How do marginalized families engage school choice in inequitable urban landscapes? A critical geographic approach

The normalization of school choice in the education system is purported to provide more schooling options for all families, particularly those who do not have the means to move into affluent areas with ‘better’ schools. Nonetheless, it is still unclear to what extent the policy of school choice has been effective in achieving the goal of providing more choices for marginalized families. This paper aims to fill this gap by examining the K-12 school choice practices and patterns of marginalized urban families, with a focus on their spatial positions and dispositions, in what is arguably one of the most rapidly diversifying and polarizing cities in the world, Vancouver, Canada. An innovative mixed-methods critical geographic approach is used to better understand the families’ school choice participation and related mobility patterns geo-spatially , while exploring their choices phenomenologically .


Introduction
The development of market mechanisms in K-16 education systems is growing worldwide, including in Canada (Adamson, Astrand, & Darling-Hammond, 2016;Ball, 2003;Forsey, Davies, & Walford, 2008).The deregulation of the private sector has accelerated while government subsidies have increased in the forms of tax credits, vouchers, or tuition payments to make private school choice more viable (Adamson et al., 2016;Fullan & Rincon-Gallardo, 2016).In the public sector, government spending and school board budgets have been cut.These cuts have left many school boards unable to cope adequately with the rising costs of education (e.g.labor, technology, and educating students with special needs and immigrant backgrounds).Meanwhile, many public school boards and ministries of education have expanded school choice policies under the premise of increasing school choice options for all families.This increasingly neoliberal approach to educational funding and organization are creating new structures and conditions of education.The normalization of school choice is purported to provide more schooling options for all families, particularly those who do not have the means to move into affluent areas with 'better' schools (Adamson et al., 2016;Forsey et al., 2008).Nonetheless, it is still unclear to what extent the policy of school choice has been effective in achieving the goal of providing more choices for marginalized families (Lubienski, Lee, & Gordon, 2013).Especially in the increasingly diversifying context of Canada, which is known for having had more equitable distribution of resources than other countries (Perry, 2009), there are few studies that inform us about marginalized families' participation in school choice.
This study thus aims to expand our understanding of the practices and patterns of marginalized urban families' school choice by drawing from critical social and spatial theories of education (Bourdieu, 1989(Bourdieu, , 1998(Bourdieu, , 1999;;Gulson, 2007;Soja, 1996).This socio-spatial framework is used to illuminate that choosing a school is not entirely an academic exercise, and that parents include social geography as part of their concerns (Butler & van Zanten, 2007;Lubienski & Dougherty, 2009).The study uses a mixed-methods research design, drawn from the scholarship of Critical Geographic Information Systems (CGIS) (Kwan & Ding, 2008), to answer the following question: How do the spatial positions and dispositions of low-income families shape their school choice participation patterns and practices?Spatial positions refer to families' residential locations in urban geography (formed by different levels of income, wealth, and education).Spatial dispositions refer to families' familiarity, comfort, resources, routines, perceptions, and emotions that develop based on where they reside.
This study finds that most low-income, racially marginalized families choose schools close to where they live.They do so as a result of economic constraints and housing choices, given the resources that are available in the forms of institutional and social support (family and friends).In addition, their choice tends to be rooted in their sense of schools as social spaces.Once families settle in a neighborhood, they develop a spatial disposition, consisting of commuting routines, connections to neighbors, utilization of resources, and emotional feelings, all of which contribute to creating a sense of a comfort zone around their neighborhoods.Their spatial dispositions tend to influence their school choice by shaping their decisions to choose a school within areas where they feel comfortable and where they "fit in" within the city.The findings thus indicate that as the city becomes increasingly spatialized due to growing inequality and racialized poverty, the marginalized spatial positions and dispositions of low-income families are likely to further shape and constrain their choices.

Literature: Low-income Families' School Choice Practices and Constraints
While there is a sizable body of literature on the school choice practices and challenges of low-income families, it is not yet clear to what extent school choice has benefited those families.What we begin to understand is that the primary concerns of low-income (LI) parents who choose schools are high academic quality and a curriculum that meets their children's learning needs (Teske, Fitzpatrick, & Kaplan, 2007).LI parents use as indicators standardized tests and/or evaluations of teacher quality (Smrekar & Honey, 2015).They also consider a school's learning climate (Martinez & Thomas, 1994).LI parents choose schools that have higher expectations for their children, especially when parents find that their children, as racial minorities, face prejudices.That is especially in the case with African-American working-class mothers (Cooper, 2007).In making choices, LI parents consider a few realistic options, which resembles the school choice concerns of other income groups (Teske et al., 2007).
LI parents, however, are somewhat different from other income groups in that they are more likely to rely on teachers and counselors rather than other parents to learn about school choice options (Teske et al., 2007).They tend to rely on individual schools and public media to learn about choice programs because their social networks provide limited information (Smrekar & Honey, 2015).LI parents appear to know very few market mavens, people or groups who are well-informed and active in distributing school choice information (Teske et al., 2007).These parents generally lack of awareness of school choice (Martinez & Thomas, 1994), and in fact LI parents tend to consider their children better equipped than they are in making school choice decisions, and thus they give their children more power in making those decisions (Reay & Ball, 1997;Taylor, 2002).
