‘Notorious Schools’ in ‘Notorious Places’? Exploring the Connectedness of Urban and Educational Segregation

While the statistical link between residential and school segregation is well-demonstrated, in-depth knowledge of the processes or mediating mechanisms which affect the interconnectedness of the two phenomena is still limited. By focusing on well-functioning schools in disadvantaged neighbourhoods, our article seeks to scrutinise whether reputation can be one of the key mediators of the connection between residential and school segregation. Our study combines qualitative ethnographic interviews from four (pre-)primary schools with quantitative segregation measures in four urban neighbourhoods in the Finnish capital city of Helsinki to understand the connections between lived experiences and socio-spatial segregation. The results show that there appears to be a clear link between neighbourhood and school reputation, as schools in disadvantaged neighbourhoods are strongly viewed through the perceptions attached to the place. Despite the case schools’ excellent institutional quality and high overall performance in educational outcomes, there is a consistent pattern of the schools struggling with negative views about the neighbourhoods, which seep into the schools’ reputation. Since school reputation is one of the central drivers of school choices and is also linked to residential choices, the close connection between neighbourhood and school reputation may feed into vicious circles of segregation operating through schools. The results highlight the need for integrated urban policies that are sensitive to issues concerning school reputation and support the confidence and identity of pupils, reaching beyond simply ensuring the institutional quality of schools. article Circle of Segregation: Understanding the Connectedness of Spatial Inequality across Generations and


Introduction
In many European cities, residential segregation is on the rise (Andersen, 2019;Tammaru, Marcińczak, van Ham, & Musterd, 2016). At the same time, educational equality is facing a challenge as school segregation and widening gaps in educational outcomes have become marked in many contexts (Boterman, Musterd, Pacchi, & Ranci, 2019). Segregation in the residential and edu-cational life contexts, or domains, is tightly interlinked. While increasing segregation on the neighbourhood level feeds into the growing differentiation of student composition between schools (Bernelius, 2013;Bernelius & Vaattovaara, 2016;Boterman, 2019), the residential mobility behaviour of young family households is increasingly informed by school choice considerations (Bernelius & Vilkama, 2019;Hamnett & Butler, 2013). Neighbourhoods with more popular schools attract more middle-class residents, while some neighbourhoods are rejected partly because of concerns related to schools. However, while the statistical link between residential and school segregation is well demonstrated Frankenberg, 2013), in-depth knowledge of the underlying processes or mediating mechanisms which affect the interconnectedness of the two phenomena is still limited.
By focusing on well-functioning schools in disadvantaged neighbourhoods, our article seeks to scrutinise whether reputation can be a crucial mediator of the connection between residential and school segregation, feeding into a multi-domain vicious circle of segregation (van Ham, Tammaru, & Janssen, 2018). The main aim of our study is to explore the interconnections between urban segregation and the reputations of schools and neighbourhoods. We ask how perceptions of local schools are linked to urban segregation and problems of neighbourhood stigmatisation. To what extent do these perceptions relate to objective characteristics of schools and neighbourhoods and how are they experienced by students, parents and teachers? In short, if disadvantaged places are labelled 'notorious' (Kearns, Kearns, & Lawson, 2013), will schools also be seen as such?
The study combines qualitative ethnographic interviews from four (pre-)primary schools with quantitative segregation measures in four urban neighbourhoods in the Finnish capital Helsinki. In contrast to previous research on school reputation, the combination of quantitative and qualitative data allows us to contrast pupils,' parents', and school staffs' subjective perceptions of school and neighbourhood reputations with objective, quantitative segregation measures. Through the constant dialogue between the two datasets, the everyday experiences in schools and neighbourhoods can thus be contextualised in place.
Our conceptual framework mainly draws on the concepts of reputation, image, and stigma elaborated by Kearns et al. (2013) in their study on 'notorious' places in the UK. We use these concepts to interpret how pupils, families, and school staff express their views and experiences of their neighbourhoods and schools and position themselves concerning their own communities and outsiders. Particularly, the differentiation between personal beliefs and meta-beliefs, referred to as reputation, allows for a better understanding of how values are attached to both neighbourhoods and schools and how these values are socially reproduced.
Helsinki is an ideal location to study the relationship between neighbourhood and school characteristics. As the local educational system consists mainly of local public schools with individual catchment areas, the interconnections between neighbourhood and school allocation are very strong. Public funding and a shared curriculum make institutional variation between schools low in international comparison. Based on egalitarian ideals, the municipal educational authority seeks to ensure equal academic institutional quality in all schools and ranking lists are not published. It can therefore be argued that school reputation is less dependent on institutional variation than in more differentiated education systems which, for instance, rely more on private schools with strong school competition and varying institutional qualities. This local context thus allows revealing effects that are almost exclusively tied to the social (re)production of reputation through the composition of both schools and neighbourhoods. Since rumours and reputations play a significant role for parents' school choices even in education systems with official league tables and in which the variation between the institutional quality of schools is higher (Butler & Hamnett, 2007;Vincent, Braun, & Ball, 2010), this study provides an understanding that can likely be transferred to systems with more institutional variation.
