Private school choice among Muslim parents: the public–private school decision in Delhi, India

This article examines Muslim parents’ private school choice, their understanding of public–private schooling and how they navigate the choice between fee-free public schools and fee-charging private schools. This article draws on qualitative data from open-ended, semi-structured interviews with 38 parents from Muslim-majority areas in Delhi, India. The findings show that parents choose private schools for several reasons, such as their proximity, discipline, emphasis on Islamic teachings and values, safety and caring teachers. The analysis suggests that structural and social factors influence and construct parents’ choice of a particular school. The neighbourhood where they reside, their minority status, their socio-economic and demographic profile, and the type of schools that are available to them influence their decision making.


Introduction
An increase in the number of private schools around the world has given many parents the opportunity to exercise choice in the educational market and send their children to their preferred schools. Choice is no longer limited to the privileged, as low-income households have started to participate equally in choosing schools for their children. In the Indian context, the low-fee private sector has become 'phenomenal' (Woodhead et al., 2013), and lower-income groups can increasingly choose low-fee private schools for their children.
In countries in the Southern Hemisphere, two key reasons are often cited for choosing private schools: teacher absenteeism and less teacher effort in the public sector (Muralidharan and Sundararaman, 2013). Empirical evidence in the Indian context shows that a large number of parents have started to send their children to low-fee private schools due to 'failing' state schools (Choudhury, 2019;Singh and Bangay, 2014;Srivastava, 2008). It is argued that parents send their children to private schools because these schools are located in close proximity to their homes (Alderman et al., 2001;Goldring and Phillips, 2008) and due to 'quality' concerns (Day Ashley et al., 2014). However, accessing any type of school is not easy, and in addition to cost, accessibility and affordability, other factors may hinder children from attending school or cause them to drop out.
In several contexts, private schools serve the lower socio-economic group by providing a service that parents feel is worth paying for, in contrast to the cheaper government school option (Härmä, 2013). In her study conducted in Lagos, Nigeria, Härmä (2013: 549) asserted that 'choosing to go private is not an ideological choice in many instances; rather it is an acceptance that to get a reliable service means to purchase it privately'. Joshi (2014) states that sociopolitical conditions and stigma around public schooling make parents choose private schools. James and Woodhead (2014: 75) argue that: The concept of 'school choice' is not straightforward. The capacity of households to make choices for their children is shaped by multiple considerations: school availability; perceived quality; accessibility; and most importantly affordability. Whilst trends in India are often framed as a choice between the government versus private sector, individual households may be faced with a much more complex array of options (especially in urban communities), including one or more government schools as well as several private schools of varying financial and/or practical accessibility. The very notion of choice assumes -to a certain extent -the making of informed decisions between viable alternatives.
In India, the majority of out-of-school children are from lower castes and Muslim minority groups (Government of India, 2009). Historically, low levels of educational attainment among Muslims have been a consequence rather than a cause of their marginalisation and discrimination against them (Jahan, 2016). Regarding education for Muslims, the Sachar Committee report concludes that the Muslim community has lower socio-economic and educational status than the Scheduled Castes. (Scheduled Castes are the Dalit [untouchable] groups, and are disadvantaged people in India. Dalits are outside the Hindu caste system. The term 'scheduled' is used because these groups are included in one of the schedules of the Constitution of India. Traditionally marginalised groups in India are singled out as Scheduled Castes, Scheduled Tribes and Backwards Classes, which also include Muslims.) The Sachar Committee report finds 'that Muslims are at a double disadvantage with low levels of education and low-quality education; their deprivation increases manifold as the level of education rises' (Government of India, 2006: 50). (The Sachar Committee was commissioned by the 'Union Progressive Alliance' government. The committee was asked to report on the socio-economic and educational conditions of the Muslim community in India.) This article focuses on the decision-making process of Muslim parents concerning school choice in Delhi, India. Although this article specifically examines the school choice of Muslim parents, it does not discuss the nuances of religious factors that can influence their choice of schools. We explore the reasons that make parents decide to send their children to low-fee private schools instead of fee-free government schools.

School types in India
In India, schools are primarily categorised into three types: government, private aided and private unaided. Schools operated by the government are usually called government/municipal schools and offer formal schooling. These schools are fully financed by the state government, and they are mostly attended by the poor (Härmä, 2011;Vasavi, 2003). Private schools can generally be categorised into two types: private aided and private unaided. Private-aided schools are privately managed but state subsidised. Of all the schools in Delhi, only 5 per cent are private aided, and these schools cater primarily to high-and lower-middle-class clientele (Ohara, 2013). Private-unaided schools mainly cater to poor households and claim to use 'English-medium' instruction and charge 'low tuition fees'. Private-unaided schools are also categorised as recognised or unrecognised. These schools are the focus of this study. In this article, the terms 'public schools' and 'government schools' are used interchangeably.

