Understanding School Segregation
Understanding School Segregation

Xavier Bonal

Xavier Bonal is Professor of Sociology at the Universitat Autònoma de Barcelona (UAB), Spain, and Special Professor in Education and International Development at the University of Amsterdam, the Netherlands. He is Director of the Globalisation, Education and Social Policies (GEPS) research group of the Department of Sociology of the UAB and a founding member of the Interdisciplinary Group on Educational Policies. He is the coordinator of the Erasmus Mundus Joint Master Degree on Education Policies for Global Development (GLOBED). Author affiliation details are correct at time of print publication.

Search for publications
and Cristián Bellei

Cristián Bellei is Associate Researcher at the Center for Advanced Research in Education and Professor in the Sociology Department, both at the University of Chile, Chile. He previously worked at the Chilean Ministry of Education and UNICEF in Chile; he has also been a consultant for UNESCO, the World Bank and the Inter-American Development Bank. He is the author of The Great Experiment: Market and Privatization of Chilean Education (2015). Author affiliation details are correct at time of print publication.

Search for publications

Bloomsbury Academic, 2019


Content Type:

Book chapter



Related Content

The Production of Socio-economic Segregation in Chilean Education: School Choice, Social Class and Market Dynamics


Cristián Bellei

Cristián Bellei is Associate Researcher at the Center for Advanced Research in Education and Professor in the Sociology Department, both at the University of Chile, Chile. He previously worked at the Chilean Ministry of Education and UNICEF in Chile; he has also been a consultant for UNESCO, the World Bank and the Inter-American Development Bank. He is the author of The Great Experiment: Market and Privatization of Chilean Education (2015). Author affiliation details are correct at time of print publication.

Search for publications
, Mariana Contreras

Mariana Contreras

Search for publications
, Manuel Canales and Víctor Orellana

Víctor Orellana

Search for publications

DOI: 10.5040/9781350033542.ch-011
Page Range: 221–240

1. Chile: a highly socially segregated market-oriented educational system

Estimates show that Chile has the lowest level of social inclusion in their schools among OECD countries, which implies a low degree of variation in student socio-economic status within schools (OECD, 2016). Despite the prominent presence of social segregation in Chilean education, this issue has only recently received some attention from both scholars and policymakers. This chapter has two objectives: first, based on a literature review, to provide an overview of the evidence about the causes of socio-economic segregation in Chilean education; and, second, based on qualitative evidence, to explore in depth the role of parents’ school choice in this phenomenon. Although we recognize the complex nature of school segregation, in our analysis we emphasize how market-oriented reforms have fuelled and shaped socio-economic school segregation in Chile.[1]

In fact, since the early 1980s, Chilean educational policies have been recognized as among the most radical cases of market reform in education. The reform (implemented during Pinochet’s dictatorship within a profound neoliberal restructuring of the Chilean economy and society) included almost all the components associated with market policies in education, was applied nationwide and has remained in place for more than three decades. Although the idea of ‘market’ in education is somewhat vague, it fundamentally encompasses family school choice, school competition for family preferences, and the relevant autonomy of schools to differentiate themselves in order to compete (Chubb and Moe, 1990; Bellei, 2009; Verger et al., 2016).

In primary and secondary education, Chilean market reform included the creation of a voucher-like mechanism as the single state funding system for schools. Irrespective of their zone of residence, families can freely choose the school for their children since the voucher is the same for public and private schools – which includes for-profit or not-for-profit private, as well as religious schools. In theory, all schools – in fear of being left without financing – would compete among themselves to enrol students. To facilitate comparison between schools and guide the supply-demand dynamic towards quality indicators, since the late 1980s the government has applied SIMCE, an annual standardized academic achievement test (at least in Language and Mathematics) to all fourth-, eighth- and tenth-grade students, and since the mid-1990s has published school rankings and has sent letters to inform parents of school-level test results. Finally, in order to limit the role of national government in education to functions of regulation, funding and evaluation, during the 1980s the administration of all public schools was transferred from the Ministry of Education to the municipalities. The rationale was that local authorities would manage public schools to compete with private schools in a highly decentralized school market, and that the Ministry of Education should adopt a ‘neutral’ position, treating private and public schools in the same way, thus promoting the emergence of a self-regulated market (Bellei and Vanni, 2015).

Market policies caused a dramatic and sustained privatization in Chilean education. When the system started in 1980, around 15 per cent of students attended private schools; by the end of the 1980s this proportion had more than doubled, and by 2017 at least 60 per cent of Chilean students were educated by private institutions (i.e. public education was proportionally reduced by half in the same period). The growth of private education has been primarily driven by the creation of new for-profit private schools (Elacqua, 2012). Finally, although several democratic governments have implemented a great variety of school improvement and compensatory programmes in Chile since 1990, there has been no policy focused on diminishing the privatization process – quite the opposite, since some more recent policies can be interpreted as a promotion of private schools. For example, democratic governments allowed private schools to receive public funds for investing in infrastructure, increased the incentives to subsidized private schools to charge tuition to families, and increased the value of the voucher for low-income students whose families cannot afford this co-payment (Bellei and Vanni, 2015). Also, the state imposes minimum requirements to open new voucher private schools and voucher schools are not required to accept all applicants, all of which shape a very radical scenario for the educational market in comparative terms (Bellei, 2015).

As mentioned, the research on socio-economic school segregation in Chile is somewhat recent; nevertheless, both international and national studies have produced a sizeable and convergent amount of evidence showing that – in comparative terms – Chile has a highly socially segregated educational system. For example, Valenzuela et al. (2014) estimated a Duncan Dissimilarity Index of 0.54 for the 2008 fourth-grade students pertaining to the lowest 30 per cent of family socio-economic level, and of 0.60 for their peers of the 30 per cent highest socio-economic level. They also showed a slight increase in those Dissimilarity Index estimates during the previous decade. Additionally, PISA has shown systematically that Chile has among the least socio-economically diverse schools of all OECD and non-OECD countries with comparable data (Vázquez, 2016; OECD, 2016).

2. Causes of social segregation of Chilean schools

In this section we analyse the causes that explain the high level of socio-economic school segregation in Chile. Based on a review of recent studies (mostly quantitative) focused on identifying the factors related to school segregation in Chile, we show that, in addition to the socio-economic segregation of Chilean cities, several features and dynamics of the educational system make substantial contributions to produce the social segregation of schools.