So far, several studies have demonstrated that school segregation is to a large extent the effect of residential patterns (Boterman, 2019;Frankenberg, 2013). However, the former is not just a simple reflection of the latter; their connection is rather exacerbated by several different processes (Candipan, 2019;Oberti & Savina, 2019). Since the relationship between residential and school segregation is crucial to understand intergenerational social mobility and inequality , the topic is of high educational and socio-political relevance. Based on a conceptual design combining both quantitative and qualitative empirical data, our study allows us to gain a better theoretical understanding of the mechanisms by which urban segregation affects the widening gaps in educational attainment between urban communities. Identifying these mechanisms is central to finding novel ways to support schools and communities in urban neighbourhoods throughout Europe.

Local Geographies of Education: The Close Relationship between Residential and School Segregation
Research across many countries illustrates that residential segregation and school segregation are tightly interlinked in a 'geography of education' (Butler & Hamnett, 2007). However, while high levels of residential segregation are usually accompanied by segregated schools, low levels of residential segregation do not necessarily result in mixed schools. In contrast, school segregation is usually higher than residential segregation, which is mainly due to parents' socially selective ways of choosing schools Ramos Lobato & Groos, 2019;Wilson & Bridge, 2019).
Access to high-quality education has become a sensitive topic for many parents (Butler & Hamnett, 2007). Especially middle-class parents, equipped with the social and cultural capital needed to take full advantage of the educational market, actively navigate the system to find the 'right' schools Kosunen, 2014). Concerned about their children's exposure to lower standards of education, to children with inadequate language skills or the 'wrong' types of socialisation, many parents tend to define the 'right' school based on its social, racial or ethnic composition-which feeds into growing school polarisation (Boterman, 2013;Ramos Lobato & Groos, 2019;Vowden, 2012;Wilson & Bridge, 2019).
Parents' school choices are strongly influenced by their local social networks (Ramos Lobato, 2019;van Zanten, 2013;Vincent et al., 2010;Vowden, 2012). Personal impressions and the experiences of friends or relatives-so-called 'grapevine-knowledge' (Ball & Vincent, 1998)-are often used to compensate for missing or untrustworthy official information. Since parents' networks spread (middle-class) parents' personal opinions about 'good' or 'bad' schools rather than provide objective information, they do not only transfer but actively construct reputations of 'good' schools (Holme, 2002;Kosunen, 2014). As grapevine knowledge also provides a medium for social comparison (Ball & Vincent, 1998), schools' reputations play a significant role in shaping parents' choice strategies.

Conceptualisation of Image, Reputation, and Stigma
Based on the conceptualisations of image, reputation, and stigma by Kearns et al. (2013), we define image as a personal belief or evaluation about schools and neighbourhoods. Reputation, in contrast, is a meta-belief; a belief about what is commonly believed about a particular object, which does not necessarily reflect the speaker's own view (Kearns et al., 2013). The way in which personal opinions are transformed into socially accepted perception or reputation is based on two mechanisms: First, through institutional actors, who can spread their views publicly-such as local newspapers (Butler, Schafran, & Carpenter, 2008;Kearns et al., 2013;Permentier, van Ham, & Bolt, 2008) and second, unintentionally, through the residents themselves (Butler et al., 2018;Pinkster & Hoekstra, 2020).
While both image and reputation can be positive or negative, stigma carries solely negative connotations. According to Goffman (1963), stigma is the classification-and the subsequent discrimination, exclusion, rejection and devaluation-of individuals as 'discredited' based on the possession of symbolic and/or physical attributes. The symbolic dimension of stigma is especially crucial since it emphasises the stigma's "structural roots in broader patterns of power and its role in legitimising social inequality in society" (Kearns et al., 2013, p. 582). As stigma is understood as an intrinsic part of the stigmatised individual-even though it is just attached to a person by others-those who have been stigmatised have less power to change the stigma but rather tend to make it a part of their own identity (Bunar, 2011;Kearns et al., 2013).
With his concept of territorial stigmatisation, Wacquant (2007) adds place as an additional and partially autonomised dimension of social discredit. Territorial stigmatisation affects not only the residents but also the level and quality of service delivery, the area's symbolic representation by journalists, and scholars, and the beliefs and decisions of state officials and their public policies (Wacquant, Slater, & Pereira, 2014).