Analytical framework
One of the most important choices that parents face is whether to send their children to a public school or a private school (Chakrabarti and Roy, 2010). Choice is an amorphous term to describe a situation in which parents can choose the school for their children to attend regardless of where they live. It is assumed that parents have full information on which to base their choice of school, and the option to send their children to the school of their choice (Oplatka, 2004). Research on parental choice and market forces argues that increased parental choice likely increases school competition; as a result, schools improve (Chubb and Moe, 1988) in this context of a tide that 'lifts all boats' (Hoxby, 2003). Although increased competition and higher educational quality are positively correlated, the effects of competition on educational outcomes appear to be modest (Belfield and Levin, 2002).
Rational choice theory suggests that 'parents are utility maximisers who make decisions from clear value preferences based on calculations of the costs, benefits, and probabilities of success of various options' (Bosetti, 2004: 388). However, a criticism of rational choice theory is that parents' decision making is much more complex than rational calculations of the economic return on their investment in particular education markets (Hatcher, 1998). Furthermore, the theory fails to consider the sociocultural nature of choice (Cucchiara and Horvat, 2017).
Highly educated parents have more choice possibilities and tend to be better informed than less educated parents. The latter manage less or poor-quality information and are confronted with more choice restrictions due to price, geographical and cultural barriers (Bonal et al., 2017). Gewirtz et al. (1995) define school choosers as 'skilled choosers', 'semiskilled choosers' and 'disconnected choosers' based on their class origin. Bosetti (2004) asks whether parents with low socio-economic status have the predisposition or the will to engage in educational markets, whether they lack awareness or understanding of choice options, and whether they have the social and cultural capital to navigate the options. The priorities and possibilities of choice are significantly different for middle-class and working-class parents. The choice of school is embedded in a complex pattern of family demands and structural limitations. This is not a matter of cultural deficit, but of pragmatic accommodation . In fact, groups from different social classes engage with and understand schools in different manners .
A rationale for a school choice that is valid for a particular social group and/or social context may not be as valid for another group-context configuration (Bonal et al., 2017). Bounded rationality (see Ben-Porath, 2009;Simon, 2000) considers opposing positions in the school choice debate, and a way to build bridges between them (Bonal et al., 2017). Like comprehensive rationality, bounded rationality assumes that the actors are goal-oriented, but it considers the cognitive limitations of decision makers in attempting to achieve those goals (Jones, 1999). Ben-Porath (2009: 532) notes that 'empirical research on choice exemplifies how a variety of elements, not commonly taken into account by liberal theories of autonomy and freedom, shape the way individuals make decisions and choices.' She provides examples from an ethnographic study and shows how gender, race and class play a decisive role in the accessibility of choice, and how risk aversion conditions low-income students to remain in underperforming schools, even when the parents have the opportunity to change schools. The bounded rationality approach also challenges the assumption of school choice as a denial of deprived groups. Bounded rationality acknowledges some forms of hierarchy in the conditions of choice that different parents face depending on their class status, but this does not make the parents necessarily 'disconnected' or detached from the process of choice (Bonal et al., 2017). A choice is clearly bound by the context in which it occurs, and it is limited by the forms of rationality that the individual can utilise (Ben-Porath, 2009). A bounded rationality explanation of the barriers to utilising parental choice opportunities may be related to reference dependence and loss aversion, which are described by behavioural economists. The limitations and challenges of parents and families in making choices must also be considered (Ben-Porath, 2009).
Scholars suggest that parents' schools of choice are actually influenced by their socio-economic and demographic characteristics (Chakrabarti and Roy, 2010;Elacqua et al., 2006;Gurney, 2017). Social characteristics such as caste and religion are also associated with school choice in India (Choudhury, 2019). In her study on parental choice in Delhi, Gurney (2017) argues that public debates often ignore the dynamics between social identity and school selection, and she illuminates how parental educational biographies and self-identities may constrain parents' ability to act as consumers in education market spaces. Sarangapani and Winch (2010: 505) state: 'In India, not only economic status, but equally if not more importantly, sociocultural dimensions including religion, language, gender and regional identities and affiliations mediate interactions of families with public institutions and markets.' The Indian history of Muslims reveals how they have been visibly invisible in the democratisation processes, and prone to significant decline in socio-economic and political terms (Jahan, 2016). Bhat et al. (2011) note that low educational attainment and marginalisation are not based on religious compulsion or Islamic teachings, but on a lack of social awareness and political consciousness (as cited in Jahan, 2016). Levin (1987: 629) argues that when 'families have different political, social and religious beliefs and values, a basic incompatibility may exist between their private concerns and the public functions of schooling'.