Broadly speaking, the high social segregation of Chilean education is mainly explained by the market dynamics in which it operates (Valenzuela et al., 2014).This is not to say that those market dynamics created the social segregation of the schools, nor does it explain it entirely, but they do exacerbate the segregation and shape it. Thus, we shall argue that the market-oriented institutions that govern Chilean education (i.e. universal vouchers, high levels of privatization, and non-regulated school choice) provide a framework to interpret the available evidence on the production of socio-economic segregation in education.

In order to observe the additional contribution of market dynamics to school segregation, it is useful to consider as a baseline the social segregation of the cities, which is highly acknowledged as the main contextual factor in producing school segregation. Since parents tend to educate their children in schools close to their homes (Gallego and Hernando, 2009; Carrasco and San Martín, 2012), it is understandable that the socio-economic distribution of the schools partially reflects the residential segregation, an association that empirical research has demonstrated for Chilean cities (Valenzuela et al., 2014; Arteaga et al., 2014). In fact, studies applying statistical simulations suggest that in a scenario of random assignment of students to schools within their cities, the Duncan Dissimilarity Index of social segregation of schools in the metropolitan Santiago area would be the same as the residential segregation (Gallego and Hernando, 2008), with it increasing in a scenario where each student attended the school nearest his/her home (Elacqua and Santos, 2013). In short, residential segregation explains a part of school segregation. For example, Arteaga et al. (2014) estimate that residential segregation accounts for 8–13 per cent of the school segregation in Santiago. However, it is important to note that the available evidence also shows that social segregation of Chilean schools is significantly higher than residential segregation (Valenzuela et al., 2014).

This last finding is not obvious nor is the direction of the causal relationship between both urban and school segregation: we have to know the institutional rules of the educational system to understand this issue. In a school system in which families were obligated to educate their children in the nearest schools (all of them being free) and in which the families perceived important differences in school quality (in whichever way they defined that), families would attempt to live close to the best schools which would create an increase in home prices and would make urban segregation partially caused by the distribution of educational quality, in turn fostering more school segregation (i.e. a Tiebout model). In this scenario, school choice has been proposed to reduce social segregation: if low-income families were able to commute to send their children to schools, they would be integrated with other social groups. If this proposal were to succeed, social segregation of schools should be lower than residential segregation. Why then in Chile, where there is complete school choice facilitated by universal vouchers, is school segregation so much higher than urban segregation? Now our hypothesis about the Chilean school market demonstrates its potential explanatory power.

To begin with, the distinction of the ‘urban factor’ is somewhat artificial and only makes sense to try to identify the contribution that internal factors from the field of education make to school segregation. The educational market does not operate in a vacuum, but is strongly embedded in the geography in which people live and schools are located. In fact, the conditions that facilitate the school market dynamics (e.g. size and density of population, transportation, family income) vary significantly among socio-geographical zones, which is related to the effects of the 1980s neoliberal educational reform (Hsieh and Urquiola, 2003; Auguste and Valenzuela, 2004); also, recent research has shown a highly dynamic market with an annual ‘destruction rate’ of about 1.1 per cent of the state-funded system between 1994 and 2012 (Grau et al., 2015). The differences in the intensity of the school market dynamics among geographical areas allows for the study of their effects on the socio-economic segregation of schools. Certainly, it is very difficult to disentangle the effects of each of the different aspects in the production of segregation since market dynamics operate systemically.

Firstly, there is evidence that shows that the level of privatization in the educational supply in a given area (i.e. share of private school enrolment, number of private schools) is positively associated with the level of socio-economic segregation of schools in that zone (Valenzuela et al., 2014; Elacqua and Santos, 2013; Roje, 2014); this finding has also been replicated in comparative studies, including Chile (Vazquez, 2016). Certainly, if private schools behaved exactly like public schools, the results might be different, but in the Chilean educational market, that is not the case. In fact, the social distribution of the enrolment was segmented early on: public schools serving low and middle socio-economic levels; subsidized private, middle and middle-high socio-economic levels; and private non-subsidized, high socio-economic level (García-Huidobro and Bellei, 2003). This stratification has been exacerbated to produce a social polarization at both extremes of the distribution (Valenzuela et al., 2013). Therefore, a share of the socio-economic segregation of the schools is explained by the social segmentation among these different types of schools (Vazquez, 2016; Paredes et al., 2013).

Nevertheless, given the growing relevance of private schools in the provision of education in Chile, school segregation is increasingly associated with differences among private schools and the operational differences between private and public schools. In fact, research shows systematically that social segregation is higher in non-subsidized private schools than in subsidized private schools, and that public schools continue to be the most socially integrated (Valenzuela et al., 2014; Elacqua, 2009; Elacqua and Santos, 2013; Roje, 2014; Mineduc, 2012; Paredes et al., 2013).[2]

One of the key differences among private schools, essentially linked to market dynamics, is the existence and the amount of tuition charged to families, which triggers a predictable tendency for price differentiation among schools. Accordingly, research has found that subsidized private schools with co-payment are the most socially segregated voucher schools (Elacqua, 2012; Roje, 2014) and that the higher the proportion of schools charging tuition to families in a city, the higher the school social segregation (Gallego and Hernando, 2008; Valenzuela et al., 2014; Elacqua and Santos, 2013; Roje, 2014; Arteaga et al., 2014). Strictly speaking, evidence suggests that school tuition should be interpreted as a continuum going from elite schools to the low-fee subsidized private schools which makes tuition a variable that affects school social segregation across the price range (Valenzuela et al., 2014; Paredes et al., 2013). The effect of co-payment subsidized schools on school segregation in Chile is so relevant that it has been estimated to be even higher than the effect of residential segregation (Valenzuela et al., 2014; Arteaga et al., 2014).

An additional difference among private schools linked to market dynamics is whether they are for-profit and, more generally, whether they are sensitive to monetary incentives when they make decisions about the socio-economic condition of the students they admit. Evidence shows that for-profit voucher schools are more socially segregated than non-profit schools, a relationship that is also present for both tuition and free voucher schools (Elacqua, 2012). More precisely, evidence suggests that for-profit schools tend to hyper-specialize in ‘market niches’; thus, although they as a group are increasingly heterogeneous in socio-economic terms, each school tends to be highly segregated (Elacqua and Santos, 2013). An indirect indicator showing that economic motives are a relevant factor in school segregation is the fact that as the economic value associated with low-income students increased (as a consequence of the 2008 Preferential Voucher Program), private schools started to expand their presence in this market niche (Roje, 2014). Certainly, Roje also estimates that this process did not contribute to school desegregation, since most low-income students now attending private schools come from public schools that were less segregated, although this issue needs to be studied in the medium term.