Notorious Schools in Notorious Neighbourhoods?
In the case of a bad reputation, most residents or pupils are aware of their neighbourhood's or school's negative reputation, the stereotypes associated with it, and the position of such schools and neighbourhoods in the local hierarchy (Hollingworth & Archer, 2010;Kearns et al., 2013;Kosunen & Carrasco, 2016). A noticeable body of research shows that objective characteristics seem to be good predictors for both neighbourhood and school reputation Kearns et al., 2013;Kosunen, 2014;Permentier et al., 2008)-with schools' educational quality as one important exception to the rule. As parents tend to associate a school's educational quality with its social, racial or ethnic composition rather than its institutional quality (Boterman, 2013;Vowden, 2012), there are schools with bad reputations despite parents' positive experiences, and their good performance (Bernelius, 2013;Kosunen, 2014). To understand the source of this bad reputation, it is thus not sufficient to pay attention to objective or institutional characteristics, but rather to the schools' interrelations with the neighbourhood and their position in the social and symbolic local hierarchy (Bunar, 2011).
So far, there is only limited information about the social processes by which school and neighbourhood reputations are discursively constructed. Certain schools seem to become stigmatised via a complex interconnection of material conditions, educational outcomes (e.g., league tables) and neighbourhood reputation (Bunar, 2011), although the impact of the latter does not seem to be so straightforward (Hollingworth & Archer, 2010). Schools' reputations seem to be connected to their relative position in the local educational hierarchy (Bunar, 2011;Kosunen, 2014); however, how this hierarchy is connected to socio-spatial characteristics has not yet been studied.
In numerous cities, residential and school segregation are two strongly interrelated phenomena (Bonal, Zancajo, & Scandurra, 2019;Boterman, 2019;Frankenberg, 2013;Oberti & Savina, 2019). Regarding the underlying mechanisms of this relationship, we ask whether reputation might be one of the key elements to understanding this strong connection. So far, there are no previous studies in the Finnish and international context where school and neighbourhood reputations have been studied together, or directly connected to quantitative measures of neighbourhood segregation in a single research setting. In this article, we thus seek to explore reputation and stigma as a potential link between urban and school segregation, feeding into vicious circles of segregation. We ask to what extent neighbourhoods' and schools' reputations are related to objective characteristics and how they are experienced by students, parents, and teachers. Our mixed-methods approach allows us to analyse both how subjective values are attached to neighbourhoods and schools and how these are connected to quantitative measures of neighbourhood segregation and thus fill the knowledge gap in research.

Methodology
To capture the interrelationship between neighbourhood and school segregation and reputations, we combine two sets of data in a common analytical framework: quantitative GIS-data to analyse the socio-spatial structure of the school catchment areas, and qualitative ethnographic interview data from two related research projects: "Well-Functioning Local Schools" (2014)(2015) and the "Mixed Classes and Pedagogical Solutions MAPS'' (2018)(2019)(2020)(2021).
The quantitative data consist of Statistics Finland Grid Database (250 m grid-cell data) on block-level socioeconomic indicators for the years 1999-2019 with additional information on registered languages of residents for 2012. In the quantitative analysis, we modelled sociospatial segregation in school catchment areas by aggregating block-level urban statistical data into the catchment area level, producing a segregation analysis of all primary school catchment areas in Helsinki. Our analysis extended over several years to check for consistency in the spatial development trajectories. The analysis software was MapInfo and QGIS, combined with SPSS for statistical analysis.
The case schools were selected based on the catchment area segregation analysis, school characteristics, and educational outcomes from 2012. The educational outcomes assessments have been carried out by the National Board of Education, and the institutional academic quality of the schools by the Helsinki City Education Council with the criteria of well-functioning school leadership, high teacher satisfaction, and low staff turnover. We selected schools, which are located in mixed or disadvantaged neighbourhoods but have been assessed to achieve good educational outcomes and to be of excellent academic quality. As previous research has demonstrated (Bernelius, 2013), the schools' educational outcomes are usually highly correlated with the socio-economic status of the catchment area in Helsinki. We searched for schools, which perform exceptionally well and exceed the outcomes which would have been statistically expected based on the catchment area's socio-economic composition. The final selection criteria were (1) a high level of local and relational socio-economic disadvantage in the school catchment area and, simultaneously, (2) educational outcomes that exceeded the level statistically associated with the quantitative measures of local disadvantage.