We use the bounded rationality approach to understand the school choice decisions of parents, and to contribute to an understanding of the challenges of low-income Muslim parents in choosing schools for their children due to their social positioning in Indian society. As Jones (1999) explains, instead of making assumptions about decision making and mathematically modelling the implications for aggregate behaviour, bounded rationality adopts an explicit behavioural stance. With this approach, this study provides insights into the school choice and preferences of parents of a particular social class, and the method by which these parents look for schools for their children.

Parental choice in the Indian context
Findings in the Indian context suggest that parents are moving away from poor-quality government schools (Choudhury, 2019;Gurney, 2018;Srivastava, 2008) and accessing private schools. Mehrotra and Panchamukhi's (2006: 438) findings from eight states, namely, Rajasthan, Madhya Pradesh, Uttar Pradesh, Bihar, West Bengal, Assam, Andhra Pradesh and Tamil Nadu, show that 'whether it [is] the quality of school buildings, one-classroom schools, drinking water or toilets for staff and students (especially girls), government schools come out looking worse than private unaided ones'. Poor schooling infrastructure, low teacher motivation and teachers' lack of interest in teaching have led residents in Delhi to lose faith, driven children to drop out (Chugh, 2005) and may have pushed families towards low-fee private schools Kusakabe, 2017, 2019). The social composition of government schools is often confined to low castes and minority groups, as shown by evidence from Uttar Pradesh (Härmä, 2011;Srivastava, 2007) and Rajasthan (Jones, 2017).
In several studies conducted in India, the findings show that parents send their children to private schools because these schools are located close to their homes (Chugh, 2005;Mousumi and Kusakabe, 2019), and have dedicated teachers (Härmä, 2011;Srivastava, 2008), better educational facilities (Mehrotra and Panchamukhi, 2006), smaller class sizes (Muralidharan and Sundararaman, 2013), higher attendance (Kremer and Muralidharan, 2008), better learning outcomes (Kremer and Muralidharan, 2008) and English-medium instruction (Day Ashley et al., 2014;Mehrotra and Panchamukhi, 2006). Parents consider the quality of private education to be better than that of government schools (Day Ashley et al., 2014;Mehrotra and Panchamukhi, 2006). However, critics of private schools, particularly low-fee private schools, extensively argue against and voice suspicions about their quality (Nambissan, 2013;Sarangapani, 2009;Sarangapani and Winch, 2010).
The choices of parents, particularly Muslim parents, do not differ significantly; however, discrimination and schools' emphasis on religious teachings (Mousumi and Kusakabe, 2019) often play a greater role in the choice of a specific school type. In India, Muslim parents' experiences in accessing schools are known to be influenced by their religious status (Sarangapani and Winch, 2010). Tsujita's (2013) findings show that structural obstacles prevent Muslim children from accessing schools, and they are most disadvantaged in terms of receiving education. Borooah and Iyer (2005) assert that Muslims might not use the formal education sector as intensively as other communities in India due to a lack of Urdu-language teaching. Gurney's (2017) findings from Delhi show that Muslim children are discriminated against in government schools because of their religion. Scheduled Castes, Scheduled Tribes, and Muslim minority children are routinely discriminated against in seating, eating, leadership and participation in their schools. They are also segregated, discouraged and humiliated in schools (Namala, 2018).
A plethora of research has been conducted on parents' school choices in the Indian context; however, previous studies have not clearly shown how low-income Muslim parents decide to choose private schools for their children. Building on prior research on school choices in Indian society, this study focuses on the factors associated with choosing schools. Because such research is currently not available in the Indian context, in this article, we discuss private school choice and how these parents navigate their choice between fee-free public schools and fee-charging private schools.

Method
This study was conducted in Delhi, India, and used a qualitative method. The study involved 38 Muslim parents whose children attended low-fee private schools. The selection criteria were Muslim parents from a lower socio-economic background who were sending their children to low-fee private schools located close to the families' homes. Knowledge about this group of Muslim families in relation to school choice is limited, and this group is at the lower end of the spectrum in terms of educational attainment; hence, we discuss their school choice.
An interview guide was used for open-ended, semi-structured interviews with 38 parents from 13 Muslim-majority areas. In this case, 22 fathers and 16 mothers were included in the study (see Table 1). However, it was anticipated that fathers would be more likely to be interviewed due to the structure of Muslim society, in which decisions are mostly made by male family members. This small study only considered Muslim parents who sent their children to low-fee private schools (the children's age range was 6-10 years, and they were attending primary schools). However, greater nuance could have been achieved if parents who sent their children to government schools or considered changing from government schools to private schools were included. The parents were recruited via a purposive sampling technique. After obtaining institutional approval to conduct the research, we visited the field for data collection. The parents included in this study resided in slums and informal settlements (that is, unauthorised colonies where government facilities are not available; these areas mainly have private schools that only offer lower grades), which are low-income areas by definition.