The families that choose schools and fall at the bottom of the income distribution ladderthat is, those whose annual incomes are below US$20,000-are twice as likely to have proximity as an important factor than are school choice families with an income level between US$20,000 and US$50,000 (Teske et al., 2007).Also, this former group reports that they do not have the tools and means to participate in school choice, do not have useful information about school choice, and are the least satisfied with school choice process and least likely to visit schools (Teske et al., 2007).LI parents further stress the importance of a sense of safety, familiarity, and the presence of their children's friends (Martinez & Thomas, 1994;Teske et al., 2007).
Across racial groups, minority LI parents are more likely than white parents to cite the availability of child care and extracurricular activities as important reasons for choosing schools (Martinez & Thomas, 1994).Culturally relevant programs with a focus empowering minority groups are also critical (Cooper, 2007).Hence, while minority groups tend to choose schools where a majority of the school's population shares their racial background (Bunar, 2010;Henig, 1996), it is also the case that racialized low-income families are more likely to choose schools located in wealthier and white neighborhoods (Ndimande, 2016;Smrekar & Honey, 2015;Yoon, in press).As such, enrollment in magnet schools in mostly white neighborhoods reflects the racial composition of the district as a whole, while enrolment in mostly non-white neighborhood schools reflects the racial composition of the neighborhoods (Smrekar & Honey, 2015).
In terms of parent characteristics, choosing LI parents tend to have higher incomes than non-choosing LI parents (Teske et al., 2007).Also, choosing LI parents have some college education, while half of all non-choosing LI parents did not graduate from high school (Martinez & Thomas, 1994).They also tend to be "advocate" parents who see school choice as empowering their children to be successful, especially African-American mothers who perceive the current public school system as having low expectations of African-American children (Cooper, 2007).
In terms of constraints, LI racial minority parents identify financial constraints as an important factor in school choice (Martinez & Thomas, 1994;Ndimande, 2016).Also, some African-American mothers note that they have faced barriers from charter school administrators when they try to enroll their children (Cooper, 2007).Furthermore, research shows that schools' decisions about location and attendance boundaries tend to limit the high-quality school choice opportunities available to low-income families.For instance, charter schools (i.e.choice public schools, mostly in the U.S.) tend to avoid the most impoverished neighborhoods (Gulosino & Lubienski, 2011;Lubienski & Dougherty, 2009).Also, schools tend to under-enroll disadvantaged students by creating enrolment boundaries (Lubienski et al., 2013).
Overall, although school choice has had some, albeit limited, positive academic and social mobility effects on disadvantaged students, overall it has had marginalizing consequences for lowincome and ethnic minority youths and neighborhoods (DeLuca & Dayton, 2009;Reay & Lucey, 2004;Yoon, 2013).In some cases, school choice has gained international momentum through the grassroots actions of low-income minority communities who wish to create "emancipatory" school spaces for ethnic minority families.School choice has led to some positive outcomes in strengthening cultural identities and improving the academic outcomes of marginalized communities (Dei, 1995;Rofes & Stulberg, 2004;Wilson, 2016).Nonetheless, these schools often face challenges of inadequate funding and public resistance (DeLuca & Dayton, 2009;Gulson & Webb, 2012;Wells, Lopez, Scott, & Holme, 1999).
This paper thus builds on the existing research on low-income families' school choice with the aim of further illuminating the nexus between the spatial positions and dispositions toward school choice by low-income families, an understudied population, especially in the Canadian context.

Theory: Socio-spatial Perspective on School Choice
The importance of geography as space and place has been highlighted in the recent studies of middle-class families' school choice.That is, school choice is not simply an economic decision, but also reflects social group affinity, feelings, and behaviors of group belonging, as manifest in urban geography (Bell, 2009;Butler & Robson, 2003;Gabay-Egozi, 2016;Reay, 2007).As such, while middle-income parents have the means to choose a school anywhere, they seldom choose schools in disadvantaged neighborhoods, even when those schools are high-achieving (Holme, 2002;Smrekar & Honey, 2015;Yoon & Gulson, 2010).School choice outcomes are thus reflective of families sorting themselves out spatially in their particular urban contexts (André-Bechely, 2007;Holme, 2002;Lubienski & Dougherty, 2009).
Building upon this geographic perspective, this paper aims to illuminate the school choice of low-income families.In particular, in conceptualizing spatial positions and dispositions in understanding school choice practices, this study draws from Pierre Bourdieu's concepts of habitus, capital, field, and site.Disposition is primarily based on Bourdieu's concept of habitus, which is the perceptions and thoughts that generate meanings and guide people's everyday practices and choices (Bourdieu, 1989(Bourdieu, , 2000;;Bourdieu & Wacquant, 1992).Dispositions are shaped in the structured and stratified social world and social class system, which Bourdieu calls field.The notion of field denotes the social world as objectively discernible social spaces where classes or groups of individuals occupy different and stratified positions depending on their possession of profitable objects, status, and/or properties (Bourdieu, 1989(Bourdieu, , 2006(Bourdieu, /1986)).What creates the relationships between individuals are the possession of material and symbolic properties and power, which is referred to as capital (Bourdieu & Wacquant, 1992).The spatial manifestations of social classes (or field) with different lifestyles thus have specific spatial locations.This spatial manifestation of social class, conceptualized as site, corresponds to geographic spaces with different housing styles, shopping areas, services, and types of schools (including private schools) (Bourdieu, 1984(Bourdieu, , 1999(Bourdieu, , 2005)).