The qualitative data consist of ethnographic interviews that were conducted in four selected neighbourhoods and pre-/primary schools. Ethnographic interviews mean that they were conducted in projects in which the relationships between the researcher and the interviewees were established during longer observation periods within the schools. Thus, the duration and frequency of contacts with the interviewees distinguish them from interviews that are set up only for that purpose (Heyl, 2007). The overall qualitative ethnographic data were collected in the two research projects and include both field notes from observations in schools and ethnographic interviews (n = 125) with pupils, their parents, and schools' staff. For this study, we limited the analysis to the latter, in which the topic of reputation was explicitly dealt with.
In all schools, we interviewed the staff volunteering to participate (n = 47) during the observation periods. This included teachers, school leaders, and other professional personnel. The interviewed pupils (n = 51) were fifth-and sixth-graders (11-13 years of age). In Finland, children start their obligatory educational paths when they enter pre-primary school at age five or six. Pre-primary education in this study was organised on the same premises and in close co-operation with primary school. Since the interviewed pupils were about to enter lower-secondary education, which starts at age twelve or thirteen, we discussed their experiences of both primary school and the transition phase. In Helsinki, pupils are mainly allocated to their nearest lower-secondary school; however, they can apply to other lower-secondary schools. The group of parents (n = 27) included the parents of the interviewed pupils and several pre-primary school pupils, whom we contacted at parents' evenings. All interviews lasted between 30 and 90 minutes and were recorded and transcribed. All Finnish quotes in this article were translated into English.
In the interviews, we talked about the interviewees' perception and experiences of the schools and neighbourhoods. In the analysis, we first utilised coding in Atlas.ti software as a means to organise the extensive dataset and then moved on to inductive thematic content analysis (Schreier, 2012). At first, we coded talk about school(s) and neighbourhood(s) concentrating particularly on how they were described in the interviews and in relation to other schools and neighbourhoods. Next, we coded the data excerpts by using 'reputation,' 'image' and 'stigma' as codes generated deductively from our conceptual framework. Afterwards, we moved on to the inductive thematic content analysis to capture the formulations used by the interviewees themselves. To differentiate between reputation and stigma, we understand stigma as a negative reputation that has already been internalised by an individual, which points to the underlying unequal power relations and their structural roots (Bunar, 2011;Kearns et al., 2013).
The quantitative analysis was combined with the qualitative analysis in two ways: First, we used the quantitative data to analyse socio-spatial patterns in the school catchment areas to find suitable areas and schools for case studies. Secondly, we contrasted and contextualised the qualitative findings with the segregation patterns to understand the relationship between socio-spatially structured segregation and the individual interpretations of places and their images and reputations. The research findings are thus based on a mutually complementary dialogue between the structural quantitative analysis and ethnographic interview data to examine, for example, the importance of a school's or neighbourhood's relative position in the city to produce a certain reputation.

Geography of the Case Study Schools
Our four case schools are called here by their pseudonyms Thyme, Caraway, Pimento, and Rosemary. They have all been assessed by the education authorities to have excellent school leadership, low teacher turnover, high parental satisfaction, and good educational outcomes. Based on these institutional factors and educational performance, school reputation should not be negatively biased by any characteristics related to the institutional quality of the schools.
Caraway, Pimento, and Rosemary are all located in the larger district of East Helsinki. East Helsinki has a strong, rather stigmatised reputation as the 'notorious' part of the city, where most neighbourhoods are clearly more disadvantaged than the city average. In the national media, East Helsinki has become almost synonymous to urban disadvantage and segregation, although there is internal variance in the socio-economic status of the different neighbourhoods in the eastern parts of the city.
According to their catchment area characteristics, Caraway, Pimento, and Rosemary share a distinct disadvantage concerning the city averages in income, unemployment, and share of residents with Master's-level education (Figure 1). The catchment areas are also among the ones with the highest local share of residents with a foreign mother tongue. Other available socio-economic indicators demonstrate the same disadvantaged status: The share of adults with only basic education is distinctly higher in the selected areas than in the city in general, and cramped housing conditions are more common. According to the longitudinal observations of all catchment areas, the neighbourhoods' relative disadvantage has deep roots. While segregation has increased between the catchment areas from 1999 to 2019, the relative position of these catchment areas has remained in the lowest quartile of the city. Previous studies in Helsinki have highlighted the risk of vicious circles of segregation in these types of catchment areas, as many of them are avoided in residential decisions or they experience a migration loss of middle-class families (Bernelius, 2013;Bernelius & Vilkama, 2019).
In contrast, Thyme's catchment area is close to the city average by all its socio-economic indicators (see Figure 1). However, while the other schools are all surrounded by catchment areas that are relatively similar to each other, the Thyme catchment area is located relatively close to the city centre, between well-off areas in Helsinki. Through Thyme, it is thus possible to explore the meaning of relative local disadvantage and the effect of local hierarchies on school and neighbourhood reputation.