Initially, the first author and research assistant visited 13 Muslim neighbourhoods (those with a high proportion of Muslims residing there), approached people on pavements and in front of shops, explained who we were and the study and asked whether they would be willing to participate. Only those who agreed to participate in the research were interviewed. We also assured the participants that we would protect their identities, and that the findings would be used for research purposes only. On receipt of consent, we conducted the interviews. All interviews took place at the interviewees' homes (as all of the interviewees chose), and each interview was estimated to last 40-50 minutes. We employed the snowball method (for example, we asked the parents to provide the names and contact information of other parents eligible for inclusion). This sampling method gradually allowed us to recruit parents from different parts of Delhi. Moreover, finding specifically Muslim parents would have been time-consuming and unrealistic; therefore, this method helped us to enrol parents of similar types, which further helped us to include many areas with a high concentration of Muslims. Parents whose children were attending private schools were considered for the study.
As we interviewed parents, we probed for further information and clarified the responses. The majority of the interviews were conducted in the Hindi and Urdu languages, with the help of a research assistant. The research assistant also helped with data translation. Field notes were taken throughout the interviews to capture parents' expressions and anything that could contribute to the research.
The interviews focused on how the participants chose schools, and if social, economic and religious identity, neighbourhood, location, and sociocultural dimensions might govern their ultimate decision. The first part of the interview guide included questions on the participants' demographic information, such as age, gender, educational qualifications, occupation, number of children and school type. The second part contained questions concerning the reasons for their choice of the schools that their children were attending, sources of information used in making the choice, number of schools visited before making the choice, satisfaction with the present school, reasons for not choosing a government school, and whether their religious beliefs influenced their decisions about the type of school to which they would want to send their children. From the above specific questions, this article mainly focuses on the reasons for the current school choice, parents' decision making, their satisfaction with the present school and their reasons for not choosing a government school (the other questions are addressed in another article). Whenever needed, probing questions were asked to understand whether their religious beliefs and values influenced their choice processes. The interview responses were systematically organised. The collected data were analysed following a thematic data analysis procedure regarding parents' private school choice and their decisions between public and private schools. The codes were generated through an inductive process and the links between the themes and subthemes were analysed. Most literature on the Indian context does not focus specifically on Muslim parents' school choice, and a significant gap within the school choice literature has been identified with regard to Muslim parents. Instead of providing generalisable data, the sample of 38 parents gave us an in-depth understanding of the nuances of their public/private school choices.

Results
This section includes a description of Muslim parents' private school choices, their reasons for making these choices and how they navigate decisions between public and private schools. This study examines specific factors that influence parents' private school choices. The parents in this study sent their children to private-unaided schools that charge low fees (Rs. 300-1,000). All schools that they chose were Muslim-managed institutions. They usually conducted observations of public and private schools, and evaluated many options before choosing a school. Parents preferred schools that offered all grades. However, they chose the private schools that were available to them, charged low fees, and satisfied their expectations and preferences. The government schools in the areas were poorly maintained and provided only primary sections. The findings show that parents consider school proximity, school environment, school quality, Islamic environment and important non-educational characteristics when choosing schools for their children.

Characteristics of parents
Information about the parents is important, and it informs our attempt to understand their school choices. Table 1 presents information on the parents who participated in the study, their educational level and their occupation.

School proximity
Many parents state that their most important reason for choosing private school is the distance from home. Alderman et al. (2001: 305) assert that 'parents treat the travel time of their children exactly how they treat out-of-pocket fees.' Commuting to schools that are further from home is difficult for parents and children due to safety and security issues, even though government schools are located within 1 km of their homes. As the National University of Educational Planning and Administration (NUEPA, 2014: iii) report notes, '98% of habitations have a primary school (class I-V) within 1 kilometre and 92% have an upper primary school (class VI-VIII) within [a] 3-kilometre walking distance'. However, in most cases, children's ages and busy roads (either congested with traffic or containing railway lines) make it difficult for them to travel by themselves to government schools, rendering these schools virtually inaccessible. On this issue, parents note: 'I am sending my children to this school because of the distance; it is just next to my home'; 'The school my children go to is in this area [near the parent's home]. It is not possible to go to schools that are far away from home; it is not convenient for us'; 'It is not possible to send the children to far-flung schools. Look at their age; they cannot go by themselves'; and 'This school is near my home. Many children are studying in that school, and I have also heard good things about the school from my colleagues'.
