The Need to Strengthen Government Schools in India
By: V Santhakumar
Some economists advocate the use of financial incentives and privatization to address the issue of poor schooling in developing countries like India. The advocacy for a greater dependence on private schools through vouchers or other ways 1 is part of this approach. There are others who defend government schools with normative views and also by indicating that public education plays an important and dominant role 2 in the developed countries. This essay uses mainly the insights from Economics (including those from recent developments in the discipline) to argue for a greater dependence on government schools. This is written for general readers who may not have a background in Economics.
In Economics, government intervention is advocated when the ‘market fails’ to provide the goods and service that are needed by the society. ‘Public goods’ are those goods or services which are non-excludable and indivisible for example, the street light. To exclude one person from getting the benefit of a street light is not only difficult but also costly. One person’s use of the street light does not reduce the availability of this light to others unless there is congestion. However, education is not such a public good. It is excludable and divisible. Excludable, because it is not difficult to exclude a person from the provision of schooling, and divisible, because given that the number of students that can be accommodated in a class is limited, each child’s use, reduces the seat available to the others.
But Education is a public good in another sense – in what is called, ‘positive externality’. In cases of such externality, one person’s action creates benefits not only for him or her but also for others in the society. The latter get these benefits ‘free of cost’. This is true for schooling too. There are multiple benefits for the society as a whole when the people it comprises, receive education. It may help in increasing the pool of skilled workers leading to possible positive benefits on the economic development or it may reduce family size, infant mortality rate, and improve health indicators for all. It may also reduce crime rates. These positive external benefits of schooling are a feature of public goods. It is difficult to exclude somebody from enjoying the benefits of an educated society, and one persons’ gain in this regard may not deprive others. This could be the reason for calling education a quasi-public good.
There is also the issue of ‘negative externality’ warranting government intervention because the lack of schooling of some people may impose a certain cost on others in society3. At the same time, certain kinds of ‘schooling’ could be harmful to others in society. For example, if schools inculcate intolerance towards religions other than one’s own, it can have a negative impact on the society. Hence, a society may decide the broad outline of the ‘content’ of education provided in all schools.
Education can also be part of, what is called, ‘merit goods’. These are cases where the society is not willing to leave the decision of consumption to individuals. If it is left to them, some may not consume adequately and some, more than what is desirable because of their lack of information or ability to decide what is suitable for them. For example, poorer people may not use schooling for their children, this may be due to their lack of awareness of how education contributes to the life of an individual and the society. This is the most substantial argument in favour of mandatory schooling. This ‘merit goods’ argument can also be extended to the discussion on the pedagogy of education. How do children learn or what should be taught or the methods of teaching (specific subjects) are matters of professional knowledge, not accessible to all and therefore these matters cannot be left to individual discretion. For example, whether to have the pre-school education or not? The affirmative answer to this is backed by professional knowledge which, if followed, will be beneficial to all. This could be another reason for societal or governmental regulation of not only the content but also the nature of schooling and pedagogy.
There is information asymmetry in education4. This is the reason why standardized tests are conducted for students and teachers. Standards in teacher education signal certain information on the attributes and abilities of the people who work as teachers. Testing of students by examination boards provide information, or reliable signals, on their `abilities’ (for example, to their potential employers). But the performance of schools or teachers cannot be evaluated and ensured merely by the users (for example, the parents).
All these reasons justify the need for societal/governmental intervention to see that schooling is available and used by all children of a relevant age group5. This may require enforcement of mandatory schooling and ensuring that a certain level of educational facilities is available and accessible to all. Moreover, governments may decide the broad outline of the content and nature of education provided in all schools.
2.1 Can there be government intervention in schooling without government schools?
It may be argued that ‘education for all’ is possible with governments providing money to poorer parents who can then choose a school (government or private) for their children6? Or the government can fund different kinds of organizations (including non-governmental and private) which want to start a school and these can cater to the needs of students in different localities.
This is the principle behind government-aided schools in India. Such schools comprise the majority – 55% of all schools – in the state of Kerala. In addition to the funding, whether to parents or to schools, governments may be regulating education in all schools (by prescribing curriculum, outlines of pedagogy, qualifications of teachers, regulation of teacher training and so on). Hence, the issue is whether there is any additional need for government schools. We can analyse this issue by understanding the limitations of private and not-for-profit schools.