One last characteristic that differs among private schools – although also present to a lesser extent in public schools, especially highly prestigious high schools – is the intensity of the selection processes implemented by the schools to the students and their families (in addition to the direct effect of tuition) either during the admission or the schooling processes. Studying the effect of these practices on the social segregation of schools involves additional challenges. Firstly, the information available is not reliable because there is no official data about them (many of these practices are very subtle and informal) and some of them are illegal (subject to certain regulations since 2004 and some direct prohibitions in primary education since 2008). Secondly, school selection practices confound easily with family school choice practices with which they interact even before a family enquires about a space for their child, as a self-selection effect. And finally, family/students selection can produce social segregation effects, without being directly or intentionally oriented by socio-economic status. For example, if a religious school selects students based on a family’s faith, but in turn, the families who practise that religion are biased towards one socio-economic level, school social segregation would be an indirect and unintended effect caused by the religious-based school admission processes.

At least since the beginning of the 1990s, there is evidence that the school admission processes in Chile are partially associated with the socio-economic status of students’ families (Gauri, 1998; Bellei, 2009), practices that have continued on a large scale despite being prohibited by law at the primary school level (Carrasco et al., 2012). A recent study confirmed the aggregate effect of those practices by showing that a proportion of the social segregation of schools is explained by the presence of ‘selective schools’ in a given area (an effect that is independent of the tuition effect), increasing school segregation beyond residential segregation and the geographical distribution of the schools (Elacqua and Santos, 2013). That is to say, the school selection of families/students increases the bias against the population with comparatively lower socio-economic status within the school’s zone of influence, making selective schools even more socially homogeneous (Carrasco et al., 2014; Elacqua and Santos, 2013).

As mentioned, it is extremely complex to distinguish between the segregation effect caused by school admission and the families’ school choice processes. Studies implementing statistical simulations are convergent in showing that this supply–demand interaction is a very relevant additional factor to explain the social segregation of the school in the metropolitan area of Santiago (Gallego and Hernando, 2008; Flores and Carrasco, 2013; Arteaga et al., 2014), although researchers disagree as to whether this should be interpreted as an effect of the bias distribution of the supply or differences in family preferences. Strictly speaking, the market dynamics we observe are the interaction between supply and demand reciprocally shaped over time.

Finally, from the demand side, there is quantitative evidence (although not always direct) that families take into account the social composition of the schools when deciding the school they will apply to for their children, suggesting that families give value to schools where students’ socio-economic levels are similar to or somewhat higher than their own (McEwan, 2001; Gauri, 1998; Elacqua and Fábrega, 2006; Elacqua et al., 2006; Gallego and Hernando, 2009), although Flores and Carrasco (2013) did not find a consistent pattern for this. The notion that families choose schools partly because of the socio-economic status of the student body and other associated characteristics that negatively affect schools (like the perception of unsafe conditions, lack of order and social marginality) has also been a consistent finding of qualitative research (Espínola, 1993; Raczynski and Hernández, 2011; Córdoba, 2012; Canales et al., 2014; Corvalán et al., 2016; Contreras, 2015). Since this line of enquiry has been less developed, we shall expand on it in the next section, presenting recent qualitative evidence on the relationship between school choice and segregation.

3. The contribution of parents’ school choice to school segregation

In this section, we analyse how the school choice of parents from different social classes may be an additional factor to explain school segregation from the demand side.[3] Certainly, we shall relate family preferences, rationale and decisions to the school admission policies and – more generally – the socio-institutional context within which school choice operates in Chile. We organized the analysis by social classes – low, middle and high – distinguishing two sectors within the middle class, the middle high or traditional and the ‘new middle sectors’.[4]

Broadly speaking, because the common family conceptualizes their vision and value of education and school for their children from a social point of view, when choosing schools, parents from all social classes affect school segregation. In general terms, the choice of a school is meant to maintain or attain a certain position in the social structure. However, the rationale of school choice varies among social classes because in practice, it is shaped according to perceptions about the positions that families have within the social structure and their relationship with other social classes as well as the resources they have and the restrictions/opportunities they face.

3.3.1. Lower social class

Living mainly in marginalized areas of the cities, low social class families are fully aware that they occupy a subordinate position in the social structure and that they live in neighbourhoods marked by poverty and the stigma of violence, delinquency and drug dealing. Within this context, for a significant number of families, the meaning of school choice is essentially linked to providing their children with spaces for socialization that avoid the risks of the neighbourhoods in which they live and foster adequate socialization, as Córdova (2012) also documented. The key to their choices is looking for institutions that function in a regular way and are able to reduce the threats associated with growing up in a marginalized environment.

Using those criteria, public or free schools do not appear to be an effective option for this objective. Whenever possible, these parents exclude those types of schools because the children of the poorest families who cannot pay or are not willing to pay attend those schools. Not paying for school would be an indicator that those parents are not concerned about the education of their children and have low expectations of being able to overcome their current situation.

The parents who go to public schools, they aren’t well educated, many of them do not work formally; maybe they want to be able to pay or maybe they can but they won’t because they are not so interested in educating their children in a better way.

 --mother, private subsidized school

Above all, this is because in those free schools there is a concentration of both families who do not uphold moral discipline and quite the opposite, show behaviour associated with alcoholism, drug dealing, delinquency and sexual precocity, or impoverished families who live unbecomingly, disregarding personal appearance or using vulgar language.

You know, here there is a lot of delinquency and drug addiction and public schools get all those kinds of people, children.

 --mother, public school

In this way, these parents draw a clear distinction within their own social stratum based on moral and cultural differences and not economic ones, and attempt to prevent their children from socializing with children of the families who show behaviours and values that they consider reproachable. According to our evidence, most parents from the low social class fear that free and non-selective public schools may produce a mix that ultimately creates some kind of contagion of those marginal and marginalizing signs that in the future will obstruct normal integration into the social structure.