The narratives suggest that parents prefer schools that are close to their homes. The parents consider school proximity crucial, as it not only ensures their children's safety, but it is also convenient. If parents select schools that are further away from their neighbourhoods, and must pay school fees in addition to transport fees, then the choice is both expensive and inconvenient for them. In addition to convenience, parents place importance on their networks, and they send their children to schools based on their trust in the schools and on recommendations from acquaintances. However, the close proximity of schools is not the only factor; other factors are also considered in selecting a school.

Parents' perceptions of the school environment of government schools and low-fee private schools
The school environment is important for parents, as it can either exacerbate children's conduct or provide discipline. Parents perceive that the government school environment does not meet their desired standards, as they observe children attending government schools to be disobedient and not well mannered. They express dissatisfaction with the schools, saying that they are unhygienic, and that the teachers are rude, careless and indifferent, are not concerned with teaching children, and do not show care for, or interest in, disciplining children. Parents perceive that government schools do not have a child-friendly learning environment. Parents gather this information from their own experiences of attending these schools, and from children in their neighbourhood that attend government schools: 'Children play a lot and don't study at all, and the teachers hardly teach, and they are not even punctual'; 'If the teachers scold the children, they [the children] hit the teachers'; and 'The teachers at these schools are unfriendly and rude'.
Several parents reported that the teachers show negligence in their duties and that the schools are dirty. Parents have observed disciplinary issues among the children from their neighbourhood who attend government schools. For instance, they say that many children who attend government schools insult their teachers, are disobedient and hardly care about school. Parents have also seen children taking illegal drugs outside the school premises. On this topic, parents note: 'These schools are not clean; [they are] unhygienic. Teachers are not polite to the students and the parents, children are always engaged in using slang words and are fighting with each other, and there is no safety in these schools'; and 'Boys only show up for attendance and take drugs in the schools'.
However, most parents in this study state that government schools' teaching and non-teaching activities are currently improving. They feel strongly that the new government in power is strict with schools and carefully monitors school activities. They also state that government officials often make official visits to schools; as a result, these schools are careful and attempt to provide a better learning environment. In this regard, parents note: 'Earlier, these schools were very bad, but now, they are improving. Teachers are also paying attention to the children'; and 'Government schools are improving. Children who were not studying are studying now, and this is because the new [name deleted for ethical reasons] government is strict with them and constantly monitoring them'.
With regard to private schools, parents deeply appreciate their management, the teachers' behaviour and the well-mannered and obedient students. For these parents, discipline and friendly and dedicated teachers matter most: 'Teachers are friendly and dedicated; they are positive by nature. The management of the school is also good; they have better infrastructure, and the principal of the school is very cooperative'; 'These schools are clean and maintain hygiene, while government schools are hardly clean'; and 'Good environment; the children are clean and use proper words [i.e. do not use slang words]. They are also disciplined'.

Preference for an Islamic environment
Regarding the absence of an Islamic ethos in government schools, some parents state that they want their children to adhere to Islamic norms and values, and to acquire knowledge related to Islamic core beliefs and practices. Nevertheless, parents do not want to send their children to a madrasa (Islamic religious school). These findings suggest that these socially and economically disadvantaged groups want to send their children to private schools to provide better future employment opportunities and an Islamic environment. Although these parents value a religious atmosphere, they prefer schools that offer religious teaching consistent with the social context, and they desire to see their children overcome their disadvantaged conditions. However, if parents select a madrasa, their children will lag behind their counterparts who attend private or even public schools, and, in terms of livelihood, a madrasa education is likely to provide limited help in securing good jobs in the future. Furthermore, to avoid discrimination and social distancing, parents prefer to send their children to a formal school (in this case, a private school) rather than to a madrasa. Parents perceive that a madrasa education is often not viewed positively, and is regarded as providing low-quality education. Parents fear that a certificate earned from a madrasa would not allow their children to secure a job. The opportunities to pursue higher education are also limited for students who attend a madrasa. Furthermore, madrasa degrees are often not recognised in India, as the religious curriculum does not tap into the skills and competencies required in the present-day job market. Moreover, a madrasa education could have adverse effects because it is often associated with breeding grounds for terrorism. Parents seek to secure their children's future, and to help their children to navigate the discrimination that Muslims are likely to experience more generally in Indian society. However, parents are disappointed by the absence of an Islamic environment in government schools: These [government] schools are not good for our children as they do not have an Islamiyat environment'; 'It [the school] has an Islamiyat environment, and it is a girls' school'.
