read with an additional year of schooling, and in Uganda only one in twelve. Even after eight years of schooling in India, almost one in three pupils still won’t be able to read a simple paragraph like this.


Students at a BRAC boat school in Shunamganj, Bangladesh. In 2011, BRAC launched boat schools to offer primary education in areas affected by flash flood.

Kids are going to school; they’re just not learning anything.

In the past, there has been little evidence to suggest the private schools provide better service than the public sector, but that’s changing.

For places including India, Pakistan, and Kenya, there are at least two types of evidence that private schools offer much better service. Call them the direct approach and the indirect approach.

The direct approach simply compares test scores between public and private schools. For instance, by grade three, students at the Bridge International Academies in Kenya—one of the schools Pearson will support—are scoring roughly 90 percent higher on reading fluency and 45 percent higher on comprehension than their neighbors.

The obvious follow-up question is whether this is just “garbage in, garbage out,” whereby private schools pick the best pupils and produce the best scores. To test this hypothesis, my colleagues Tessa Bold (Goethe University), Mwangi Kimenyi (Brookings), Germano Mwabu (University of Nairobi), and I did some relatively simple analysis of the gap in test scores between public and private schools. We find a performance gap between public and private schools of roughly one full standard deviation, which is more than seven times larger than the impact of the best-documented, successful interventions to improve public schools in Kenya, which find an effect of about 0.14 standard deviations. More importantly, this gap does not appear to be due to the selection of smart kids into private schools. When private enrollment goes up in a district, overall test performance rises as well—a phenomenon that can’t be explained merely by the sorting of good pupils into good schools.

In a similar study of test scores in India, Alex Tabarrok from George Mason University finds not only that private primary schools vastly outperform public schools, but again, that this result is mostly not due to “cream skimming” of the best pupils. As more and more students flock to private schools—exceeding 70 percent of all pupils in some districts—the public-private gap doesn’t narrow.

The direct approach clearly signals huge gains from private schooling. How about the indirect approach?

By economists’ logic of “revealed preference,” the stampede of pupils from the public to the private sector would seem to be a strong indication of what parents think of the quality of public schools. A study by Michael Kremer of Harvard and Karthik Muralidharan of University of California-San Diego found that over 80 percent of government-school teachers in India send their own children to a private school.

As Muralidharan summarized for a New York Times reporter, “What does it say about the quality of your product that you can’t even give it away for free?”

The major challenge for any private education system is to find the right price point. In the case of the developing world, that means finding a figure low enough to draw in lower-income families while still meeting the costs of running a school. For Westerners used to associating the word private with expensive, it can be eye-opening to realize how low tuition fees can be in the developing world.

Research by Tahrir Andrabi (Pomona College), Jishnu Das (World Bank), and Asim Khwaja (Harvard) shows that the average fee of a rural private school in Pakistan is less than a dime a day. They also find that villages where private schools emerge have a significantly smaller gender gap in enrollment. In Kenya, my colleagues and I examined fees paid by parents, and calculated that two-thirds of private schools cost less to operate than the median public school.

In India, the previous study by Kremer and Muralidharan finds that provinces and districts with lower per capita incomes have higher rates of private schooling. Rather than being driven by wealth and privilege, they find that demand for private schooling is associated with the failure of public school systems. Private schools are significantly more likely to emerge in villages where public school teachers are frequently absent, or frequently not teaching when they show up.

There’s a risk of overselling the case for private schools. After all, they do charge fees.

Separating the provision of education (by private schools) and its financing (by government) requires initiatives like voucher programs and charter schools. The challenges of designing a good voucher or charter school program are not trivial. Countries as diverse as Sweden and Colombia have introduced vouchers in a way that improves overall quality without compromising equal access. Chile got things wrong—after introducing vouchers, Chile saw a massive exodus to the private sector, and increased socioeconomic stratification between schools, arguably because Chile allowed schools receiving vouchers to conduct selective admissions and charge top-up fees, encouraging schools to compete on who they could attract, not what they taught.

The real frontier in research is not about whether private schools work, but how we can support this market so it ends up looking more like Sweden’s rather than Chile’s.