The environment has families that in some way all come from delinquency and if I expose my daughter to that environment, she will change the way she talks, swear words will become normal for her.

 --father, private subsidized school

More than peer effects, other parents openly fear for the security and integrity of their children inside the public schools because those schools are seen as chaotic institutions lacking basic management. This in practice implies that schools adopt weak and permissive disciplinary policies that fail to control the potential danger coming from some students.

The risk of being assaulted or having a bad experience, of seeing something very bad, these things I would avoid.

 --mother, private subsidized school

Ultimately, the management problems of the public schools have academic repercussions. According to the perceptions of these parents, public schools are less demanding and have fewer opportunities to learn, inadequate teaching materials and many overwhelmed teachers who are tired or not committed to their work. Thus, the cycle of social exclusion is closed with the lack of educational opportunity.

Public schools are so busy preventing children from fighting and insulting each other, that teaching is increasingly set aside.

 --mother, private subsidized school

For the parents from low social classes, to be successful in the school choice process is to be able to find institutions that guarantee their basic operation (e.g. reduced teacher absenteeism and class cancellation) and ensure the kind of socialization and security that they seek for their children. For them, public schools in their neighbourhoods are not acceptable, and even the label of private schools may not be enough if they are free. In this context, the co-payment, even if small, is basically perceived as a mechanism to restrict access to families that they want to avoid having in the school space.

Schools are like this: if you pay more, the quality of the people is better, if you pay less, the category of people also goes down because public schools receive all people and the other ones don’t.

 --mother, private subsidized school

Nevertheless, for some parents, all these strategies are not enough and they enrol their children outside of their neighbourhoods (even in a public school), where they perceive the risk from the environment is lower and the somewhat higher social-cultural level of the families could enrich the socialization process.

The people who want a better future for their children, we have to get them out of here. I send them to a school in [low middle class neighbourhood] because the environment is better, the children come from families, they don’t look like gang members, they don’t have bad behavior, it’s another thing, the mothers are concerned about school, that the children do their homework and don’t have bad peers.

 --mother, private subsidized school

As is seen, while we can identify an explicit intention by some low social class parents to segregate themselves, in others what is predominant in school choice seems to be a strategic response to an educational system in which the public schools available in their neighbourhoods do not provide the conditions that guarantee minimum standards of safety and learning for their children. For these parents, the public schools they know socialize not for the integration into the whole society but instead for a social sector in which the norm is to be excluded. So for those who experience daily exclusion, private subsidized schools (especially with co-payment and located outside their neighbourhoods) are seen as a pathway through which the children can attain basic incorporation into the social structure.

These kids, in addition to attending a [public] bad school [located in a marginal neighbourhood] with poor discipline, bad results, have parents with little education who don’t support them much because of their jobs. It is like leaving them in the back yard and burying them. Here, those who can escape from it.

 --mother, public school

Certainly, not all low social class families demonstrate the dispositions in relation to school choice we have described. In the ‘inferior’ part of the school system, there are parents who cannot or who do not feel it is important to separate themselves from those families that the others want to avoid. When they live in small cities, it is common that they do not perceive the level of risk and marginality characteristic of the densely populated metropolitan areas. At the other extreme, some parents living in metropolitan areas consider the risks so omnipresent in their neighbourhoods that it is better to learn how to face them. These families enrol their children in the schools closest to their homes in order to be near in case problems arise and choose public or free private schools, either because they trust in public education or because they don’t see significant differences among the schools available to them, including low-cost private schools. In other words, they are the families who are where and with whom nobody wants to be. After all if they didn’t exist, from whom would the parents in the first group want to separate?

I have always emphasized that schools are all good, it depends on you, if you want to study . . . there are good and bad in every place. It is a mixed bag.

 --mother, public school

3.3.2. Middle social classes

The middle classes are the largest segment of the Chilean population and also the most heterogeneous. They include the traditional and the new middle classes. They are also the most active group in relation to the dynamic of the school market in Chile, associated with the rapid growth of subsidized private schools. Although with different emphases, the basic meaning of school choice for these families is to consolidate a position in the social structure that is self-perceived as ‘middle class’ and, through education, to avoid a decline in social status, which is consistent with Rojas et al. (2016).

Generally, middle-class parents choose paid private schools with or without public subsidy, spending the maximum amount they can afford to pay. They are willing to commute far distances and also to satisfy many requirements of the schools to which they are applying. In the case of the new middle class, they feel high levels of stress when they cannot have access to the schools they select because they perceive the risk of imminent loss of social status.

The ‘new middle class’

The new middle social class identify themselves in opposition to the lower social classes. Since the new middle classes have recently overcome poverty, they have some experiences in common with the lower social class, from which they want to distance themselves in order to create a dignified and legitimate social space: a middle class of ‘work and effort’. Although in many respects, the cultural background of their school choice strategies are similar to those of the more active parents of the lower social class, the key difference lies in the fact that for new middle class families, it is not so much to ‘leave’ where they are but to ‘consolidate’ the position they have attained. That is why the effort to distinguish themselves and not to be seen as part of the lower social class is so marked.

The parents of the new middle class are the clientele of the subsidized private schools to the point that choosing this kind of school appears for them to be a kind of moral obligation. To pay for the education of their children – with the effort that this implies – is seen as a duty, and at the same time results in differentiating themselves from the lower social class who, according to their perception, live totally dependent on state subsidies. For middle social class families, their achievements are explained essentially by their own individual efforts and merits.

All my friends told me: ‘don’t be a fool, don’t spend so much money!’ But it is important; I want the best for my daughter and today this school is the best I can give her. I have four jobs, but if tomorrow I get a different job and I can afford something better, I will do it.

 --father, subsidized private school

These parents choose subsidized private schools as a category and rule out public schools because of their low academic results and the scarcity of teaching materials, but mainly because of the social environment produced by accepting all kinds of students and particularly low-income students to whom they attribute social and behavioural problems. After all, these parents assume that all public services are low quality and they try to avoid them.

I can’t imagine; I haven’t seen good schools that are free, there are very few or at least they are not where I live. My neighborhood is not bad, so there are not many free schools.

 --father, subsidized private school

They prefer subsidized private schools with co-payment since they associate these characteristics with better academic results, more resources (better infrastructure, psycho-social professionals, extracurricular activities, among others) and essentially families with a similar or somewhat higher social status. Since parents from the new middle class consider that the higher the tuition, the better the school, they are always willing to pay more, even more than is allotted in their family budget. They also attribute moral integrity to families who pay for education or at least consider that these families have more concern and value for the education of their children.

