Private Schools for the Poor: A Challenge to the Global Orthodoxy

Prof. James Tooley

Introduction

Many believe that the private sector has very little to offer in terms of reaching the United Nations Millennium Development Goal of “education for all” by 2015.  Private education is often assumed to be concerned only with serving the elite or middle classes, not the poor. And unregistered or unrecognized private schools are thought to be of the lowest quality and hence demanding of detailed regulation, or even closure, by governmental authorities.  My findings from a two-year in-depth study in two Asian countries – India (Hyderabad and Delhi) and China (Gansu Province) – as well as in three countries in sub-Saharan Africa – Ghana, Nigeria and, Kenya – suggest that these conclusions may be unwarranted. Private schools, I argue, can play – indeed, already are playing – an important, if unsung role in reaching the poor and satisfying their educational needs.

The research has challenged two “myths” that appear to be part of the consensus about private education and the poor, challenges that form the basis of this brief outline[1].

Myth 1: Private education for the poor does not exist

Undertaking this research was disheartening at first. In each country I visited officials from national governments and international agencies that donate funds for the expansion of state-run education denied that private education for the poor even existed. Usually I met with polite, if embarrassed, apologies that always went something like, “Sorry, in our country, private schools are for the privileged, not the poor.” In each venue, however, I struck out on my own and visited slums and villages and there found what I was looking for: private schools for the poor, usually in large numbers, if sometimes hidden from view.  In this section, I describe the situation in two settings – Hyderabad, India, and Gansu Province, China.  Similar results were found in each of the other settings surveyed.

Hyderabad, India

Visit the ultra-modern high-rise development of “High Tech City” and you’ll see why Hyderabad dubs itself “Cyberabad,” proud of its position at the forefront of India’s technological revolution.  But cross the river Musi and enter the Old City, with once magnificent buildings dating to the 16^th century and earlier and you’ll see the congested India, with narrow streets weaving their way through crowded markets and densely populated slums. For our survey, we covered three zones in the Old City, Bandlaguda, Bhadurpura, and Charminar, with a population of about 800,000 (about 22 percent of Hyderabad), covering an area of some 19 square miles. We included only schools that were found in “slums,” as determined by the latest available Census and Hyderabad Municipal guides. There were areas that lacked amenities such as indoor plumbing, running water, electricity, and paved roads.

In these slum areas alone our team found 918 schools: 35% were government-run, 23% were private schools that had official recognition by the government and, incredibly, 37% that flew under the government radar (“unrecognized”).  The latter group of schools is in effect a black market in education, operating entirely outside of both state funding and regulation. (The remaining 5% were private schools that received a 100% state subsidy for teacher salaries, making them public schools in all but name.) In terms of total student enrollment in the slum areas of the three zones, 65% of all the school children attended either recognized or unrecognized private schools that received no government assistance – with roughly the same percentage of children in the unrecognized private schools as there were in government schools.

Two points need to be made about these numbers: first, although we know that we found all government and recognized private schools (because there were government lists with which to compare them), there was no way to verify whether or not all unrecognized private schools were found. So these numbers provide a conservative estimate of the number of private schools; the true number is almost certainly higher. Second, because the unrecognized schools are not included on any government lists, there are significantly more children in school than is recorded in official statistics. This clearly has implications for any discussion of how to achieve universal primary education (“education for all”): In the ‘slum’ areas of three zones of Hyderabad, we found 79,851 students in private schools that were not on government lists (around 30% of the total number of school children in those areas).  But a recent report from the Azim Premji Foundation, using official figures, suggested that for the 35 zones that make up Hyderabad District, 129,000 children are out of school, that is, 15.4% of the total 837,212 school-age children (aged 5 to 15) in Hyderabad. It is likely that many of these children would be in the three zones surveyed – we chose them because they were reportedly some of the poorest. If all the out of school children were in these zones, then this would mean a maximum of about 49,000 children actually out of school in Hyderabad, if our figures are correct – the remaining accommodated in private unrecognized schools that are missed in official figures. Instead of 15.4% out of school, the figure would be sharply reduced to only about 6%. (See table 1 below). More realistically, if some of the officially ‘out of school’ children are spread over the 32 other zones, then the actual figure of out of school children would be even lower. It is surely easier to bring 6% or fewer children into school than it is to bring in 15%. India’s ‘education for all’ goal may thus be much easier to reach than official sources claim.

Table 1 Education for all in Hyderabad

In the slums of Hyderabad, India, a typical private school would be in a converted house, in a small alleyway behind bustling and noisy streets, or above a shop. Classrooms are dark, by western standards, with no doors hung in the doorways, and noise from the streets outside easily entering through the barred but unglazed windows. Walls are painted white, but discolored by pollution, heat and the general wear-and-tear of the children; no pictures or work is hung on them. Children will usually be in a school uniform and be sitting at rough wooden desks. Generally, there are about 25 students in a class, a decent teacher-to-student ratio, but the tiny rooms always seem crowded. Often the top floor of the building will have various construction works going on, to extend the number of classrooms that can be accommodated. The school proprietor will usually live in a couple of rooms at the back of the house.