Urban neighborhoods, theorized as sites, reflect varying concentrations of symbolic, social, economic, and academic capital available within one's geographic parameters (Bourdieu, 2006(Bourdieu, /1986)).Further, these neighborhoods embody certain spatial histories and meanings that are associated with social and racial divisions and inequality (Bell, 2009;Good, 2016;Gulson & Symes, 2007;Reay, 2007;Soja, 1996).As such, neighborhoods generate a sense of inclusion and belonging for some groups but not others (Kwan & Ding, 2008;Matthews, Detwiler, & Burton, 2005).The locally circulating meanings, stories, and images of different urban neighborhoods, that is, urban imaginaries, influence how individuals perceive different neighborhoods, their identities, and their sense of where they belong, all of which in turn influence the reproduction of space and place (Lefebvre, 1991;Yoon, 2015).By linking Bourdieu's theories with these critical spatial theories we get a more layered theoretical framework for analyzing how social groups who reside in different neighborhoods with particular historic and social place meanings may reproduce urban geography, as well as how these patterns may shape school choice.
Hence, in this paper, we refer to spatial position as neighborhoods where one can afford to live given one's level of capital (especially economic capital).We use the concept of spatial disposition to theorize how residing in different neighborhoods as sites shapes individual habitus, that is, dispositions, comfort, and sense of belonging in particular neighborhoods.

Methodology: Critical Geographic Information Systems
To understand how spatial positions and dispositions shape low-income families' school choice practices, this study applies a mixed-method research design in order to understand lowincome families' school choice practices, both as spatial dispositions that can be studied qualitatively by discerning meaning-making, and as spatial positions that can be examined quantitatively by measuring physical relationships in the urban context.Qualitatively, this study explores the phenomenology of low-income families by examining what meanings they construct about school choice.Quantitatively, it measures what proportions of low-income families choose schools and how far the children travel to attend a non-catchment public school at the secondary-school level.Bridging these two approaches, this study's overarching mixed-methods research question is: How may the spatial dispositions and positions of low-income families influence their participation in school choice?This question is asked in an exploratory sequential research design (McMillan, 2016).It is also grounded in the tradition of critical Geographic Information Systems (CGIS), which provides an epistemological framework to consider a neighborhood as having multiple layers of meanings, demographic information, quantifiable features, and measurable distances (Elwood, 2009;Harvey, Kwan, & Pavlovskaya, 2005;Pavlovskaya, 2009).
Using a mixed-methods research design, we collected data from interviews with parents and their children at the elementary-school level while simultaneously collecting data about student school choice participation patterns at the secondary-school level.This design allowed us to maximize the scope of the research in order to understand low-income families' school choice from Kindergarten to Grade 12.Where the data converged was at the transition from elementary to secondary school.In other words, we examined how low-income parents and their children perceived choosing schools starting from Kindergarten, and then at the transition from the elementary to the secondary level, when more parents are likely to choose.
For the qualitative part of the study, we recruited 40 participants from marginalized communities and interviewed them as pairs of one parent and one child at the elementary-school level.Most of these parents were recruited at neighborhood public schools or community centers through flyers and/or snow-ball sampling.The participants' information is summarized in Table 1.
Of the families interviewed, eight of them lived in subsidized public housing, and two lived in nonprofit, co-op housing where tenants pay reduced monthly rent based on their income (government funds make up the amount that the tenant cannot pay).Also, five families lived in a house they shared with extended family members, including grandparents and other adult siblings.The adult interviewees were all mothers, except for two fathers.Half of the parents were unemployed, while the other half generally worked part-time doing low-end service work, including cleaning, beauty services, cashiering, clerical work, and social services (including graveyard shifts).Three mothers had a First Nations background, and the rest were racial minorities.All but one were immigrants from Asia, Africa, or the Middle-East, and all had functional-level English.Two mothers had university degrees, but their degrees were granted by a foreign university and were not recognized in Canada.One immigrant mother, originally from South Asia, had a university degree from the US prior to immigrating to Canada, but was a stay-at-home mom with three children.Two mothers had some college education in Canada.The rest of the parents either had a high school diploma or less than a full high school education.
All of the families in this study are classified as low-income families, according to Statistics Canada and the City of Vancouver (City of Vancouver, 2016).Half of the participants had an income of less than CAN$10,000 per year; 20% earned an annual income between CAN$11,000 and CAN$20,000; 15% earned an income between CAN$21,000 and CAN$30,000; and 15% earned an income between CAN$31,000 and CAN$40,000.(It should be noted that this last group of families with three young children, lived in either social housing or co-op housing, and their income was less than the low-income cut-off of CAN$51,272 for a family of five).