Sweden has had success with its private education system, introduced in 1992, according to Sandefur. But a number of cases, like Chile, demonstrate that success is not at all guaranteed.

We should avoid the urge to squash the private sector. This urge is reflected in Kenya’s decision to penalize private school graduates in secondary school admissions, and in some of the more controversial pieces of India’s Right to Education Act. Private schools are laboratories for approaches that can be applied in the public sector, and can be an important tool in extending access to quality education. [Editor’s note: The Kenyan policy was designed to correct an imbalance whereby seven times more graduates of private school enter higher education than from the public school system.]

No one is calling for mass privatizations, whatever that means. I don’t think we currently have sufficient evidence based on the impacts of vouchers in low-income settings to even justify rapid expansion of these programs. But the UN Declaration says everyone has a right to education. From Lahore to rural Balochistan, and Nairobi to the farthest reaches of the Rift Valley, parents are no longer waiting for superman. They have accepted that neither the public sector nor the aid industry is going to sweep in and save their children from broken schools, so they’re taking matters into their own hands

Kevin Watkins

I share many of Justin’s concerns. The learning crisis is one of the great development challenges of our day. And, like Justin, I want to see some bold new experiments in education—a sector paralyzed by the conservatism of aid donors, government indifference, and weak leadership from the UN and the World Bank.

I have no interest in defending the indefensible quality of public education provided in many of the poorest countries. But when public education systems are broken, they need fixing, not bypassing or franchising out to the private sector. And if we care about equity, there is no credible alternative to a public system that offers opportunity for all rather than choice for some.

There are currently around 61 million primary-school-age children out of school. Since 2005 progress towards universal primary education has slowed globally and stalled in sub-Saharan Africa, where out-of-school numbers are rising. Of the 126 million primary-school-age children in sub-Saharan Africa, around two-thirds are likely to enter adolescence unable to read, write, or do basic numeracy, irrespective of whether or not they complete primary school. Research by Jishnu Das in India and Pakistan suggests that close to one half of the children in these countries face the same prospect.

What is driving the crisis in learning? The causes are complex and vary across countries. Many of the children entering public education systems over the past decade arrived carrying huge handicaps, including household poverty, parental illiteracy, and pre-school malnutrition—an affliction for around 175 million children. These are disadvantages that impact heavily on learning.

The school environment is another concern. Across much of sub-Saharan Africa, the pupil-teacher ratio exceeds 40 to one and there is less than one book for every three children. Overcrowding is typically worst in the early grades where children from nonliterate homes need most support.

The quality of teachers and teaching is one of the most critical school-based determinants of learning. Unfortunately, teacher absenteeism is rife. Many teachers lack subject knowledge. Trained to deliver teaching by rote, they are ill-equipped to deliver active learning. In-service support is lacking. To make matters worse, public education systems typically skew resources and the best teachers towards the most advantaged pupils, best-performing schools, and most prosperous regions.

Sadly, fee-paying schools are no heal-all. As far as I am aware, no credible commentator has ever questioned that there is a private-public school performance gap. The question is whether that gap disappears with the introduction of appropriate controls for differences in the school environment, student characteristics, and the household environment.

Furthermore, there is no shortage of research documenting the struggles of poor households to pay “low-fee” providers. One village-level survey in rural western Uttar Pradesh, India, has found that low-fee schools are unaffordable to the poorest two quintiles, and that the growth of private provision has reinforced education inequalities linked to wealth, caste, and gender. When asked, many of the parents paying for low-fee private school say that they would prefer the option of sending their children to a public school that offered decent education—a revealed preference that advocates of private provision often ignore.

Evidence from urban informal settlements is equally compelling. In Lagos, it costs the equivalent of around 10 percent of the minimum wage to send one child to an approved, low-fee private school. This is in a country where one third of households with children who have dropped out of school cite education costs as the reason for their nonattendance. In Kenya, Moses Oketch and others have highlighted the lack of access to public education for low-income households in informal urban settlements, leaving them with no choice other than to attend private schools. The resulting cost barriers are restricting progress towards universal primary education. The bigger question is this: why should we tolerate a state failure that leaves some of the world’s poorest households facing prohibitive user-fees to secure their children’s right to education? Voucher-based systems, which Justin seems to prefer, are more a recognition of chronic inequality that a prescript for increasing access to all. (The only national-scale study of vouchers in the United States, conducted by Stanford University, reported that only 17 percent of charter schools outperform matched neighborhood public schools.)