What is clear from our research is that these private schools are not mom-and-pop day care centers or living room home schools. The average unrecognized school had about 8 teachers and 170 children, two-thirds in rented buildings of the type described above. The average recognized school was larger and usually situated in a more comfortable building, with 18 teachers and about 490 children. Another key difference between the recognized and unrecognized schools is that the former has stood the test of time in the educational market: 40% of unrecognized schools were less than 5 years old while only 5% of recognized schools were this new. Finally, tuition in these schools is very low, averaging about $2.12 per month in recognized private schools at first grade and $1.51 in unrecognized schools.

While these fees seem low, they must be measured against the average income of each earning member within the extended family households.  In the unrecognized schools, this was about $23 per month (and about $30 per month for recognized schools and $17 for government schools).  Given the fact that the official minimum wage in Hyderabad is $46 per month, it is clear that the families in our private schools are poor.  Fees therefore amount to about 7% of average monthly earnings in a typical private unrecognized using household. For the poorest children, the schools provide scholarships or subsidized places – we found that 7% of children paid no tuition and 11% paid reduced fees. In effect, the poor are subsidizing the poorest. 

Gansu, China

With 25.3 million people spread out over an area the size of Texas, Gansu Province is a mountainous region situated on the upper and middle reaches of the Yellow River in Northwest China.  It has an average elevation of over 3,000 feet and 75 percent of its population is rural, with illiteracy rates among people aged 15 or higher at nearly 20% for men and 40% for women. Roughly half of its counties, with 62% of the population, are considered “impoverished.”

Figures from the Provincial Education Bureau show only 44 private schools in the whole province, all of which are for privileged city dwellers. Given the paucity of information on private schools, the research teams surveyed each major town in each of the counties designated as impoverished (over 40 of them), and to visit as many of the villages that were accessible to them from these towns. In the early stages I wasn’t worried about getting precise estimates of the numbers of schools or the proportion of children in them, but rather wanted to see if such schools even existed.

In the major towns and the larger villages, crowded and bustling, there is a public school, usually a fine two-story building that often sports a plaque marking it as a recipient of some foreign aid. But then I abandoned my car and either walked or hitched a ride on one of the ubiquitous and noisy three-wheeled farm vehicles to travel up the even steeper mountain paths to clusters of houses in smaller villages. And there, nestled on mountain ridges, were stone or brick houses converted to private schools, with the proprietor or headmaster living with his family in one or two of its rooms. Occasionally, the school had been built, by the villagers, to be used as a school. Over and over again researchers followed these trails high into the arid mountains and, in the end, discovered  a total of 696 private schools; 593 of them serving some 61,000 children in the most remote villages.    

The comments above about not finding all of the schools apply even more strongly here. Notably, in Ding Xi county alone, one of the poorest in Gansu, the researchers here found 293 of the 593 village private schools – which suggests either that they are more common there, or that the researchers assigned to that region did a more thorough job of finding the schools. In one of the localities in this county, Wei Yuan, population 300,000, we were confident enough that we had found all of the private schools to estimate that 15% of school children were in private school. We currently have no way of checking if a similar proportion is in private school elsewhere, but have no reason to believe that Wei Yuan is significantly different from other localities.

Not surprisingly, the vast majority of Gansu’s private schools were set up by individuals, or the villages themselves, because government schools are simply too far away or hard to get to.  Significantly, the majority of the private schools found were in the three poorest regions of Gansu, where net income per year ranges from $125 to $166 -- private schools serving some of the poorest people on the planet.  But, surprisingly, these private schools, which depend on tuition, are also cheaper than government schools.  Average fees for a Grade 1 elementary school student are about $7.60 per semester, compared to about $8.00 in the public schools, not an insignificant difference to someone living on $125 per year.

Myth 2:  Private education for the poor is low quality

It is a common assumption among development experts that private schools for the poor are worse than public schools. This is not to say that they have a particularly high view of public education. Indeed, the World Bank’s World Development Report 2004: Making Services Work for Poor People calls it “government failure,” with “services so defective that their opportunity costs outweigh their benefits for most poor people.” Yet this just makes the experts’ dismissal of private schools for the poor all the more galling.

The Oxfam Education Report published in 2000 is typical.  While the author acknowledges the existence of high quality private providers – no one seems to dispute this – he contends that these are elite, well-resourced schools inaccessible to the poor. As far as private schools for the poor are concerned, these are of “inferior quality”; indeed they “offer a low-quality service” that is so bad it will “restrict children’s future opportunities.” This claim of low quality private provision for the poor has also been taken up by Prime Minister Tony Blair’s Commission for Africa, which recently reported that although “Non-state sectors … have historically provided much education in Africa,” many of these “aiming at those who cannot afford the fees common in state schools… are without adequate state regulation and are of a low quality.”