The children who were interviewed with their parents ranged in age from 6 (Grade 1) to 13 (Grade 7, the last year of secondary school); however, in this paper, we focus extensively on the interviews with parents in order to understand their school choice practices and patterns.The interview data with the parents provide an in-depth understanding of spatial dispositions that include daily routines, convenience, relationships, and feelings that underpin school choices.Their perceptions and comments provide us with a deeper understanding of factors such as academic quality, program choice, proximity, and location, which were identified in the literature review section as important to LI parents' school choice.
For the quantitative part of this study, we focused on school choice participation rates and the distances traveled by students with low-income backgrounds.The student data (2014)(2015) were obtained from the Vancouver School Board.The data set includes home addresses, assigned schools, enrolled schools, and enrolment in specialized choice programs.The distances students traveled from home to school were calculated by using the ArcGIS software program.Canadian Census data (2011) with student addresses were merged to add the demographic information of enumeration areas (which includes about 400 to 700 persons living in one or more neighboring blocks) (Statistics Canada, 2012).Several demographic variables such as education, housing, and language were used to analyze school choice patterns.We used the variable of median after-tax family income (for couples with children) to analyze marginalized urban families' school choices.
The mobility of secondary-school students was measured because at the secondary-school level, students become more independent and thus can travel farther for their schooling.We used the geoprocessing tool for road network analysis in the ArcMap software to calculate the difference in distance traveled when driving between the public school where the students were enrolled (i.e., their enrolled school) and the public school to which the students were assigned (i.e., their catchment school).This method allowed us to consider students' residential locations in relation to their enrolled schools, and thus to assess the relative distance added when they chose a school outside their catchment.Also, by measuring the distance by driving, we aimed for consistency in comparing the students.

The Local Context of School Choice: Increasing Polarization
Vancouver, the site for this study, is important for two major reasons.First, school choice has been practiced in Vancouver's public school system for an extended period of time, and is now increasingly practiced in the private sector.Yet, little is known about the extent to which low-income families are choosing schools.Second, the city is facing challenges related to increasing income disparities and a housing affordability crisis, which have resulted in gentrification within the city and low-income families becoming further marginalized spatially and economically.Therefore, it is timely to assess how increasingly marginalized low-income families in Vancouver are included in or excluded from school choice in the public system.
Vancouver implemented its school choice policy in the forms of open enrolment and district specified alternative programs starting in the 1970s.While all families living in Vancouver are guaranteed that their school-aged children have a spot in their neighborhood public school, they can also participate in the School District's open enrolment policy, which allows parents to choose cross-boundary schools if space is available in those schools.Also, families can choose to enroll their children in a French Immersion program, or choose a district choice program offered by elementary and secondary schools.
More specifically, at the elementary-school level, parents are given a form on which they can select up to three options when enrolling their children for their first year of schooling starting with Kindergarten.The enrolment in choice programs at the elementary-school level is based on space availability, or otherwise by lottery.However, at the secondary-school level, most district specified alternative programs, including mini-schools, arts programs, academies, and gifted programs, require certain academic aptitudes and well-developed talent and portfolios to gain admission.Even qualified students have to compete for spots in these programs since space is limited (Yoon, 2011(Yoon, , 2016)).There are no public subsidies or transportation options for those who choose public schools outside their catchment areas.
Outside the public school system, parents may choose homeschooling or a school in the growing private sector (Federation of Independent Schools Association, 2016).In 2016, 47 private schools operated in the city of Vancouver, 22 of which enrolled children up to Grade 7, 11 of which enrolled students in Grade 8 and above, and 14 of which enrolled students from Kindergarten all the way to Grade 12 (Federation of Independent Schools Association, 2016).Some of these schools are well-established elite private schools, some are religion-based schools, and others focus on particular learning needs, styles, and methods.The tuition fees at private schools can range from less than CAN$500 annually to over CAN$23,000 for some of the top elite private schools (Federation of Independent Schools Association, 2016).The government of British Columbia subsidizes these independent and private schools up to 50% for a full-time enrolled student assuming the school meets the criteria set by the provincial government for student eligibility, cost, and evaluation requirements (BC Ministry of Education, 2016).
Meanwhile, Vancouver has experienced widening inequality, which has disadvantaged historically, racially, and economically marginalized families even further.The inequality and polarization between the city's rich and poor have grown, with an increasing concentration of residents at the bottom and top ends of the income distribution (and a shrinking middle class) (Ley & Lynch, 2012).This inequality has had a spatial dimension.On the west side of the city, with its predominantly white, professional, and middle-class neighborhoods and growing number of affluent, elite, business-class immigrants from East Asia, residents' incomes have grown by more than 15% relative to the metropolitan average (Ley & Lynch, 2012).In contrast, in the east side neighborhoods of Vancouver, where a high proportion of residents are marginalized racial groups of Asian and Aboriginal ancestry, average incomes have fallen by more than 15%, relative to the growth of the metropolitan average income (Ley & Lynch, 2012).There has also been a notable increase in diversity: 51.2% of Vancouver residents are visible minorities (Statistics Canada, 2016).Also notable is the current racialization of poverty, with the relative incomes of recent immigrants to Canada's major metropolitan regions dropping substantially; In 2011, recent immigrants earned only about 65% of the earnings of native-born Canadians, compared to 90% in 1981 (Ley & Lynch, 2012).