It [the school] is providing good education; they discipline the children, offering Islamiyat Deeniyat [Islamic teachings]. In government schools, there is an Urdu subject, but there is no teacher to teach this subject. Teachers at government schools are overloaded with work, and that is why they don't teach. Parents do not want to send their children to private schools because of their high fees. However, private schools hold parent-teacher meetings frequently, and this makes the parents happy about these schools.
The school has an Islamic environment, uses English-medium instruction, and [has] good infrastructure, although it does not have a playground.

Parents' perceptions of quality in government schools and low-fee private schools
Parents consider school quality to be an important factor in their private school choices, but they perceive government schools to be gradually improving under the new government's regular monitoring. However, problems persist with regard to quality in both government and private schools, and parents see great disparities in quality between the two types. Some parents have experience attending these schools, have been exposed to word-of-mouth or have watched the schools located in their neighbourhood; the teaching-learning aspect of government schools is often determined by observing neighbourhood children attending these schools. In this article, we define school quality based on school facilities, such as well-established infrastructure, smaller class sizes, low teacher-student ratios, good examination results and teaching activities, improved teaching and learning materials, and other features that schools can offer.
Parents criticise government schools and observe how poorly the schools function in the area. The government schools in the neighbourhood mainly offer lower grades (primary section) and function under abysmal conditions. Parents note that government schools have poor and inadequate infrastructure, and that their teachers are overloaded with administrative work, and thus do not teach. Some of their observations are as follows: 'They [students] don't know anything [they are not learning]; they are being promoted to the next level but not learning anything. Teachers are not worried about their jobs, and that's why the teachers are not teaching'; 'My brother-in-law is studying in a government school in Grade 8, and if you ask him to say A to Z [the alphabet], he won't be able to say it fluently'; 'These schools have poor infrastructure, although [the schools] have a playground'.
However, the parents' narratives suggest that government schools are improving and have therefore become more attractive to parents. Muslim parents regard schools that offer separate shifts for boys and girls highly, because they do not want boys and girls to be able to intermingle freely. It should also be noted that low-income households find it difficult to send all of their children to private schools. Studies show that in India, due to financial constraints, parents prefer to send their sons to private schools and their daughters to government schools (Härmä, 2011;Kumar and Choudhury, 2020;Mehrotra and Panchamukhi, 2006;Vasavi, 2003;Woodhead et al., 2013). Hill et al. (2011) assert that parents may be reserved in their explanations for why their daughters and sons attend different schools, but the role of gender as a determinant of school choice is clearly revealed in their study. Parents state: 'The school is not good, but has nominal fees and provides mid-day meals'; or: Government schools are good for girls because there are separate shifts for boys and girls. However, these schools are not good for boys because the boys are always fighting with each other. The teachers currently are teaching properly because the new government is monitoring them. These schools are also providing scholarships to the students.
In contrast, many parents praise private schools with regard to their quality and say that they are especially satisfied with the teachers at private schools, as they care for the children and teach them certain codes of behaviour. These schools also hold regular parent-teacher meetings, which inform parents of their children's education and behaviour. In terms of school quality, parents connect quality with teachers' credentials, professional training and level of care for the children. Parents do not associate school quality with their children's academic performance, and they do not discuss the teaching-learning processes of the schools. Parents also rarely refer to their children's educational activities in the schools: 'Teachers are teaching and are attached to the students, as they are afraid they may be fired from their jobs.' However, parents also state that private schools are no better than government schools, since they operate with insufficient rules and regulations; teachers are not highly educated, but they teach well (they show special care for the children). The interviews also show that some parents view private schools as focused on doing business, instead of providing education. In addition, these schools provide only lower grades, and parents must eventually transfer their children to schools that offer higher grades. In most cases, parents opt for government schools to send their children to higher grades. Parents strongly dislike the fact that private schools do not have 'qualified' and 'professionally trained' teachers. This raises concerns among parents about their children's education. Parents firmly believe that only qualified and trained teachers are able to teach effectively: 'Schools are running for business purposes only'; 'These are like shops; if they provide good service, they will get more customers. Since parents pay the fees, they listen to the parents. Otherwise, they don't truly care. They only care about the fees. It is not easy to enrol children in private schools'; '[There are] limited physical facilities. There are no qualified teachers; they are untrained. They charge high tuition fees, and do not provide any scholarships'; 'Teachers are not teaching properly. The school environment is good, but teachers do not have sufficient knowledge of the subjects. Teachers are qualified, but they are not trained'.
The data reveal that parents associate quality with teachers' credentials, and care about and regard these factors, rather than specific teaching-learning processes or academic achievement, as indicators of quality. Despite feeling contempt for private schools, parents continue to send their children to these schools because these families are economically disadvantaged, and they do not have the option of schools that provide a nurturing learning environment. Parents are aware of the facts, and they understand the quality of both government and private schools.