Finally, it is clear that the more you pay the higher the school’s test scores. So, you have to try to choose the best and simply pay; that way you guarantee results.

 --father, subsidized private school

Their definition of ‘quality education’ is closely linked to the type of families that make up the schools; thus, families with more resources and education directly or indirectly contribute to produce a better school.

Sadly, the economic factor, the price, tells you how good a school is. So, you have to look at the tuition to know a school, what children it serves, what families go there.

 --mother, subsidized private school

In short, for these families of the new middle class, the meaning of the school choice is to avoid the lower social class in the school space, justifying this by the threat they represent to the socialization and education of their children, because they typically associate them with characteristics of the marginalized sectors.

If those children attended the school of my son, they would pull down the current students, who have values and other things that you teach them at home, while those children do not come with this background. More than a low socioeconomic level, they have a low sociocultural level.

 --father, subsidized private school

In this context, the co-payment and to some extent the student selection are explicitly valued as mechanisms to avoid social diversity.

I don’t want to discriminate but I think that the requirement of money is also important because this discriminates against a certain type of people (children with bad education or with family problems). So the fact that there is a little bit of money excludes a huge amount of people.

 --mother, subsidized private school

Although some parents are open to paying more to obtain a better social environment, many parents also recognize the limits of this strategy, since they feel comfortable in the schools interacting with families from a similar socio-economic situation. Hence, the combination of both dispositions reduces markedly the range of eligible schools.

I saw similar people, hardworking people, I saw nothing abnormal. You always search for the average or higher, I was looking for people like me. I don’t want my daughter to feel that she doesn’t belong to the group.

 --mother, subsidized private school

Lastly, their willingness to commute to schools outside of their neighbourhood is closely linked to the social composition of their own neighbourhoods and the level of risk they perceive from their social environment. When they live close to or within low-income neighbourhoods, they assume they have to go far in order to find a satisfactory school. On the contrary, when they live downtown or in rapidly growing new neighbourhoods with homogeneous socio-economic composition (which is the most common situation for this group), they perceive that schools in their proximity are satisfactory. Certainly, this is the social space that mostly explains the expansion of for-profit subsidized private schools in Chile, the prototypical situation of the new school market.

The traditional upper middle class

Traditional upper middle class parents normally prefer non-subsidized private schools. They choose educational projects with a wide scope and they have the expectation that this will allow their children to pursue the professional career they want in a prestigious university, something that they see as essential for a successful positioning in the social structure.

They rule out public schools in general, since they do not meet their quality standards, which in addition to academic results, include the acquisition of a second language and various extracurricular activities. However, there is an exception: traditional academically selective public high schools located in the centre of the city.

Today there is no academically strong public primary school that gives me confidence that my children will get a good education and not only in the academic sense but also that fosters other skills. The only school that I know with those characteristics is too far. It’s the only one.

 --mother, subsidized private school

Whenever possible, they also avoid subsidized private schools and attend them only when family economic restrictions or students’ low academic performance or behaviour problems prevent them from accessing a non-subsidized private school. They complain that subsidized public schools provide lower-quality education compared to the non-subsidized private, do not have satisfactory infrastructure, place for-profit motives above educational aims and do not guarantee the comprehensive development of their children. Similarly, for some parents, the socialization of their children with children from families of lower socio-economic and cultural status would be undesirable.

It is a subsidized private school but, because the school has social programs and gives scholarships to some students, they receive children from poor neighborhoods so suddenly it is full of little scoundrels. No – forget about it!

 --mother, public school

Nevertheless, within the category of non-subsidized private schools, parents from the traditional middle class also rule out elite private schools because they believe that those schools do not pertain to the middle-class social space and do not provide the socialization that they are seeking. Hence, similar to new middle-class parents, they want to meet families in the school space who have the same or somewhat higher socio-economic status, but never significantly lower or higher.

There are schools that are much more elitist than others. You have to search for schools where the children are in an adequate social level, and this school is more or less where we are socially and economically comfortable.

 --father, non-subsidized private school

In fact, some parents perceive, within their private and tuition charging schools, relevant differences (cultural more than economic) among the parents at those circuits of schooling.

People who can afford a non-subsidized private school mainly have a socioeconomic and cultural status similar to us. That is different from people who with effort go outside of their cultural environment, but you notice that in the way they interact with others, the tone of their speech that is different, different expressions and that puts us off.

 --mother, non-subsidized private school

Even parents who are aware of the segregation effect of their decisions will still send their children to private paid schools. In the end, the family interests prevail over any kind of concern about the collective good. These parents think that the future social positions of their children heavily depends on academic trajectory, which starts as early as when they enter the school system.

I will not sacrifice my son. I want to give him the best and maybe I would send him to a public school to make the education less segregated, but I know that if I send him to this school and pay, he will receive a better education and will have more opportunities.

 --mother, non-subsidized private school

Generally, these parents consider that the schools in their neighbourhoods satisfy these requirements; actually, many of them choose to live close to those schools and to avoid a long commute for the children. These are comparatively expensive areas of the city.

I know that here and in [high social class district], you can enroll your children in any school and they will be good always, but in other districts you have to do a search.

 --mother, public school

Besides the analysed patterns, some parents from the traditional middle class also contribute to the reproduction of school segregation even if they do not give value to the social composition of the schools. Some families prefer to enrol their children in schools close to their homes, but since they live in already segregated neighbourhoods in which most of the schools are private, they unintentionally reinforce the social segregation of the schools. Other parents choose schools exclusively based on educational criteria and look for academically highly demanding schools or alternative schools (e.g. Montessori and Waldorf schools); these parents also contribute to segregation because they are not satisfied with the public or subsidized schools, and decide to choose expensive private schools.

3.3.3. High social class

Chile is a country with extreme income concentration and high social class families live in hyper-segregated neighbourhoods. For high-class parents, school choice is a sociocultural issue that involves world views (including the pedagogical dimension although less precisely), and look for the cultural coherence between the family and the institution.