In addition, Vancouver's local property market has been the most expensive in Canada (Ley, 2010), which makes it more challenging for low-income families to buy or rent a home anywhere, especially in the affluent part of the city where more popular and 'top' schools are located (Fraser Institute, 2016).Housing affordability on the west side of Vancouver is extremely low, while the schools there are in increasing demand from globally mobile middle-class and elite families (Ley, 2010).Most secondary schools on the west side of the city are over-subscribed (Author, under review), while schools on the low-income, racially marginalized east side face low enrolment and have been recommended for school closures (see Figure 1).

Figure 1. School Closures, 2016
It is within this changing and polarizing urban context that this study examines how these families, who are concentrated heavily in the east side of Vancouver, participate in school choice.Does school choice policy enable marginalized families to choose schools that are ranked 'higher' and located in affluent west side neighborhoods?How do their spatial dispositions and positions shape their school choice?

Locating 'Schools of Hope' and Spatial Dispositions
About one-third of the low-income families we interviewed chose schools other than their assigned public schools at the elementary-school level.A couple of families chose low-fee religious private schools in their neighborhoods because they desired higher academic outcomes, more traditional moral values (e.g., less emphasis on sex education and same-sex marriage), and stricter discipline in their children's education.The three Aboriginal families chose Aboriginal-focused schools that provide more culturally-relevant education for their children, even though they were not the closest neighborhood school.Similarly, families with French cultural backgrounds, especially those from former French colonies in Africa, chose a Francophone school to maintain their cultural identity and heritage.Additionally, some families chose other public schools for programs such as "free lunch" and "free piano and violin lessons."Underpinning these varying kinds of choices was one consistent theme: parents' search for a 'school of hope.'The parents considered that for their children to have a successful life and career in Canada, education was important, and that completing secondary-school education was necessary.The families, especially those who chose, emphasized the primacy of a positive schooling experience, which helps them accumulate emotional capital (Reay, 2004).In addition, they hoped their children would acquire a reasonable level of educational credentials (e.g. a high school diploma), resonating with Bourdieu's (2006Bourdieu's ( /1998) ) notion of academic capital.The low-income parents who chose thus actively sought out schools where their children could get a 'good' education and become high-achieving, echoing the literature discussed earlier, including Cooper (2007), Teske et al. (2007), and Martinez & Thomas (1994).Nonetheless, when prompted, few has searched the internet (e.g. by going to the Vancouver School Board's website) to learn about different schools.Rather, their information was based on word-of-mouth except one who considered the schools appeared on a local ethnic newspaper.
In making school choice, parents mentioned that the distance that their children had to travel to a school of choice was an important concern.If transportation had not been a problem and/or was readily available, they might have made different choices.In the absence of free school bus services, those who were able to choose a school located far away did so because the school offered free transportation or because the parents owned a car.Otherwise, they chose schools because choice schools were available close to where they lived.The families who lived in the working-class areas of downtown had more options.Downtown areas have more churches with charity missions that run schools, offering subsidies to families who struggle financially.Also, those areas have more bus routes that enable a greater choice among public schools.In comparison, those who lived in the southeast part of Vancouver mentioned that private schools were far away and expensive.The spatial locations of low-income families thus mattered to their school choice.
The transcripts further indicate that LI parents would not feel very comfortable sending their children to elite private schools or other public schools in the wealthier neighborhoods.They were concerned that their children may not fit in because they did not have expensive and brandname clothes, bags, or toys.They would thus prefer choices within the areas where they lived, both for proximity and also for the comfort that comes with being in a working-class neighborhood within the socio-economic range of where they could afford to live (Henig & MacDonald, 2002;Schneider & Buckley, 2002;Witte, 2000).
Elaine, a (part-time) working immigrant mother, was one of the parents who chose an elementary school for her children.Her choice was indeed socio-spatial because she chose a school that she considered to be "good" but not "top" because she could only afford to live in southeast Vancouver.
The schools on the top, usually they are on Vancouver's west side.Elaine's narrative echoes that of most other parents who wish to choose a school they can afford and where their children fit in economically.Parents were quite conscious of economic disparities and how it may affect their sense of belonging and their children's sense of belonging and schooling experiences in a city with growing inequality.Hence, when considering their options, they preferred a school in the areas where they felt included in a socio-economic sense.Their spatial dispositions thus shaped their school choice.
Similarly, Aboriginal families in this study mentioned that their comfort level in different neighborhoods in the city intersected with their school choice.Specifically, they were concerned that their children may experience social exclusion in neighborhoods that do not have many Aboriginal people.A sense of trepidation about choosing a school outside the neighborhood was noted in the interview with Pam, an unemployed single mother with an Aboriginal background who lived in the public housing complex.