However, these development experts have little hard evidence for their assertions about private school quality.  They instead point out that private schools employ untrained teachers who are paid much less than their government counterparts, and that buildings and facilities are grossly inadequate, both of which are largely true. But does that mean that private schools are inferior -- particularly against the weight of parental preferences to the contrary? One Ghanaian school owner challenged me when I observed that her school building was little more than a corrugated iron roof on rickety poles and that the government school, just a few hundred yards away, was a smart new school building.  “Education is not about buildings,” she scolded. “What matters is what is in the teacher’s heart. In our hearts, we love the children and do our best for them.” She left it open, when probed, what the teachers in the government school felt in their hearts towards the poor children.

Facilities and Resources

The issue of the relative quality of private and public schools was at the core of our research, and we relied on both data on school resources and day-to-day operations and on student achievement scores. Our researchers first called unannounced at schools and asked for a tour, noted what teachers were doing, made an inventory of facilities, and administered detailed questionnaires.

Certainly, in some countries the facilities in the private schools were markedly inferior to those in the public schools. In China, where the researchers were asked to locate a public school in the village nearest to where they had found a private school, often many miles away, private school facilities were generally worse than in those publicly provided. This was predictable, given that the private schools undercut the public ones in terms of fees, and served the poorest villages where there were no public schools. In 88% of private schools in Gansu Province desks were available in classrooms, compared to 97% of public schools; 66% of private schools had chairs or benches in classrooms, compared to 76% of public schools.

In Hyderabad, however, on every input, including the provision of blackboards, playgrounds, desks, drinking water, toilets and separate toilets for boys and girls, both types of private schools – recognized and unrecognized – were superior to the government schools. While only 78% of the government schools had blackboards in every classroom, the figures were 96% and 94% for private recognized and unrecognized schools, respectively. In only half the government schools were toilets provided for children, compared to 100% and 96% of the recognized and unrecognized private schools. 

When it came to the key question of whether or not teaching was going on in the classrooms, in China there was no statistically significant difference between the two school types: 92% of teachers in private schools were teaching when our researchers arrived, compared to 89% in the public schools.  When researchers called unannounced on the classrooms in Hyderabad, however, 98% of teachers were teaching in the private recognized schools, compared with 91% in the unrecognized and 75% in the government schools.  Teacher absenteeism was also highest in the government schools.     It was also the case that private and public schools in China had more or less the same pupil-teacher ratio, about 25:1.  In Hyderabad private schools, including the unrecognized ones had significant advantages over the government schools: the average pupil-teacher ratio was 42:1 in government schools, compared to only 22:1 in the unrecognized and 27:1 in the recognized private unaided schools.

Student Achievement

To compare the achievement of students in public and private schools in each location where we conducted research, we first grouped schools by size and management type: government, private unrecognized and private recognized in Hyderabad and government and private in Gansu, China.   We tested a total of roughly 3,000 students in each setting in English or Chinese and mathematics. All children were also given IQ tests, as were their teachers. Finally, questionnaires were distributed to children, their parents, teachers, and school managers, seeking information on family backgrounds.

Our analysis of these data is still in progress. However, in all cases analyzed so far – including Hyderabad and Delhi – students in private schools achieved above the levels achieved by their counterparts in government schools in both English and mathematics, (see Table 2) as well as Hindi in Delhi. Moreover, the private school advantage was maintained after taking into account differences in an unusually rich array of characteristics of the students, their families’ economic status, and the resources available at their schools.[2] 

Table 2 Hyderabad – raw scores

Source: Survey of Achievement Data

In short, it does not seem to be the case that private schools serving low-income families are inferior to those provided by the state. In all cases analyzed, even the unrecognized schools – those that are dismissed by the development experts as being obviously of poor quality – seem to outperform their public counterparts. 

Lessons for Education for All

So, the accepted wisdom appears to be wrong. Though elite private schools do exist in impoverished regions of the world, private schools are not only for the privileged classes.  From a wide range of settings, from deepest rural China, through the slums of urban India and Kenya, to the peri-urban areas of Ghana, private education is serving huge numbers of children. Clearly the evidence presented here may have implications for the ongoing policy discussions over how to achieve universal education worldwide and for development policy. William Easterly, in his Elusive Quest for Growth notes the ineffectiveness of past investments by the international agencies and developing country governments in public schools, pointing out that “Administrative targets for universal primary education do not in themselves create the incentives for investing in the future that matter for growth,” that is, in quality education. If development agencies could think instead about investing in private schools, then genuine quality educational improvement could result. Strategies to be considered include offering loans to help schools improve their infrastructure or worthwhile teacher training, or in creating partial vouchers to help even more of the poor access the private schools that are ready to take them on.