2.1.1 Private, for-profit schools
The general belief that private schools are better is based on a number of reasons. These schools are likely to be more ‘efficient’ because though they have no incentive to spend more than the minimum expenditure required to attract and keep children7, they may have an incentive for a higher level of `customer orientation’. Hence, these are more likely to provide the kind of education desired by parents who send their children there. Similarly, the managers of for-profit private schools may have the incentive to monitor and penalize those employees (mainly teachers) who deviate from what they are expected to do. There are two assumptions here: (a) the work of teachers can be monitored easily by school managers; (b) teachers can be motivated through financial incentives or penalties (including the removal from the job). These assumptions have to be analysed critically.
188.8.131.52 Monitoring the work of teachers
Education as an output is not completely tangible, that is, only some aspects of it are intangible. As such, we may consider test scores or passing examinations as tangible outputs8, which can be monitored by parents or managers and can be used to evaluate the performance of teachers. On the other hand, such monitoring can be difficult for intangible outputs of education, for example, the inculcation of creative thinking or decision-making that makes a student socially responsible. Since it is not possible for parents and school managers to measure the non-tangible aspects of education, this may work against ensuring the effectiveness of teachers in this regard.
We expect the for-profit schools to come under governmental or societal regulation to see that the education provided by them conforms to certain frameworks and standards. However, this requires the state to monitor the behaviour of private schools (or their teachers). This monitoring will involve a cost especially when such schools or teachers have the incentive to violate the established frameworks and standards. For example, the practice of teachers to compel students to take private tuition on payment9. Monitoring and avoiding such practices could be difficult for the government.
184.108.40.206 Motivating teachers through financial incentives
It is an accepted fact that people are motivated by incentives – financial (or material) and non-financial (such as awards or social recognition). There is also intrinsic motivation like doing the right thing, in some people. It is believed that teachers need to have high levels of intrinsic motivation in order to be effective.
There is a common perception that financial and non-financial incentives work in the same manner, that is, the motivation they provide is the same. It is also believed that a person interested in doing a better job due to intrinsic motivation can be further motivated by giving financial incentives. However, incentives and intrinsic motivation do not work towards the desired result in the same way in every situation. An interesting case came from a systematic study in Israel (Gneezy et al, 2011). A small section of parents of children in a playschool had been coming late to pick their wards after school. This created a problem for the caretakers since they had to stay back even if one child was left behind. The playschool instituted a fine to address the problem. Parents who came late were required to pay a fine. Surprisingly, many more parents started to come late. Even those who were coming on time earlier because they felt it was not right to hold the caretakers back beyond their normal working hours, began to come late. This was because when the fine was introduced, they thought that the extra time of the caretakers could be bought without any ill will. This is a case where the introduction of a financial/market incentive led to the disappearance of intrinsic motivation10.
In the case of schooling too, financial incentives working against the expected work is not an uncommon phenomenon in India. Let us consider this hypothetical case: A government allows teachers in schools to charge a fee to provide extra attention to those children who need such assistance. One would not be surprised to see some teachers compelling even those children who may not need extra-assistance to pay the extra fee (probably by not delivering their expected task completely). Hence, we need to be careful in instituting financial incentives as a way to improve the performance of school teachers.
There are several challenges in depending on the for-profit, private schools to improve schooling. Like if there is only one private school in a given locality, it may act in a monopolistic way by charging a higher fee (more than the cost) and providing a lesser quantum of service (compared to the demand of the society). This may mean a reduction in the enrolment of the school and children from poorer backgrounds may be able to access the school only if the government provides financial support. However, this requires contracting between the state and the private school. Such contracting may encounter a number of problems11. How to ensure that the private schools which get grants from the government meet the needs of the state and what should be the nature of funding, are complex issues which may have an impact on the effectiveness and efficiency of the service (of providing education). Secondly, though these schools have an incentive to minimize the costs, it need not translate into a reduced expenditure for the state which funds these schools. There are a number of incentive problems12 here:
- If the money is provided per student, a school may admit all students but may not provide quality education. This happens when schools appoint teachers who are willing to work at lower salaries.
- If teachers’ salaries are directly given by the state, profit-oriented private managers may appoint teachers who pay higher bribes to them. This practice is widely prevalent in Kerala where the government directly pays the salary for teachers in aided private schools (Mathew, 1989). It is difficult for public authorities to monitor and control such private transactions between people who both have an interest in such transactions even if these are illegal.