Place-Based Stigma and Real-Life Consequences
Caraway, Pimento, and Rosemary are all located in East Helsinki. While the socio-spatial characteristics of their catchment areas show quite similar patterns of socioeconomic deprivation and higher-than-average shares of ethnic minorities, the interviews, in contrast, illustrate multidimensional aspects affecting images, reputations, and stigmas (Kearns et al., 2013). Among them, Caraway differs regarding its shared positive self-image that covers both the neighbourhood and the school. School and neighbourhood images are connected strongly, and this connection is intentionally maintained. The school staff describes the school in a consistently positive way. Both staff and parents are proud of the school's place-related roots and traditions, which are used in the identity building of pupils as the following teacher's quote illustrates: "You can see Caraway in these children. It is a certain pride for many that they are from Caraway; you certainly don't have to hide it." One aspect that is important to mention is that many of the staff members live in the neighbourhood as well. Along with the parents, they applaud the diversity and social mix of the neighbourhood and the school. Pupils in Caraway describe their neighbourhood similarly positively; most of them can even imagine staying there as adults.
The other two East Helsinki schools, Rosemary and Pimento, have good images as well; however, they need to maintain them despite the neighbourhoods, which are mostly talked about in a negative way. Even though Caraway's school catchment area characteristics are like those of Rosemary's or Pimento's, the relationship between the school and the neighbourhood differs significantly. In Rosemary and Pimento, the interviewed staff, none of whom live in the neighbourhoods, describes them through social problems. This talk relates to the existence of "problematic" places in the neighbourhoods, including local public transport stations, which generally tend to gather problematic phenomena in Helsinki, such as substance abuse. Similar patterns can be found in pupils' interviews, which mention difficulties in finding positive comments about their neighbourhoods and rather describe incidents with intoxicated adults, even harassment, as one of the sixth graders illustrates: I'll tell you a story. When my sister was in Pimento…there was this woman she didn't know…she said to [my sister] that she will burn her hair and kill her when she sees her the next time….The woman chased her, and our dad called the police.…I'm afraid to go to Pimento nowadays.
Even though these problems are not related at all to the quality of the schools, they seem to directly seep into them and consequently, to impact the schools' reputation. How strongly doubts about the quality of schools are shaped by the neighbourhoods' overall bad reputation is illustrated by the following quote. When asked about the school's reputation, one parent explicitly states: "Mostly that it is located [in Rosemary's neighbourhood], and then people already start thinking whether it's a good school." Hence, a reputation as a bad place is powerful enough to socially construct schools as notorious regardless of their actual quality (see also Bunar, 2011). In order to prove themselves to be better than outsiders' expectations, school reputations are often deliberately constructed and maintained against other schools within the area. Because of these local comparisons, it may be difficult for all schools in these neighbourhoods to be perceived as 'good', because the problems that are attached to the neighbourhoods' stigma need to be located somewhere in the local discourse. Consequently, the socially constructed relative positions of schools in the local educational hierarchies become important, which we will refer to in more detail.
While staff members talk a lot about problems, they also made clear that they believe they are doing a good job, that the schools' everyday life functions well and the atmosphere is said to be better than in some schools with an "easier" pupil composition. The staff's perception or image of the schools is an overall positive onedespite the difficulties they may experience due to the neighbourhoods, as the next quote illustrates: Interviewer: How would you describe this place to someone who doesn't know it?
Teacher: Nice people, everything works well, not at all like, like I had the impression, of course…when I came here to Pimento school what it must be like, but the image is much more positive now.
The quote illustrates how image and reputation differ. This teacher heard about the Pimento neighbourhood's reputation before s/he first entered the school and therefore had doubts about the school's quality. Nevertheless, s/he ended up working there and now perceives the school from a different viewpoint.
The 'notoriousness' of East Helsinki creates another layer in the place-based problems of the schools. Not even Caraway, which manages to positively connect the school and the neighbourhood, can ignore the stigma attached to East Helsinki. 'East Helsinki' as a term is referred to several times in the interviews. However, 'East Helsinki' does not only or mainly refer to a certain area or place; rather, it is used as an attribute describing something challenging either in the schools or in the schools' reputations.
The stigmatised position of East Helsinki and thus the schools located in it becomes obvious in the interviews when the interviewees themselves use this stigmatisation as a self-explanatory concept when they describe (potential) problems. This is visible in the following teacher's quote where s/he describes teaching: "And it's probably more challenging and more difficult and takes more time in an East Helsinki school than somewhere else." In the quote, East Helsinki is used as a synonym for a challenging school and potential failures. The interviewee even treats this as common sense so that the interviewer is expected to understand why something is "more difficult" in an "East Helsinki school." This self-explanatory concept is internalised as the teachers believe themselves to be in challenging circumstances due to the school's location. This stigmatised position also becomes visible in parents' frustrated comments about being tired of the talk related to East Helsinki: Interviewer: Were [things that outsiders find suspicious] suspicious to you? Parent: No. I think, East Helsinki is pretty peaceful, even if many say otherwise, but it's [fine].