They understand education as an essential part of their children’s personal development and their incorporation into the community to which they pertain. In addition, they feel that the success of their family projects heavily depends on the enrolment of their children in the right educational institutions which are extremely scarce. So for these families, eligible schools are limited to a small group of non-subsidized private schools that charge high incorporation fees, enrolment fees and tuition to families. Surely, they perceive a significant diversity within this group and seek multi-dimensional educational projects that integrate community, individuality and academic achievement.

Interestingly, for some high-class parents, there is practically no ‘choice’ of the school since they perceive themselves as members of sociocultural communities of which the schools are also a part (i.e. Catholic Jesuit, Catholic Opus Dei, specific nationalities). Other parents, either because they have a rational approach to school choice or because they have entered the high class more recently, look for more detailed information, accumulating a large body of knowledge about the schools. Certainly, high academic performance is taken for granted; even more, in some ways they do not trust the institutional version of quality education linked to standardized testing. At the end, both paths drive them to elite schools.

Notwithstanding, these parents perceive that the main challenge is not choosing a school but to be accepted by it. The access to an elite school is in no way guaranteed just because the family has the economic resources to pay for it. The use of strict additional filters like academic requirements for students, interviews with the parents or being part of a certain religion or community (ethnic-linguistic or former students of the school) is almost universal in these schools.

non-religious group

In my case it was very hard, really I felt very intimidated because since they first give a psychological test to the children, when you go to the meeting with the principal, they interrogate you a lot to see whether you have the same impression that the principal has of your child and you start providing explanations for everything.

In this sense, the choice happens from the family to schools and from the schools to the family in a highly demanding process of mutual identification that ultimately attempts to certify membership of the family to a certain sociocultural group.

They choose homogenous families and provide the child a homogenous environment because it is easier for the school.

 --non-religious group

Ultimately, it is the schools that decide. The parents find the schools to which they apply to be chosen; the schools choose whom they accept. Conscious of the complete asymmetry of power in the process, these families from the high class are willing to undergo all kinds of requirements from the schools. Thus, elite schools are by definition hyper-segregated communities.

Practically, in Chilean education, it is not that we choose, in the end it is the school that chooses you.

 --non-religious group

4. Conclusions

Academic research has accumulated a considerable amount of evidence to understand the factors associated with the production of socio-economic school segregation in Chile, which is comparatively very high. We interpret those findings as robust evidence of the segregation effects caused by market dynamics prevalent in Chilean education. Certainly, other potentially relevant dimensions, such as the strong test-based school accountability and the curricular specialization in high schools (comprehensive versus vocational tracks), require additional research to be linked to socio-economic school segregation in Chile.

On the supply side, the most relevant segregation forces are the increasing presence of private schools, the for-profit motivation of most of the new private schools, the relevance of tuition-charging private schools (including state-funded schools), the discriminatory school practices during the admission and the schooling processes, and the competition for family preferences among all kinds of schools (including public schools) in order to obtain public funding via vouchers. All of these factors are more relevant than the urban segregation that also contributes to produce school segregation. On the demand side, families from all social classes apply strong socially oriented criteria for school choice: broadly, high social class families prefer to educate their children in socioculturally homogeneous elite schools; middle-class parents reject free and non-selective public schools to avoid low-income families that they associate with risky behaviours; and while some low social class families share a similar rationale with the middle classes, others do not see school choice as a relevant means to avoid social risks and enrol their children in neighbourhood schools, which in large cities tend to be highly segregated environments. Although the centrality of social aspects for school choice is decisive in most cases, there are families for whom the meaning that guides the school choice does not fall into that dimension (e.g. religious or academically oriented families); nevertheless, given the institutional characteristics of Chilean education, the actions of those families inadvertently also have an effect on segregation.

A key policy implication of the previous analysis is the need to recognize the complex educational, sociocultural and political challenges a policy agenda oriented to de-segregate Chilean schools would face, since segregation forces are strongly embedded in both institutional and social practices. This became apparent during the polarized debates around the current educational reform focused on diminishing the relevance of market-oriented dynamics in Chilean education, including the end of for-profit schooling, co-payment, student selection during the admission processes in subsidized schools, and the reinforcement and expansion of public schools (Bellei, 2016).

Certainly, some of the discussed findings about parents’ school choice as a factor in school segregation relate to international literature, since middle- and high-class parents would reinforce it mainly through the departure from some public schools. Actually, although the middle and high social classes would be closer to the prototypical market agent favouring educational quality based on an instrumental rationality, findings point out that socio-economic school composition is a key factor for them. In particular, those families who had more in-depth knowledge about school work were more conscious about the peer effects (Gombert, 2008; Felouzis and Perroton, 2009; Van Zanten, 2009). Hence, they link school effectiveness and prestige not only with school management and pedagogy but also with the kind of public that they serve, and students from the low social class are usually associated with academic and behavioural problems by middle- and high-class families. In addition, their negative regard for the cultural patterns of low-income students in the educational space encourages them to avoid social diversity in school.

Likewise, school segregation varies significantly among countries and evidence suggests that some of the market dynamics we discussed in the Chilean context could partially explain those differences (Waslander and Van der Weide, 2010; Musset, 2012). Basically, school systems that provide parents with greater school choice possibilities have higher probability of incrementing socio-economic, ethnic or academic school segregation (Jenkins et al., 2008; Alegre and Ferrer, 2006); also the discretion given to schools to manage their admission processes is an additional segregation factor (Maroy, 2008). Furthermore, before the admission processes, schools affect student distribution by specializing their offerings in order to attract the demand of certain types of families by designing specific educational projects and locating them in local school markets, reinforcing segregated ‘circuits of schooling’.

From this perspective, what the Chilean case clearly shows is the enormous transformative potential of market dynamics in education. Conceptually, in the traditional model based on public education, schools should be equal to guarantee equity and to allow a meritocratic contest among students, but in the market model, schools should be different to compete for students. Market proponents expect that the internal homogeneity of schools and maximum heterogeneity between schools will be both a facilitating factor for the educational process and a reason for parents’ satisfaction (Chubb and Moe, 1990). This is a radical departure from the functionalist liberal vision: for Durkheim and Parsons, that education should introduce students to the normative order of society; from market theory, education should be a communitary extension of the family. That is why school segregation can be considered an expected result – not a side effect – of market-oriented education, even if unintended or undesirable for some education actors.


Alegre, M. and Ferrer, G. (2006), School regimes and education equity: some insights based on PISA. British Educational Research Journal , 36 (3): 433–61.