- If the financial support is given on the basis of the number of students a school admits, there can be an over-reporting of student enrolling and attending schools. This may happen also if the state provides the salary of teachers based on the number of students. This will require continuous monitoring of the attendance in all schools so that children from other schools are not `hired’ to demonstrate an exaggerated attendance on a given day.
- If the government provides a lump-sum grant to a school, it may cut costs by reducing enrolments because there is no incentive to enrol all eligible children from the locality.
- If extra money is given to parents based on the attendance of their children in school, it may require supervision so that there is no manipulation of attendance records in the schools through the collusion between parents and school authorities. This is especially so when parents do not have a great interest in sending their kids to school regularly. However, this scheme will not address all incentive problems of the school discussed earlier.
- If the funding is based on the performance of students, the school may have an incentive to deny admission to those students who do not do well in examinations.
Though private schools may be efficient in terms of expenditure, they may give lower salaries to the teachers. The higher salaries in government schools may encourage many teachers to join private schools or join as contract teachers in government schools while aspiring for permanent appointments in government schools or any other better-paying jobs. Consequently, a major part of their time may be taken up by preparations for this and it reduces the time available to them to enhance their capacities for a long-run, effective teaching career. For all these reasons, the strategy of using for-profit, private schools (with or with financial support from the government) to improve schooling may not be suitable.
2.1.2 Not-for-Profit Schools
There are non-governmental or not-for-profit organizations (NGO/NPO) which own and manage schools with or without governmental support. For such organizations, since increasing their profit may not be the objective, some public concerns which are neglected in the for-profit schools may be better addressed. However, providing a service like education may not be their prime objective or it may be a by-product of the work they do. Consequently, the social priorities in terms of education need not be fully taken into account by them. Also, the content or the priorities in the provision of this service may be driven by the `ideals’ of the organization. For example, there can be more `religious content’ in the education provided by a religious NGO than what may be desired by the larger society.
Even those NPOs which focus exclusively on education and do not function along religious lines may have their own ideas on ‘desirable education’. For example, a number of schools opened in independent India following the ideals of Mahatma Gandhi. There are a few well-known alternative schools functioning in India which follow J. Krishnamurti’s ideas on education. Even if there may be nothing explicitly wrong with these ideas from a social point of view, there could be sections of the society which may object to such ideals. For example, the basic schooling experiment promoted by the Gandhian schools was not attractive to many parents. Parental ideas or norms about what constitutes desirable education are likely to be influenced by their socio-economic status. For example, those who send their children to alternative schools like the ones run by the Krishnamurti Foundation of India (KFI) come from specific socio-economic, intellectual or cultural backgrounds13. What is important for us is to note that the ideals of secular NPOs can also limit the reach of the educational institutes run by them, even if they have no intention to exclude any section of the society.
There could be other reasons which limit the reach of schooling provided by the NPO/NGOs. There may be geographical areas where such organisations are not willing to work. This may be driven by the debilitating environment in a specific social context. There may also be areas where these organizations cannot work even if they wanted to. For example, the social situation could be adversarial to the functioning of a church-run school in many parts of India. These are the reasons that there are lakhs of villages in India where there are no schools run by NPO/NGOs.
Coming to their financial capacity, schools run by NPO/NGOs have three generic sources of funding of schools: (a) cost recovery from parents; (b) other sources of charitable funding; and (c) the government. If the funding is provided by the parents, the school can cater to the needs of only those who can afford to pay. That itself may limit the scope of these organizations in meeting the societal needs of schooling, which is that children from all backgrounds should get schooling irrespective of their ability to pay. A few of these organizations may run schools by following a policy of cross-subsidy whereby wealthier parents pay a higher fee so that the cost of education of the children from poorer backgrounds is subsidised. However, such cross-subsidisation policy among the children who use a single school may not generate substantial resources to support all the children who need such financial support.