I would feel a little bit nervous about west side…just for my son.I am a very protective mom.I don't put up with kids bullying at all.I would be scared for him to be bullied, or, you know, because it's west side.There are not too many natives out there.Mostly Asian or White, so no… I would lose it… Which is sad, because I should feel good about living anywhere, so should he, but because of the color of our skin… as native people, because bad apples out there, making native people stupid.They think all native people are like that.We are not all like that… On one hand, parent accounts indicate that their spatial dispositions, which consist of spatially-formed, everyday routines, relationships, feelings, and senses of belonging, were key to locating their schools of hope for their children's present schooling and future opportunities in a city with increasing income disparity, post-colonial racial tensions, and growing diversity.On the other hand, their sense of fitting in -having a school community where they felt they belonged -was important.Parents' narratives indicated that a neighborhood is not just a Census boundary or physical boundary; it generates "circuits of belonging, attachment, and a strong sense of 'people like us,'" thus, shaping people's perceptions of who belongs where (Wilkins, 2011, cited in Gabay-Egozi, 2016, p. 17;Yoon & Gulson, 2010).

Spatial Positions and Housing Options
An increasingly polarizing city constrains low-income families' school choice because their housing options are limited.Most parents said that housing decisions came first, while they gave some consideration to school options when making choices about housing.For families who are renters, monthly payments were the primary concern.Other families could not afford their own homes so they moved in with their extended family members, and their children's schools were chosen based on where their family owned or rented.For those who lived in subsidized public housing, their housing options were determined by availability, meaning they had scarcely any choice at all.Housing options and constraints thus determined the low-income families' spatial positions and their choices of schools in the city.
Keith, an immigrant single father from a Southeast Asian country, moved to Vancouver from Toronto a few years ago.At the time of the interview, he was unemployed, so rent was key concern for his choice of residence and thus school.
I started looking for a place in newspaper, Vietnamese magazine, and found this place.Thought if this place is not good, move again, but I like this place, so I stayed… I looked at the paper: how much is for one month rent?Cheap, so okay.I come and looked around.It's clean, so I rent… We need a place to live first, so look for other things, school and jobs… I chose a school close to home.My son could walk from home to school.Car can make him sick.Busing can make him sick.So, it's good for his exercise.
For Keith, the only reason he would have chosen another school was if there was violence in the current school and his son was involved with it.
A quarter of the study's participants lived with their extended family, and they chose a school where they lived, guided by their family members who had more information about the neighborhood public school.They mentioned that their families were able to provide good feedback on the neighborhoods where they lived, so they were happy to choose the public schools in their catchment areas.Also, for some parents, moving in with their mother or father meant that they were going back to the neighborhood where they grew up, so they were already familiar with the school, and they were content with choosing that school.
Some parents who lived in subsidized public housing felt that they could not reach what they perceived to be their schools of hope because those schools were too far away.Pam, the abovementioned mother with the Aboriginal background, considered using her friend's home address to enroll her child in a public school with a specialized arts program; however, she felt the school was too far away, so she did not pursue that option.
Another mother, Hana, who came to Canada as a refugee from the Middle East, said that she was feeling quite unsafe living in the subsidized public housing complex because of the presence of drug addicts nearby and the racial attacks that appeared on her door (eggs thrown and "racist" graffiti), so she hopes to move to another location; however, she did not have any choice.Nonetheless, she has applied for another subsidized public housing unit in a 'better' neighborhood and with a 'better' school for her children: I am scared for my kids.We don't go out much.Anytime not safe here.After 4:30 [PM].I can see it from here, man and woman [doing] drugs, cocaine.I want to move to near Van Tech.We asked BC Housing [for a new subsidized public housing space], and been waiting for four years… I like [the neighborhood of] Skeena Terrace.…Because it's quiet there, there are no drug addicts like here.Also, I have my Arabic friends over there.Sometimes when I get sick, my friends can come and help.
Some of these families experienced a feeling of being "trapped" because of a lack of housing options in the city for low-or no-income families, especially in the neighborhoods they considered to be safe and quiet (Tate, 2012).Accordingly, their spatial positions limited their options of school choice since they cannot move to the neighborhoods where they could benefit from the profits of localization, that is where there are choices of public schools that are enriched, respected, and popular (Reay & Lucey, 2003).

Social Benefits of Localization
Despite the challenges, dangers, and risks that are associated with living in low-income neighborhoods, including constrained school choice, most low-income parents mentioned the importance of institutional resources and social benefits in their neighborhoods, which strengthened their sense of belonging and helped them develop strong social networks there.Most of the families from subsidized public housing also noted the significance of the social and institutional resources that were readily available to them in their neighborhoods as being critical to raising their children.In the areas with considerable drug addiction problems, street violence, and prostitution, the neighborhood school offered a 'walk to school' program which made them feel assured that their children would be safe.These options were better for them than sending their children by bus (on their own) to other schools in different or affluent neighborhoods.
Parents also identified the benefits of having free after-school programs or spaces for their children at the neighborhood community center.They felt comfortable that their children were going to these centers after school where they would be safe and learn new activities, while also benefitting from extra tutoring programs and free childcare.A variety of sports and arts activities in community centers in their neighborhoods were offered for reduced fees.A community center in the downtown area also offered a free pick-up program so they could bring children from school to the community center for after-school programs, and parents could pick their kids up later.The lowincome families mentioned that they benefited from such social and welfare programs that were available in their neighborhoods.