Since all schools in this sample are doing well in terms of their educational outcomes, it might be argued that the neighbourhoods' bad reputation and stigmatised position remain on a symbolic level. However, the interviews illustrate that these symbolic meanings have real-life consequences for the schools. First, even if the principals say that their schools are popular among job applicants who are familiar with East Helsinki schools, there are, nevertheless, place-related problems concerning recruitment, as this principal puts it: This school has had a reputation as a good school, but geography plays a role here, because even though this school has much better social networks than many schools in the [city] centre, we have to try to sell the school when we recruit.
A second consequence becomes apparent in talks about school choice. An example shows how Rosemary pupils fear that they may end up in their nearest lowersecondary school, Dandelion. Teachers and parents recommend pupils to apply to "higher-level" schools and talk proudly about pupils who have "succeeded in getting into better schools." In the following discussion, sixthgraders talk about their choices:

Pupil 1: [Dandelion has a] really bad [reputation].
Interviewer: Do adults also talk about it somehow?
Pupil 2: Some parents do, our friend said that…she won't come to Dandelion because her mother won't let her. You know that schools are given scores or something. Dandelion got six, like a really bad score, that's why she's not coming to that school. It's only because there are some foreign pupils…everyone says that Dandelion is a very bad school.
Pupil 1: Everyone who goes to Dandelion will end up smoking and stuff.

Pupil 2: And my sister's friend…said that…you can't get a proper education or [a proper] job afterwards.
The discussion shows how reputations enhance the selfperpetuating circle of neighbourhood and school segregation (Bernelius & Vilkama, 2019;Kosunen, 2014). Pupils and parents stress about the transition phase, and the reasoning for their choices appears quite random since there are no public rankings, for instance. Stigmas become visible in everyday discussions and are manifest in the choices of those who can choose. The example also shows signs of inferior educational opportunities in Helsinki. If people think that some schools are unable to offer "proper education and work," this might label pupils in those schools regardless of the actual quality of the school (Bunar, 2011).
A third example follows Hollingworth and Archer's (2010) findings showing that the pupils' location within 'pathologised' places decreases some pupils' confidence in their own abilities. We found similar patterns in our data. Since pupils seem to have internalised the stigma attached to certain places, teachers in East Helsinki schools need to build up confidence in their pupils for them to learn, as the following teacher's quote illustrates: Despite [others'] expectations, we need to drag these children to the level in which they themselves get to decide how they want to continue….So that they would themselves see 'I can do this.' Sometimes it's very challenging. The trust that 'I can do something' is missing completely….Raising one's self-esteem is important here.
Despite their similar socio-economic position, these consequences are more apparent in Rosemary and Pimento than in Caraway, which highlights the importance of further studying the nuances of how image and reputation affect schools and places. In Caraway, the positive connection between the neighbourhood and the school with its long roots and intentionally maintained traditions appears to soften some of the negative consequences attached to its location in the 'notorious' East Helsinki, of which Rosemary and Pimento suffer from. Also, the fact that several Caraway staff members live in the same neighbourhood, while none of the interviewed members in the case of Rosemary and Pimento do, might facilitate to identify with the neighbourhood for the former and to distance themselves for the latter.

Local Hierarchies and the Relativity of Reputation
Unlike the other schools in this study, Thyme school is mostly affected by its relative position in the local hierarchies. The school and the neighbourhood suffer from a bad reputation or even stigma, as the socially mixed neighbourhood stands out in comparison to the prestigious, middle-class neighbourhoods that surround Thyme. The stigma appears to relate especially to racialised ethnic minorities, while the neighbouring areas are described as "white." The quantitative analysis shows that regarding its social structure, Thyme neighbourhood, in fact, follows the city average closely. Socioeconomic indicators are neither particularly low, nor does its ethnic composition differ. Thereby, Thyme provides an example of how the relative position is socially constructed (Bunar, 2011;Kosunen, 2014;Kosunen & Carrasco, 2016) and how the symbolic representation matters (Wacquant et al., 2014). The interviews demonstrate a positive self-image in the Thyme neighbourhood. Nevertheless, the school is rejected in school choices and stigmatised in public discourse. People living in the neighbourhood and working in the school are aware of this, and comment on it in the interviews, as this parent does: Good connections and all, I don't see any reason why this should…have a bad reputation, but evidently, if you follow [public discourse], the reputation is very bad. Because there are so many people with immigrant backgrounds, but I don't know, I think this is a nice area.