The other source of funding is the charitable contributions from individuals and organisations. This source of funding can also be linked to the ideals of the organisation, and then the argument about ideals limiting the reach and access to schooling discussed earlier, come into play. If the government funds these schools, as in the case of the aided schools in Kerala, there will be the problems of monitoring and regulating the functioning that have been discussed. So, whatever their source of funding, there is the issue of transparency. A school’s claim that it is not-for-profit needs to be verified. Most private schools (or private educational institutes in general) in India claim to be not-for-profit, but one can identify practices which are aimed at increasing `profit’ for the owners. In order to ensure that an educational institute is really not-for-profit, there may be a need for somewhat intrusive regulations or monitoring on the part of the government. For example, there may be a need to know actual costs and to regulate fees and other payments collected from the students. Such intrusive regulations may not be liked by the NPO/NGOs that run these schools. However, the absence of regulation can be costly, especially if the government funds schools with certain social objectives (for example, to ensure that children from poorer backgrounds get education in such schools). Therefore, the incentive problems that governments encounter while contracting with the for-profit private schools (discussed earlier) may be encountered in an even worse form in schools claiming to be not-for-profit but which are actually for profit14.
The NPOs may start sc hools with certain broader social objectives, but there can be a gradual decline in their mission in the long run and the schools may begin to function like other enterprises that run for their own sake. It may also bring into play perverse incentives which may work against the realisation of social objectives. For example, a number of aided schools in Kerala started by religious and caste associations have started appointing teachers by taking bribes from them since teachers get salaries from the government. The management may also be inclined to appoint teachers belonging to a specific caste or religion flouting regulations to base the selection on merit or the reservation policy.
If the government funds the major part of the expenditure of schools run by NPOs, there is no reason why they should remain non-governmental schools. It may lead to a situation where the government cannot reduce funding on the one hand and cannot regulate (the appointment of teachers, salaries, and so on) on the other. Hence the provision of education by NPO/NGOs has to be seen as complementary to what is to be done by the state, and not as a substitute for the governmental provision.
2.2 Why is there more privatization of schooling in India than in the US or Europe which are known market-economies?
The percentage of children studying in private schools in India is significantly higher than that in many developed countries including those in the US and Europe. Schooling in India is already highly privatized (Kumar, 2014) with around 30% of schools in the country in the private sector. This trend is likely to continue in the future. How do we explain this situation?
There will be no surprise in this regard if we look at the other services in India for which governments in different parts of the developed world play a major role. The use of private health-care in India is much higher compared to the European countries. Even the poor in India, who are expected to be the beneficiaries of the government provided health-care system, depend on various kinds of private health and medical-care providers. In that sense, privatization in health-care is also more `advanced’ in India. In summary, for goods and services which ideally require different kinds of state intervention, people may even depend on poor-quality (and higher cost), unregulated private providers when they are at lower levels of development. This is part of what can be called an inferior private provision.
The reasons we don’t see enough societal action to create an effective public provision in services such as schooling in India are varied. The protracted underdevelopment of some sections of the society is one. The lack of demand from some people as reflected in their unwillingness to make effort to create and sustain public systems that provide quality service could be another. A collective effort is important even for exercising the voice option which is needed to make public provision more accountable and more responsive to the needs of people. The social fragmentation that makes an across-the-board collective action in society difficult and adds a cost, could be yet another reason. We need to understand the relevance of these different factors in the context of schooling in India, and that may require more research and reflection.
2.3 Why are learning achievements higher in private schools in India?
There are studies which note that the learning achievements of students (as evident from test scores) in private schools are better than those of students in the government schools (Tooley and Dixon, 2005). The annual surveys on learning achievements collected and compiled as Annual Status of Education Reports (ASER) too note that achievements in private schools are significantly higher than those in government schools. However, another study (Muralidharan and Sundararaman, 2013; Karopady, 2014) – a randomised control trial – conducted in Andhra Pradesh could not see a significant improvement in the scores of children who were randomly shifted to private schools (as part of the intervention). On the other hand, students who were in private schools originally (at the start of the experiment) scored significantly higher than the students in government schools. How do we interpret these contradictory results?
International studies have established that educational achievements are influenced not only by what is happening inside the school but also in the households, besides other social factors. For example, Hanushek and Luque (2003) have noted that many factors outside the control of school are responsible for determining academic achievement. The influence of out-of-school factors on learning achievements is confirmed by recent reviews too (for example, Lee and Shute, 2010). Parental expectations and aspirations too matter in this regard (Fan and Chen, 2001).