For instance, Elsa, an immigrant mother who came from southern Africa, resided in subsidized public housing.She was working full-time (the graveyard shift) after obtaining a college degree in Canada.She summarized the supportive programs that were available in her neighborhood as the following: My neighborhood is great because we have all kinds of social services we need.Most low-income parents were not sure if similar programs would be available to them if they moved to wealthier parts of the city.Hence, they benefited from the localization of social benefits (rather than private profits, in market terms) in their low-income neighborhoods.
In addition, low-income parents mentioned the importance of having their friends and relatives nearby.The families who moved in with their family members noted that their extended families provided them emotional support and extra childcare support.Jocelyn, a single mother with an East Indian background, said that although she wished she could choose academically high achieving schools elsewhere, she ended up choosing a school in the neighborhood where she could share a duplex with her grown-up siblings and their children because her family members could support each other in various ways: My parents are here.As parents develop strong networks in their neighborhoods, feeling safe is spatial and being far away means feeling less secure.Like Jocelyn, other parents mentioned that they would be worried about their children attending schools so far away from their neighborhoods and not knowing what is 'there' in those neighborhoods.
The interviews with marginalized urban families reveal a certain level of benefits, spatialized social capital, and a sense of safety, belonging, and resources in their spatial positions, all of which help with the upbringing and education of their children.Their neighborhood was a fertile social space for developing social networks, belonging, dispositions, and habitus, and the various types of in-kind, non-monetary, and hard-to-measure social and institutional resources that were available nearby account for why a majority of marginalized urban families chose a school in their neighborhoods.

Spatially Constrained Secondary-school Choice
When asked about choosing secondary schools, the theme of finding a school of hope continued, but most parents also mentioned that their children would become more independent when they entered secondary school and would be able to travel farther to attend a non-catchment school.Almost all parents mentioned they would respect their children's choice and would ask their children what schools they wanted to attend.This echoes the findings of other studies of workingclass family decision-making dynamics of school choice, especially in England (Reay & Ball, 1998;Taylor, 2002).Nonetheless, when compared with other income groups, the school choice patterns of students from low-income backgrounds were the most limited.
Using the student data, we examined the spatial distances traveled by those with low-income backgrounds, compared to the distances traveled by children with middle-and high-income backgrounds.Overall, students from low-income neighborhoods had three distinctive patterns of participation in school choice.(1) For the families who chose, the distances they traveled were shorter than those from higher income groups.(2) Marginalized families were less likely to participate in school choice.(3) Those who chose were more likely to leave their school catchment areas to attend another regular public school rather than a selective program (French Immersion or District Specialized Alternative Programs, including mini schools).Table 2 compares school choice participation patterns by income, using data from the Vancouver School Board and Statistics Canada.In terms of school choice, those who lived in the lowest-income neighborhoods (with after-tax annual income less than CAN$25,000) had the lowest participation rate.As noted in Column 3 in Table 1, the low-income families' participation rate was 33%, compared to 33.4% of low-middle-income families, 35.4% of middle-income families, 38.3% of middle-high-income families, and 37.7% of high-income families, indicating a small but largely consistent pattern.Next, in terms of the kinds of choices families made (see Column 3), students from low-income neighborhoods were more likely to choose another public school in the mainstream English program than specialized choice programs that have academic admission criteria.The highest percentage of students choosing such schools were from low-income neighborhoods, at 36.8%, compared to their counterparts: 29.4% (low-middle), 30.6% (middle), 31% (middle-high), and 28% (high).In contrast, for the specialized public school programs of choice, such as mini schools or French immersion, which have selective and competitive admission processes and criteria (see Column 4), students from low-income neighborhoods had the lowest participation rate, at 20%, compared to 59.3% (low-middle), 59.3% (middle), 65.8% (middle-high), and 67.7 % (high).
In terms of distance traveled (see Column 1), students from low-income neighborhoods traveled the shortest difference, about 2 kilometers (km), to attend their schools of choice.In comparison, the difference of distance traveled by other income groups was farther: 2.6 km (lowmiddle), 2.9 km (middle), 2.7 km (middle-high), and 2.3 km (high).This socio-spatial analysis indicates that although parents imagine that their children can attend a school that is farther away when they reach secondary school, overall, the city-wide pattern is that students from low-income families do not travel as far as those from families in other income categories.Their spatialized social (peer group) network, as shaped by their residential locations, financial capital, and sense of belonging, seems to limit the parameters of what they can choose and where they can travel to attend secondary school.
The choice of housing, which is based heavily on housing market inequality and economic polarization, thus appears to have important consequences for school choice.Parents' housing decisions, which are determined primarily by economic and social capital and perceptions of where they belong, further shape their children's social networks and secondary-school choice.As such, spatial position and disposition come together when low-income families consider schools as social, physical, and phenomenological spaces.As the income gap widens, income groups are more likely to live further away from each other, physically and mentally, with implications for where their children will be schooled.Parents imagine particular schools through their dispositions (habitus), which are shaped by their spatial locations in a broadening spatial hierarchy as families seek their schools of hope.