Arteaga, F., Paredes, V. and Paredes, R. (2014), School Segregation in Chile: Residence, Co-payment or Preferences? Unpublished manuscript.

Arzola, M. P. and Castro, R. (2013), Indicadores de inclusión socioeconómica del sistema escolar. In Serie Informe Social (no141), Chile, Libertad y Desarrollo.

Atria, R. B. (2004), Crecimiento y estratificación social: la paradoja del crecimiento y el caso chileno. Estudios Sociales (Chile) , 85–116.

Auguste, S. and Valenzuela, J. P. (2004), Do students benefit from school competition? The Chilean experience . Doctoral dissertation, University of Michigan, Ann Arbor.

Barozet, E. and Fierro, J. (2011), Clase media en Chile, 1990–2011: algunas implicancias sociales y políticas . Santiago: Konrad Adenauer Stiftung, Serie de Estudios, 4.

Bellei, C. (2009), ‘The Private-Public School Controversy: The Case of Chile’, in Paul Peterson and Rajashri Chakrabarti (eds.), School Choice International , Cambridge, MA: MIT Press, pp. 165–92.

Bellei, C. (2015), El gran experimento. Mercado y privatización de la educación chilena . Santiago: LOM Ediciones.

Bellei, C. (2016), Dificultades y resistencias de una reforma para des-mercantilizar la educación . RASE: Revista de la Asociación de Sociología de la Educación , 9 (2): 232–47.

Bellei, C. and Pérez, C. (2014), ‘Democratizar y tecnificar la educación. La reforma educacional de Eduardo Frei Montalva’, in C. Huneeus y J. Couso (eds.), Eduardo Frei Montalva: un gobierno reformista. A 50 años de la ‘Revolución en Libertad’ , Santiago, Editorial Universitaria.

Bellei, C. and Vanni, X. (2015) ‘The Evolution of Educational Policy in Chile 1980–2014’, in S. Schwartzman (ed.), Education in South America . London: Bloomsbury Publishing.

Canales, M., Bellei, C. and Orellana, V. (2016), ¿Por qué elegir una escuela privada subvencionada? Sectores medios emergentes y elección de escuela en un sistema de mercado. Estudios Pedagógicos , 42 (3): 89–109.

Carrasco, A. and San Martín, E. (2012), ‘Voucher System and School Effectiveness: Reassessing School Performance Difference and Parental Choice Decision making’, Estudios de Economía , 39 (2): 123–41.

Carrasco, A., C. Flores, E. San Martín, G. Gutiérrez and F. Bogolasky (2014), Selección de estudiantes y desigualdad educacional en Chile. ¿Qué tan coactiva es la regulación que la prohibe? Informe proyecto FONIDE 711286, Chile: CEPPE, Pontificia Universidad Católica de Chile.

Castro, F. and Kast, S. (2004), Movilidad de la pobreza en Chile: análisis de la encuesta panel. In Serie Informe Social (No. 85). Chile. Libertad y Desarrollo.

Chubb, J. and Moe, T. (1990), Politics, Markets, and America’s Schools . Washington, DC: Brookings Institution Press.

Consejo Asesor Presidencial por la Calidad de la Educación (2006), Informe Final . Consejo Asesor Presidencial por la Calidad de la Educación.

Contreras, D., Sepúlveda, P. and Bustos, S. (2010), ‘When Schools Are the Ones that Choose: The Effects of Screening in Chile’, Social Science Quarterly , 91, 1349–68.

Contreras, M. (2015), La incidencia de la elección de escuela de los padres de clase media en la segregación escolar . Tesis Pregrado. Universidad de Chile.

Córdoba, C. (2012), Elección de escuela en Chile: la propuesta neoliberal a examen . Disertación Doctoral, Universidad Complutense de Madrid.

Corvalán, J., Carrasco, A. and García-Huidobro, J. E. (eds.) (2016), Mercado escolar y oportunidad educacional. Libertad, diversidad y desigualdad . Santiago: Ediciones Universidad Católica de Chile.

Elacqua, G. (2012), ‘The Impact of School Choice and Public Policy on Segregation: Evidence from Chile’, International Journal of Educational Development , 32 (3): 444–53.

Elacqua, G. and Fábrega, R. (2006), El consumidor de la educación: El actor olvidado de la libre elección de escuelas en Chile . In Educación y brechas de equidad en América Latina . Santiago: PREAL, pp. 353–98.

Elacqua, G. and Santos, H. (2013), Los efectos de la elección escolar en la segregación socioeconómica en Chile: Un análisis georreferenciado . Espacio Público, Documento de Referencia, 01.

Elacqua, G., M. Schneider, and J. Buckley. (2006), ‘School Choice in Chile: Is it Class or the Classroom?’, Journal of Policy Analysis and Management , 25 (3): 577–601.

Espínola, V. (1993), The Educational Reform of the Military Regime in Chile: The School System’s Response to Competition, Choice and Market Relations . University of Wales, Tesis de Doctorado.

Espinoza, V., Barozet, E. and Méndez, M. L. (2013), Estratificación y movilidad social bajo un modelo neoliberal: El caso de Chile . Revista Lavboratorio , 25 (14): 169–91.

Felouzis, G. and Perroton, J. (2009), Grandir entre pairs à l’école. Ségrégation ethnique et reproduction sociale dans le système éducatif français . Actes de la recherche en sciences sociales , 5 (180): 92–100.

Flores, C. and Carrasco, C. (2013), Preferencias, libertad de elección y segregación escolar . Espacio Público, Documento de Referencia, 02.

Gallego, F. and Hernando, A. (2008), On the Determinants and Implications of School Choice: Semi-Structural Simulations for Chile . Documento de Trabajo, 343, Instituto de Economía PUC.

Gallego, F. and Hernando, A. (2009), School choice in Chile: looking at the demand side . Documento de trabajo, 356, Instituto de Economía PUC.

García-Huidobro, J. E. and Bellei, C. (2003), Equidad educativa en Chile . Documento de Trabajo, Departamento de Educación, Universidad Alberto Hurtado.

Gauri, V. (1998), School Choice in Chile. Two Decades of Educational Reform , Pittsburgh: University of Pittsburgh Press.

Gombert, P. (2008), L’École et ses stratèges. Les pratiques éducatives des nouvelles classes supérieures . Rennes, Presses universitaires de Rennes.