Racialised ethnic minorities are a crucial element of both the Thyme school and the neighbourhood reputations. According to the interviewees, "immigrants" are the reason why outsiders have doubts about it. This seems to force people to comment on the subject even if they do not see it as a problem, in other words, even if it does not relate to their image of the neighbourhood. The sometimes even striking opinions that outsiders have about the Thyme neighbourhood seem to be related to their perceptions of the school, as the next quote by a teacher shows: We interpret this as another example of how strongly school and neighbourhood reputations are connected. While Thyme neighbourhood has a good image among the people living there, those people living in the surrounding neighbourhoods rather demonise it. Interviewees give several examples of this. However, when referring to Thyme as a "bad" neighbourhood, they mainly do that in discussions that relate to the schools in Thyme. We thus argue that the 'need' for outsiders to have an opinion about Thyme emerges, or at least increases, as soon as their children start their school paths. Thus, while we previously showed how the neighbourhoods' bad reputation seeps into schools, it seems that this relationship might also work the other way around. Schools can thus also be a crucial element in the production of a neighbourhood's reputation.
The neighbourhood's low relative position and its connection to the school seem to affect how staff encounters pupils. Some of the interviewees refer to the deficit perceptions and limited expectations some of their colleagues have about the pupils at Thyme school. These teachers believe that they cannot expect similar performances from their pupils as they could in schools in more well off areas, even though Thyme neighbourhood is not particularly low in socioeconomic terms. We interpret this as an additional sign of the school's stigmatised position since it illustrates that the staff has already internalised external beliefs about the 'problematic' school (see Kearns et al., 2013).
This internalisation has at least two types of reallife consequences. First, as the school staff has (partially) internalised that the pupil composition is too difficult to succeed with, it might play a part in maintaining the bad reputation of the school. Second, the school's stigmatised position creates anxiety in the interviewed preschool parents, whose children's school paths lead to Thyme unless they make other intentional choices. In the quote below, a parent who described herself to be highly active in the neighbourhood and said she was fighting back against the bad talk about the area, discusses the difficulties in deciding what stand to take on the negative reputation of the school: Parent: And since we have had the possibility to think about another school. And because of all the contradicting [opinions], some are like, 'this is a horrible school.' Interviewer: What is it related to? Parent: Bullying and, well, immigrants….I can't really tell, people's prejudices affect these things so much that they interpret things that have happened in their own way…of course, I've been thinking about whether they are right [in that the school is bad].
Among all interviewees, some actively wish to challenge the negative perceptions of the neighbourhoods or schools. Changing a bad reputation, especially a stigmatised position, is, however, difficult as it tends to become internalised and relates to larger social issues. Underlying structural issues, such as racism, are difficult to challenge by schools or neighbourhood communities.

Conclusion
Previous research has shown how closely the domains of neighbourhood and school segregation are connected.
Our school interviews-interpreted in dialogue with socio-spatial segregation analysis-offer additional insights into reputation as a mechanism mediating the connection between these two domains. There appears to be a strong link between neighbourhood and school reputation, as schools in disadvantaged neighbourhoods are viewed through the perceptions attached to the place. The same processes feeding into places becoming 'notorious,' as conceptualised by Kearns et al. (2013), also brand the schools as the 'notorious schools' of the 'notorious places' (Bunar, 2011). Despite the case schools' excellent institutional quality and high overall performance in educational outcomes, there is a consistent pattern of the schools struggling with negative neighbourhood reputations and prejudices. This directly seeps into the school's own negative reputation-a relationship that can be interpreted through the lens of 'territorial stigmatisation' (Wacquant, 2007). How closely school and neighbourhood reputations are interlinked becomes evident for example in an interview with a local parent, who feels that outsiders immediately assess the school as bad simply because it is located within that neighbourhood. At the same time, schools might be the trigger creating a need for families to evaluate the neighbourhood, as it appears to be in Thyme. By highlighting this territorial link, our research confirms previous studies and assumptions defending that the doubts about school quality are most often not related to the school's supposedly inadequate quality of education, but rather to social aspects of school segregation (Bunar, 2011;Kosunen, 2014).