Parents' perception of non-educational characteristics in government schools and low-fee private schools
Student population and school cost are grouped under non-educational characteristics. Parents differentiate themselves from others within their own communities and classes; they reject public schools and they want to distinguish themselves from families whose children attend them. Although parents reject government schools based on quality, they appreciate these schools because of the students they accept. Some parents state that government schools are non-discriminatory in terms of students' economic status, and that anyone can send his or her children to these schools. Moreover, being Muslim, the parents also appreciate that government schools are good for girls, who supposedly have certain feminine attributes, such as 'cooperativeness' and 'gentleness', and are thus less likely to fight with other children or have disciplinary issues: 'In the government schools, rich and poor can go together, but this is not possible for the private schools. There is no restriction for any of us with regard to accessing the government school'; 'This school is good for girls but not for boys'; 'Mahol [the school atmosphere] is important. There is not a good atmosphere, as children from all types of families enrol their children here [in the government school], so it is tough to discipline the children'.
At the same time, some parents reject government schools because they view the school atmosphere as unsuitable for their children's disciplinary needs. Parents strongly criticise the intermingling of students from all social groups in government schools. Government schools are often seen as undesirable because they are attended by students of all social classes. In private schools, the parents say that 'the teacher-student relationship is not good; if the students do not follow the instruction of the teachers or have a bad relation [with the teachers], the teachers will fail' the students. However, parents perceive that teacher-student relations are not based on trust and mutual respect in private schools. Teacher-student relations are distant, which often has high costs for children, as it appears that a student who does not follow a teacher's instructions may fail that subject: 'Only the parents who can afford it send their children to the private schools'; 'They [the private school authority] keep asking if there is any problem [they are accountable to the parents]. However, these schools are expensive; they have expensive books, like they charge Rs. 1,000 for a Rs. 300 book'.
Parents complain that government schools ask for documents for admission. This clearly violates the Right of Children to Free and Compulsory Education Act or the Right to Education Act (RTE) 2009, which took effect in 2010. Some key stipulations of the act are year-round admission, no discrimination, no requests for documents, no screening, no capitation fees, no detention and age-appropriate classes. However, the following quotations indicate that parents are still required to provide documents for enrolment. (Some parents do not possess identification documents because they migrated from other states in India.) With the introduction of the Act, schools should no longer be able to engage in such behaviour. Regarding government schools, parents say: It [the government school] is affordable; however, they require documents, they ask for a certificate of residence, but the parents don't want to send their children to that school [the government school near the parents' homes] as the road is not good, there is a big hole in the road [children might fall], and the school does not have a pucca [brick] building; classes are going on in a tent.
The above vignette shows that the increasingly disproportionate fees do not seem reasonable to the parents. Although parents are hesitant about the student population in government schools, they feel that these schools could be improved if the teachers looked after the children with care.

Discussion and conclusion
Parents make their school choices based on the options in the market that are accessible to them in their neighbourhoods. Bonal et al. (2017) assert that the preferences of low-income parents are constructed in a context of social and economic restrictions and that these preferences must be consonant with the market that is available to them. Bonal et al. (2017) further assert that low-income parents are guided by predefined sets of preferences, and establish priorities, reject certain schools and reflect on their final choices during the school choice process. Low-income, urban, mostly minority parents rely on the information available to them and their access to information about schools, which is greatly limited by their location and social positioning (Ben-Porath, 2009). In fact, 'Growth in the number and type of schools has expanded the education market place and changed the institutional context within which households make decisions about schooling choice. However, schooling choice is not always a "free" choice but one mediated by household and social characteristics' (Hill et al., 2011: 98).
Accessing schools, especially for the religious minority, is often not a linear process. The juxtaposition of poverty and a Muslim identity both speak to a complex form of exclusion. Moreover, sociospatial inequalities compound vulnerabilities because of minority identity and, consequently, opportunities for schooling. Spatial inequalities in Delhi have placed minority Muslim communities (and other poor families) at a disadvantage because the settlements where they reside have inadequate access to good-quality schools, regardless of whether they are government or private schools. Jahan (2016) writes that Gayer, in his study of Abul Fazal Enclave (an informal settlement where mainly Muslims reside) in Delhi, highlights the ambiguity of choice in situations where individuals are bound by multiple constraints. She also notes that the 'choiceless' nature of choice compels Muslims when making housing decisions. Indeed, these vulnerabilities and intricacies have forced and reinforced cumulative inequities. In fact, in many states in India, such as Maharashtra, Gujarat and Madhya Pradesh, there is ghettoisation of the Muslim community in terms of residence, education and employment (Jahan, 2016). The choices made by people in this particular community are complicated, and the reasons for exclusion are obscure. We have unpacked several reasons for such choices, especially for the Muslim minority community.