Matt Berlin/Flickr
A public classroom in Kenya. The issue is less about whether or not private education is successful than it is about “whether we care about equity,” says Watkins.

Much of our debate focuses on school management. Of course, the issues are important. But I can’t help wondering whether the endless dialogue on public-versus-private provision is diverting attention from far more important themes that affect education and learning outcomes—and from the policies that can make a difference.

Let me give a few examples. We know that one in every three children in low-income countries reach school age having experienced extreme malnutrition, most of them before the age of two. This has devastating and largely irreversible consequences for cognitive development of learning achievement in school. This largely ignored crisis can be tackled through early childhood interventions—a point underlined by a recent randomized trial survey from a Save the Children program in Mozambique.

If we want to get all children into a learning environment, we need to reform public spending programs that skew resources towards schools in wealthier regions; and we need demand-side financing interventions—incentives for keeping girls in school and getting kids out of child labor—that break down the disadvantages associated with poverty, gender, and other markers for disadvantage.

If we want to raise learning standards, surely the discussion should focus on how to recruit, train, remunerate, and support teachers. One of the reasons that so many of the children in school are failing to learn is that teachers are ill-equipped to develop basic literacy and numeracy skills in the early years, setting the scene for failure.

Without wanting to develop a shopping list, I would add the accountability dimension. From central ministries down to classrooms, public-sector education providers in poor countries are, for the most part, notoriously unaccountable to parents. So are their low-fee private-sector counterparts: the idea that a shift from public to private provision brings greater accountability is naïve. Civil society coalitions and nongovernment organizations have played a positive role in challenging the culture of impunity among education providers. In some cases—like UWEZO in east Africa and Pratham in India—they have done this by making available information on learning. In others, they have supported the development of parent-teacher associations and used social accountability tools to hold providers to account. And it strikes me that this empowerment aspect of reform merits greater attention.

The narrow focus of the private-public debate in low-income countries has a parallel in the United States. Here the “school reform” movement appears to attribute 90 percent of the crisis in public education to teachers, the absence of school choice, and school management issues, with 10 percent attributed to poverty, social breakdown, and social disadvantage. As Dianne Ravitch has powerfully argued, they have the numbers back to front.

So, in terms of a solution, let’s start by admitting there are no quick fixes.

Private schools clearly have a role to play in achieving the goal of education for all—and far more should be done to ensure that education strategies include a proper regulatory framework for private providers. Ultimately, though, governments need to take responsibility for fixing school systems that are failing.

We know the strategies that can lead countries out of the low-learning trap. Reform of teacher recruitment, training, and support; the development of national learning assessment systems to identify failing schools and pupils; a stronger focus on preschool provision and early grade teaching which are among the most important determinants of learning and future well-being; and more equitable public financing all have a role to play. As Barbara Bruns of the World Bank has documented, this is the public education reform path that Brazil has followed—and the country is now one of the world’s fastest climbers in the international learning assessment league table.

Now that’s what I call bold experimentation—and it has delivered results.

For a full list of hyperlinked references, please see the extended, four-part online interview, starting with


Justin Sandefur

Justin Sandefur is a research fellow at the Center for Global Development. His research focuses on the interface of law and development in sub-Saharan Africa. From 2008 to 2010, he served as an adviser...


Kevin Watkins

Kevin Watkins joined the Overseas Development Institute as executive director in June 2013. He is a former nonresident senior fellow with the Center for Universal Education at the Brookings Institution,...


Duncan Green

Duncan Green is the author of From Poverty to Power and Oxfam GB’s senior strategic adviser. He was Oxfam’s head of research from 2004-12. From Poverty to Power contains the accumulated knowledge of...

Leave a comment

Your email address will not be published. Required fields are marked *