The studies conducted in India too point to the influence of social and economic factors which are of a tangible type. The Cycle 3 of the National Achievement Survey (NAS) has shown that students from minority groups, comprising Scheduled Castes, Scheduled Tribes, and Other Backward Categories, scored significantly lower than students in the General category (as quoted in Santhakumar et al, 2016). The family size was found to be a factor that had a significant impact – children coming from larger families did more poorly than the average. Parental education was found to be important in the mid-term assessment (MAS) of the NAS. The children of college-educated parents did significantly better than those who had illiterate parents. Santhakumar et al (2016) used the control data from the RCT study in AP. It showed that these social and economic factors have an influence on student achievement – gender, being a scheduled caste or a scheduled tribe (SC/ST), the economic status of the household, parental education and paid tuition. Children belonging to the SC/ST households performed marginally poorly than children from the general category or the OBC households. The economic status of the family was measured by aspects and assets such as having access to water, a toilet in the house, type of the house, number of rooms, possessions like TV and vehicle for transport. These were then converted to an index. A higher ‘affluence’ index had a slightly positive and significant effect on achievement. The impact of such socio-economic factors was assessed after controlling for three school-level variables: school infrastructure (which is measured through an index based on facilities available in the school including electricity, water, toilets, covered classrooms etc.); student-teacher ratio; and teacher qualification.
The role of socio-economic factors evident from these studies explains the higher learning achievements noted in private schools in India. The private schools are used by parents who are not only better off in terms of their socio-economic status but also more concerned about the educational achievement of their children. A similar socio-economic background could be the reason that the children who were moved to the private school from the government school and the children who remained in the government school, showed no significant difference in their learning levels. However, the children who were already enrolled in the private school did significantly better. Presumably, this set of children had parents who displayed a greater demand for education than those who went to the government school. The difference in the learning levels between the children who originally went to the private school, and those who were randomly moved there through a lottery, can be taken as the effect of the demand for education by the parents (Santhakumar et al, 2016).
3. A Case for Government Schools in India
Considering the factors that come in the way of effective schooling in private as well as not-for-profit schools in India, it is clear that there are some issues which not only warrant government intervention but also support the case for government schools.
There are settlements in India with only 50-100 households living within a radius of say, two kilometres. The families there may not be too keen on educating their kids and they would be even more unwilling if they have to send their children to a distant school. So, a private school may not be interested in starting operations there, even if funded partly by the state. In fact, that is the situation in many villages of India. Even with the spread of poor quality private schools in many parts, there are lakhs of villages where there is no private school within a reasonable distance. There is a need for government schools in such locations.
One may argue that instead of setting up government schools, the state can support (for-profit or not-for-profit) private schools directly or through school vouchers. However, the discussions earlier substantiate the fact that having an appropriate contract with them, monitoring and ensuring that they deliver what is agreed upon, can be very difficult and financially, unviable.
It is therefore clear that certain features of schooling in India may necessitate a system that is dominated by government schools. These schools may also encounter issues of inefficiency and ineffectiveness, which brings us to the issue of making government schools accountable and effective.
3.1 Making government schools effective
We have discussed the significant role of intrinsic motivation in making teachers do their jobs well. In addition to it, there may a need for organizational control/supervision, and also vigilance on the part of the parents to see that government schools remain effective. Organizational control requires multiple layers of supervisors to ensure that the system performs its role adequately. There should be institutional penalties on those who do not do their job well. Teachers who do not come to school, do not take their classes dutifully, and do not complete their expected tasks, may have to be penalized. In order to do so, higher authorities should have the incentive to do their jobs well. The effort required to see that this multi-layer hierarchical system works may be substantial.
Also, internal hierarchy by itself is not adequate. Like in the case of other public organizations, the schools too have to be made accountable through the `voice’ and ‘exit’ options of the users. Parents are required to use their ‘voice’ to participate15 or use their ‘exit’ option to vote the elected representatives out in order to improve the performance of the government schools16. This may require for the parents to not only send their kids to schools but to also ensure that they learn; know what they want from the school to improve the education of their kids, and; come together to solve collective-action (or free-riding) problems. These may depend on a number of socio-economic factors and levels of social/human development.
If we consider this in the context of the current situation in India, a substantial section of the society is yet to express full demand for schooling, which means not just enrolment but also learning (Santhakumar et al, 2016). Given the social fragmentation wherein some sections do not value schooling and others value it highly, there can be collective-action problems in jointly exercising voice options to improve schools. Though we expect the community to have homogenous features and a strong interest in improving schooling and have made structures for community participation like school management committees (SMCs), their real participation depends on their attitudes, awareness and common interests. The ability of the parents to influence the state government to improve the functioning of schools depends on the political development there. As discussed elsewhere (Santhakumar, 2014), the political development in different states in India takes place at a different pace and some areas are yet to reach a stage where the governments are compelled to respond to the needs of people in terms of improving public services.