Grau, N., Hojman, D. and Mizala, A. (2015), Destructive Creation: School Turnover and Educational Attainment . University of Chile, Santiago.

Hsieh, C. T. and Urquiola, M. (2003), When Schools Compete, How Do They Compete? An Assessment of Chile’s Nationwide School Voucher Program . NBER Working Paper Series.

Labarca, A. (1939), Historia de la enseñanza en Chile . Imprenta Universitaria.

Maroy, C. (2008), Por qué y cómo regular el mercado educativo . Revista de Profesorado , 12 (2).

McEwan, P. (2001), ‘The Effectiveness of Public, Catholic, and Non-Religious Private Schools in Chile’s Voucher System’, Education Economics , 9 (2): 103–28.

Mineduc (2012), Medidas de segregación escolar . Centro Estudios Ministerio de Educación, Serie Evidencias.

Musset, P. (2012), ‘School Choice and Equity: Current Policies in OECD Countries and a Literature Review’, OECD Education Working Papers, No. 66, OECD Publishing.

Núñez, I. (2003), La ENU entre dos siglos . Santiago: LOM Ediciones.

OECD (2004), Reviews of National Policies for Education: Chile . OECD.

OECD (2016), PISA 2015 Results (Volume I): Excellence and Equity in Education , PISA, OECD Publishing, Paris.

Paredes, R.; Volante, P.; Zubizarreta, J. R. and Opazo, M. (2013), Financiamiento compartido en la educación subvencionada chilena . In Propuestas para Chile, Concurso Políticas Públicas, Pontificia Universidad Católica de Chile.

Raczynski, D. and Hernández, M. (2011), Elección de colegio, imágenes, valoraciones y conductas de las familias y segregacion social escolar . Santiago: Asesorías para el desarrollo.

Rojas, M. T., Falabella, A. and Leyton, A. (2016), Madres de clases medias frente al mercado educativo en Chile: decisiones y dilemas . In Corvalán, J., Carrasco, A. and García-Huidobro, J. E. (eds.), Mercado escolar y oportunidad educacional. Libertad, diversidad y desigualdad . Santiago: Ediciones Universidad Católica de Chile.

Roje, P. (2014), Segregación escolar por nivel socioeconómico y su relación con las políticas educacionales en Chile: el caso del financiamiento compartido y la subvención escolar preferencial . Universidad de Chile. Tesis de Master.

Ruiz, C. and Boccardo, G. (2014), Los chilenos bajo el neoliberalismo. Clases y conflicto social . Santiago: Nodo XXI/El Desconcierto.

Ruiz, C. and Orellana, V. (2011), Panorama social del Chile del bicentenario . Análisis del año 2010 , 27–51.

Salas, I. (1930), The Socio-economic Composition of the Secondary School Population of Chile . Columbia University, PhD Dissertation.

Schneider, M., Elacqua, G. and Buckley, J. (2006), School choice in Chile: Is it class or the classroom? Journal of Policy Analysis and Management , 25 (3): 577–601.

Torche, F. and Wormald, G. (2007), Chile, entre la adscripción y el logro. En: Estratificación y movilidad social en América Latina: transformaciones estructurales de un cuarto de siglo . Santiago: NU. CEPAL/LOM Ediciones, pp. 339–87.

Valenzuela, J. P., Bellei, C. and Ríos, D. D. L. (2014), Socioeconomic school segregation in a market-oriented educational system. The case of Chile . Journal of Education Policy , 29 (2): 217–41.

Valenzuela, J. P., Villalobos, C. and Gómez, G. (2013), Segregación y polarización en el sistema escolar chileno y recientes tendencias . Espacio Público, Documento de Referencia, 03.

Van Zanten, A. (2009), Le choix des autres. Jugements, stratégies et ségrégations scolaires . Actes de la recherche en sciences sociales , 180, 24–34.

Vazquez, E. (2016), Segregación escolar por nivel socioeconómico. Midiendo el fenómeno y explorando sus determinantes . Económica , 62, 121–84.

Verger, A., Bonal, X. and Zancajo, A. (2016), ‘What Are the Role and Impact of Public-Private Partnerships in Education? A Realist Evaluation of the Chilean Education Quasi-Market’, Comparative Education Review , 60 (2), 223–48.

Villalobos, C. and Valenzuela, J. P. (2012), Polarización y cohesión social del sistema escolar chileno . Revista de análisis económico , 27 (2): 145–72.

Waslander, S. Pater, C. and Van Der Weide, M. (2010), Markets in Education: An Analytical Review of Empirical Research , OECD. Education work papers. n°52. OECD Publishing.

[1] This work was supported by the PIA-CONICYT Basal Funds for Centers of Excellence [grant number BF0003] and FONDECYT [grant number 1130430].

[2] Arzola and Castro (2013) replicated the same results when estimating the Duncan Dissimilarity Index, but they later estimated a different index (which they preferred) and obtained inverse results.

[3] This section is based on the study ‘The sociocultural foundations of the school market’ (Fondecyt 1130430). The study collected data from 102 semi-structured interviews with parents (seventy-seven) and school administrators (twenty-five), and twelve focus groups with parents, in four different zones in the Santiago Metropolitan Area and two additional Chilean cities. Parents were selected based on their residential locations and social class situation.

[4] The sample was composed following basic characteristics of the Chilean stratification structure (Atria, 2004; Barozet and Fierro, 2011; Espinoza et al., 2013; Ruiz and Boccardo, 2014; Torche and Wormald, 2007; Castro and Kast, 2004; Ruiz and Orellana, 2011). High social class (about 5 per cent of the Chilean population) included families living in both traditional and new very expensive areas of Santiago, who are either high status professionals or business owners; middle-high class (around 15 per cent of the total) included professionals and highly qualified technicians, living in quite expensive traditional residential neighbourhoods; ‘new’ middle class sectors (currently the most dynamic and numerous, including about 50 per cent of the Chilean population) typically live in areas of recent urban expansion in the cities, are semi-skilled workers and high-school graduates working in the service sector, or in both public and private bureaucracy; finally, low-class families (around 30 per cent of the total) live in poor neighbourhoods and sometimes marginalized areas of the cities, they encompass the working class, and unskilled workers of personal services (e.g. housemaid) and subsistence economy.