The study also demonstrates how neighbourhood and school reputations are constructed in relation to complex local and regional hierarchies (see also Kosunen, 2014). Three of our case schools are located in a relatively homogenous part of the city, where most of the neighbouring areas face similar socio-economic challenges. Their shared negative reputation is consistent with the general stigmatisation of Eastern Helsinki and its disadvantage in terms of poverty and perceived social problems. The national and local discourse, in which East Helsinki has nearly become a synonym for disadvantage and failure, seems difficult to overcome in education. In this sense, the schools' negative reputation is constructed through the neighbourhood reputation, and in relation a city-wide hierarchy of neighbourhoods. Nevertheless, even in this rather homogenous area, relative positions of schools can be identified. The parents' and staff members' positive perspective on the neighbourhood and the Caraway school's long tradition in the neighbourhood might enable to withdraw from negative reputations and to draw counter-narratives and a more positive self-image and reputation about the school. These fine-grained and rather complex differences also highlight the need for further studies on the underlying mechanisms of local social hierarchies.
The significance of the relative position of schools and neighbourhoods becomes particularly visible in the case of our fourth case school, Thyme. In contrast to the three other schools, Thyme has a catchment area with a social status close to the city average. At the same time, it is wedged between areas with high status and a long history of social prestige. Despite the school's excellent educational outcomes and its close-to-average socio-economic composition, the subjective perception of disadvantage is constructed in relation to the neighbouring catchment areas and schools, which became evident in the way school staff and parents talked about the interlinked poor reputation of the neighbourhood and school. Thus, while the East Helsinki schools and catchment areas seem to be mostly evaluated and contrasted against the rest of the city, Thyme's reputation appears to be constructed on a smaller scale in relation to its immediate surroundings.
Like previous research showing that residents evaluate their neighbourhood significantly more positively than non-residents (Permentier et al., 2008), our interviews highlight the differences between the personal experiences with these schools and the perceived perspectives from outside. The conceptual framework of image and reputation helps to distinguish between the insider and outsider perspective and to understand why reputations easily become internalised as 'beliefs about beliefs.' In all case schools, most pupils, parents, and staff are satisfied with their school; however, at the same time, they are very strongly aware of, and sensitive to, the perceived negative views from outside, and how this reputation affects the outsiders' prejudices against the local school. Consequently, these socially constructed symbolic distinctions have real-life effects for the school communities and pupils. The stigmatisation of schools and neighbourhoods is felt by the teachers, who raise concerns about the load and fatigue caused by negative assumptions about the school and impairs the pupils' self-image and confidence. In some schools, pupils are worried about the effect that the school reputation may have on their future. In these discourses, places and schools are even used as self-explanatory phrases-"It's difficult to do that in an East Helsinki school"-creating self-fulfilling prophecies.
The results highlight, for their part, the interconnectedness of multiple domains of segregation, and the fruitfulness of studying these in a common setting to unearth some of the mediating mechanisms. The observed connection between neighbourhood and school reputation and their link to segregation provides both challenges and possibilities for educational and urban policies. On the one hand, this connection demonstrates how and why even an egalitarian, high-quality educational system is not automatically protected against vicious circles of educational segregation in disadvantaged neighbourhoods. As reputation is one of the central drivers of school choices, negative perceptions of schools and neighbourhoods likely affect their rejection as middleclass parents are especially sensitive to fears of choosing the 'wrong' kind of school for their children (Boterman, 2013;Ramos Lobato, Bernelius, & Kosunen, 2018;Ramos Lobato & Groos, 2019;Vowden, 2012). In this process, even institutionally well-functioning schools also appear to be vulnerable to the self-perpetuating growth of school segregation if they are located in a stigmatised neighbourhood. School reputation is also closely linked to residential choices. The interlinked process of stigmatisation of neighbourhoods and schools may thus feed into a multi-domain circle of segregation, where segregation in one domain feeds into the other. The risk of circles of deprivation may be even more profound in contexts where the institutional quality of schools varies more and school choice policies are more liberal. In less egalitarian and more differentiated educational systems, where inequalities between different social and ethnic or racial groups are bigger, such as the UK or US, or where schools partly rely on additional financial support by parents, or are characterised with a special pedagogical approach and gifted programmes, such as The Netherlands or Germany, schools operating in disadvantaged neighbourhoods may be even more vulnerable to rejection in school choices made by educationally motivated families.
On the other hand, as the basic dynamics appear relatively constant over diverse urban settings, the findings offer opportunities to support both schools and struggling neighbourhoods through active policies focused on improving schools and specifically targeting the way they are perceived by the general public. As cities are looking for ways to combat urban deprivation, investing in schools and strategically supporting positive school reputations has significant potential in policies aimed at breaking circles of deprivation in urban neighbourhoods across different types of cities.
Isabel Ramos Lobato works as a Post-Doctoral Researcher at the Helsinki Institute of Urban and Regional Studies, University of Helsinki, and is part of the Geographies of Education and Divided Cities (GED) research group. Her research interests are broadly rooted in urban and social geography with a focus on social inclusion and exclusion in the fields of education and housing. In her PhD, she examined parents' choice of primary schools and its impact on school segregation in Germany.