Conclusion
In the context of increasing privatization and marketization of education reforms, this study began with the aim of deepening our understanding of the K-12 school choice practices and patterns of marginalized urban families in one of the most rapidly diversifying and polarizing cities in the world, Vancouver, Canada.In particular, this study has focused on the spatial positions and dispositions of marginalized low-income families, one of the most under-represented groups in the current literature, in understanding their school choice.In doing so, we have drawn from Bourdieu's sociological lens and critical human geography while applying a mixed-methods research approach.Our approach was chosen to help us examine the spatial underpinnings of school choice (including dispositions, relationships, and locations) among low-income families rather than to survey what factors are considered or prioritized among those families.
The pattern found in this study corroborates the findings of the existing literature on lowincome (LI) families in that a majority of LI parents choose for reasons that are similar to those of middle-class families, such as school quality, climate, and safety; nonetheless, they end up choosing schools close to where they live for proximity and comfort, and thus most LI families end up choosing a school where a majority of the students are from their own class and/or racial group (Henig, 1996;Holme, 2002;Kotok, Frankenberg, Schafft, Mann, & Fuller, 2015).In other words, schools are not simply an abstract space of learning and teaching, but a spatialized social space where students and families come to "recognize their own space in relation to others" (Bourdieu, 1989, p. 16).In addition, parents have identified the social and institutional benefits of living in their neighborhoods and choosing schools in those neighborhoods, even though their school choice options are limited and they face urban 'dangers.'Overall, low-income families' participation in school choice is lower than higher income groups, the distance their children travel is shorter, and their choice tends to be limited to cross-boundary rather than competitive/selective choice programs.
This study complements existing studies on mobile, professional, and elite urban families' school choice practices, which they pursue for social advantage and distinction in an aspiring global city (Ley, 2010;Waters, 2006;Yoon, 2011Yoon, , 2016) ) by expanding our understanding of current practices and patterns of school choice among low-income families.Indeed, this study illuminates some of the areas of school choice policy that can be modified to increase the inclusivity of lowincome families.The findings suggest the need for more provision of transportation, social support in a school of choice, and/or greater options available in low-income areas.Indeed, this study raises questions about declining government support for social programs and educational programs while providing more choice programs that seem to be popular among high-income families.Low-income families indicate that choosing schools far away from one's home location is expensive, timeconsuming, unfamiliar, and unsafe.As such, low-income families indicate a preference for having more programs available to support their children's educational success in the neighborhoods where they (can afford to) live.A neoliberal approach to educational funding and organization thus seems short-sighted while neglecting the fundamental needs and rights of low-income families to support their children's education in areas where they feel safe.
Indeed, critical reviews of school choice have found that the privatization and marketization of education neither generates greater levels of efficiency in the education system nor improves accountability (Forsey et al., 2008;Lubienski & Yoon, 2015).This study raises additional concerns about the broader trend of growing income inequality in cities.As income inequality grows, different social groups will feel greater disparities and thus experience greater distance.In this context, school choice is more likely to reproduce inequalities and divisions among social and racial groups as spatially organized.Hence, we recommend that governments take a more fundamental and comprehensive policy approach that can generate greater equality and help ameliorate gaps in school resources and educational outcomes.

Table 1
Summary of Parents' Characteristics (Number of participants) They are private schools.Public schools are usually in the middle, middle-top, middle-low, or low level.I chose [a school in] the middle….The schools are on the west side are Parents drive famous cars to pick up [their kids].My friend has a used Japanese car, and her daughter said, 'You don't need to pick me up.If you are going to pick me up, pick me up in the next block.'My friend got really angry… I don't think it's really healthy for the children when they are growing up.
expensive.My friend said people in that school, students like to compare.[My friend is] a new immigrant mother, moved there [west side], pay higher rent.Drive a Japanese car, a second-hand Japanese car.That school, most of the people are rich.
When I moved here, I was new immigrant.I didn't know anybody.Most people I met first were Church people.They were afraid of Downtown Eastside because they know it as a drug user area.But where there are more troubles, there are more services.So my neighborhood is great for me.Childcare, [in] Strathcona, they have an emergency daycare.Free.Nowhere else in the Lower Mainland [i.e., Greater Vancouver Area], you can find a free daycare.Nowhere.When I moved here, and when I need a break, I call and ask.If they have space, they take care of my babies.Also, Ray-Cam Community Centre, it's a common space for kids.Kids come to socialize here.They don't have to go find friends in the housing.You don't know about kids' parents.You don't know about them, you don't feel safe.But, when kids meet here, it's safe here.It's just like home.
My [ex-]husband's parents are here.They could walk.While I go to work [weekends or evenings at a local grocery store], they can also take my kids to school.Grandma, she can help… Also for my older son, I didn't want him to go anywhere else [a non-catchment school]… I wouldn't have my nephew there.I wouldn't have that kind of support.They can go to school together now.My nephew is older than him.He's always watching him.He's a better example for my son.A role model.My nephew guides my son.'I see, you are hanging out with this student….he'snot good.'He also keeps me updated.

Table 2
The difference in distance traveled by students between the public school where the students were enrolled (i.e., their enrolled school) and the public school to which the students were assigned (i.e., their catchment school).** Regular program refers to those who cross their catchment boundaries to attend another public school.*** Students attend another public school for a selective alternative program, including mini schools and French Immersion. *