Parents in this study appreciate schools that have caring teachers, and teachers who instil discipline in children. Similar to Bell's findings (2009b), parents appreciate a nurturing environment in which their children will not be lost. Parents prioritise factors such as travel, distance (convenience), school environment, preference for an Islamic environment, cleanliness, discipline, safety, school quality (mostly in terms of teachers' care for, and individual attention to, the children) and non-educational characteristics (student population) when choosing a private school.
Parents favour private schools because they adhere to Islamic norms and values, and offer teaching on Islam-related subjects. Based on the interviews, parents clearly remain hesitant, and have not gained confidence in government schools because they have seen dilapidated government schools in their community for years; thus, the choice for government schools decreases because parents are discouraged. This predicament has also arisen due to this population's enduring deprivation and discrimination for generations, and their experience of systematically being set apart from participation in government services. Research on low-fee private choice indicates that traditionally advantaged groups in society are favoured by the market in education, while children from traditionally disadvantaged caste groups and minority religions, and children from poor and less educated families, are less likely to enrol in private schools (Chudgar and Creed, 2014;Härmä, 2011). Poorer households face economic constraints, an inequality of power within the school, and less ability to acquire the knowledge necessary to maintain a foothold in this new market (Fennell and Malik, 2012). On the other hand, private schools that deliver services at sufficiently low fees to attract poor families may not deliver services of adequate quality. Moreover, private schools that cater to the poor often exploit marginalised groups who are not capable of assessing whether their children are learning (Alderman et al., 2001). However, Tooley (2004) asserts that private schools are the 'poor's best chance'.
The results underpin the importance of both 'heuristics and bounded rationality' (Bell, 2009a: 205) in understanding the school choices of Muslim parents. This study reveals that despite limited resources and residence areas, and perhaps even in a disadvantaged context, parents exercise choice and select schools that satisfy their expectations and preferences. They articulate why they select private school and repudiate government schools. The parents are aware of their restricted choice due to their social class (and perhaps their minority status) and location; however, they adapt their decision to the available schools. They select schools from the choice set that is available to them. Their neighbourhood and social positioning limit their choice options. Structural and social factors influence and construct parents' choice of a particular school. This study finds that the choices of parents are bounded. There are social and structural inequalities (Bell, 2009a) that greatly influence Muslim parents in decision making. The residence neighbourhood, minority status (in addition to discrimination), socio-economic and demographic profile and type of schools that are available influence their decision making. They are not necessarily 'disconnected choosers'; the aforementioned factors impact their construction of choice.
The social characteristics of government schools strongly discourage parents from sending their children to them. Parents determine school quality based on the social composition of the schools. However, the results show conflicting narratives: some parents like social classes to mix together in government schools, while others do not. Parents may also use student demographics as a proxy for the academic quality of the school. Parents may assume (accurately or not) that schools with more advantaged students can attract more motivated students and families (Elacqua et al., 2006). Despite coming from equal social classes, they hold differing perceptions and 'can reach the opposite conclusion in a similar intuitive fashion' (Ben-Porath, 2009: 535). Interestingly, they hold this notion regardless of their education and occupation. Choice is a complex process, and the factors that influence choice may vary in the same social group. Ben-Porath (2009: 535) writes that there can be individual differences and personal processes of decision making, and that 'the reasons for either parent's choice vary, and they could be described through an agent-based or bounded-rationality model as well as through social and institutional construction model'.
The findings suggest that it is unreasonable to think that increased parental choice improves schools. It is often assumed that given choices, parents will choose the best schools for their children because they are 'utility maximisers'. However, the decision-making process is not straightforward, and factors such as political, social, structural and economic barriers must be considered. In this study, we find that the minority identity, socio-economic status and spatial inequalities of parents shape the bounds and appear to have considerable equity implications. With these statuses, Muslims in India become doubly marginalised in terms of accessing better education, which widens social and economic inequalities. The schools that parents select do not produce equity gains, since the results suggest that the selected private schools are not better than the government schools. Policymakers should consider the constraints of the availability of good schools, and introduce mechanisms to reduce structural inequalities and sociospatial segregation.
By investigating Muslim parents' choice of schools, this study contributes to understanding how social positioning and spatiality construct choice. Muslim parents' choice of schools involves introspection, as their right to education is circumscribed. This topic requires further discussion. Future studies with large representative samples should be conducted to examine the effects of sociospatial aspects on school choice options, and to understand how marginalised communities arrive at decisions about their children's schooling. Future studies should also discuss the nuances of the strong preference for the 'Islamic environment', and how this preference may contribute to potential segregation or self-segregation.