3.2 Will people ever prefer government schools over private schools?
Will an improvement in the quality of government schools be able to bring back parents/students who are attracted to or are using private schools in India? There is no clear evidence. However, we can see a few favourable signs in this regard. Among the middle-class (and not necessarily the upper middle-class) in urban India who may, or can afford, to send their children to reasonably well private schools, there is some attraction towards the Kendriya Vidyalayas (KVs), which are managed by the Government of India. If they can get admission to these schools, parents may prefer to send their children there. The attraction of the KVs comes from the quality of teachers and lower fees (in addition to its affiliation to the CBSE). There are some popular government schools17 in Kerala which children from middle-class households attend. These examples show that when quality schooling is available in government schools, a significant section of those who otherwise prefer private schooling may be interested in using the former. Those who prefer private schooling are not doing it for the sake of private control of schools, but for what they perceive is provided in these schools. If the schooling provided in the KVs has been able to attract some sections of the middle-class, the quality that can attract more parents may not be that insurmountable to achieve. Will the government schools in different parts of India be in a position to improve their quality of education to this level, is the crucial issue.
V. Santhakumar, Professor, Azim Premji University
Brennan, R. L. (2006). Perspectives on the Evolution and Future of Educational Measurement. Educational Measurement (4th ed). American Council on Education and Praeger Publishers. 1-16.
Brewer, D.J. and McEwan, P.J. (2010) Economics of Education. Amsterdam: Elsevier Publishers.
Fan, X., and Chen, M. (2001). Parental Involvement and Students’ Academic Achievement: A meta-analysis. Educational Psychology Review, 13 (1): 1–22.
Gneezy, U., Meier, S. and Rey-Biel, P. (2011). When and Why Incentives (Don’t) Work to Modify Behaviour. Journal of Economic Perspectives, 25 (4): 191-210.
Hanushek, E. A. and Luque, J. A. (2003). Efficiency and Equity in Schools Around the World. Economics of Education Review. Elsevier, 22(5): 481-502.
Hirschman, A. (1970). Exit, Voice and Loyalty. MA: Harvard University Press.
Jacob, B. and S. Levitt. (2003). Rotten Apples: An Investigation of the Prevalence and Predictors of Teacher Cheating. Quarterly Journal of Economics, 118: 843–877.
Karopady, D. D. (2014) Does School Choice Help Children from Disadvantaged Sections. Economic and Political Weekly, 49 (51): 46 – 53.
Kumar, A. (2014). Privatization of School Education: A Social Welfare Program or Industry for Profit. European Academic Research, 2 (1): 1541-1559.
Lee, J and Shute, V. J. (2010). Personal and Social-Contextual Factors in K-12 Academic Performance: An Integrative Perspective on Student Learning. Educational Psychologist, 45(3): 185-202.
Mathew, G. (1989). Communal Road to Secular Kerala. New Delhi: Concept Publishing.
Muralidharan, K. and Sundararaman, V. (2013). The Aggregate Effect of School Choice: Evidence from a Two-stage Experiment in India. NBER Working Paper No. 19441.
National Curricular Framework. 2005. New Delhi: National Council of Educational Research and Training.
Reeves, E. 2008. The Practice of Contracting in Public-Private Partnerships: Transaction Costs and Relational Contracting in the Irish Schools Sector. Public Administration, 86 (4): 969-986.
Sandel, M. J. (1998). What Money Can’t Buy: The Moral Limits of Markets, The Tanner Lectures on Human Values.
Santhakumar, V. (2014). The Roots of Ill-Governance and Corruption. New Delhi: Sage Publishers.
Santhakumar, V., Gupta, N. and Sripada, R. (2016). Schooling for All: Can We Neglect the Demand? Delhi: Oxford University Press.
Stiglitz, J. E. (1999). Economics of the Public Sector (3rd Ed). New York: W.W. Norton & Company.
Thapan, M. (2006). Life at School: An Ethnographic Study. Delhi: Oxford University Press.
Tooley, J. and Dixon, P. (2005) Private Schools Serving the Poor: A Study from Delhi, India. New Delhi: Centre for Civil Society.