Spatial structures of student mobility: Social, economic and ethnic ‘ geometries of power ’

There now exists a growing literature on educational mobilities, and this paper contributes to understanding the way contemporary youth imagine the geography of the United Kingdom and how this translates to their mobility intentions. Using Giddens and Massey and drawing on a unique multi ‐ sited qualitative dataset, we examine how these flows can be understood as embedded within narratives of the self that are situated within a particular spatial structuring of social, economic, and ethnic difference. The multi ‐ sited dataset provides a unique opportunity to see the simultaneity of these social relations across space, mutually shaping, and reshaping each other over time. We illustrate how embedded within imagined mobility narratives are deeply unequal structures of economic power, (re)producing oppressed and dominant positions across social and geographic space. Geometries of race and ethnicity are also shown to structure the ways in which different ethnic groups look upon the geography of their university choices. The patterning of these imagined spatial flows around the United Kingdom at the point of university entry can be interpreted as one further manifestation of deep ‐ seated geometries of power that pervade social life.

qualitative, providing richly detailed accounts of these specific spatial units, there also exist larger-scale quantitative work, with Gould and White (2012) carrying out one of the first studies (originally published in 1974) that attempted to quantify British school leavers spatial preferences, in terms of the regions they perceived positively and negatively.
Other work has taken a race-based perspective in examining the significance of place, exploring questions around racism encountered across diverse geographic spaces (Bonnett, 1996;Bressey, 2016;Webster, 2003), and the whiteness of space and white privilege (Lawler, 2012;Rothenberg, 2008;Slocum, 2007). In the context of educational research, there is a body of work examining the educational encounters and experiences of students of colour, exploring higher education and elite university settings, and in relation to their mobility trajectories to university, including the ethnic composition of their origin and destination geographic locations (Ball, Reay, & David, 2002;Dumangane, 2016;Gamsu, Donnelly, & Harris, 2018;Khambhaita & Bhopal, 2015).
We contribute to this work by further exploring the wider power structures within which youth's perceptions of place are situated, especially in terms of the social, economic, and ethnic divisions that pervade social and geographic space. The participants in our research are young people on the cusp of entering university; a significant life event, often regarded in the public imaginary of English higher education (Holdsworth, 2009a) as a time when young people move away from home, and so is likely to bring to the surface these latent underlying spatial perceptions and preferences. Indeed, the research presented here underlines the deep-seated spatial divisions and spatial structures that manifest in the way contemporary British youth frame their transitions to university.
There is now a well-established body of work on the geographies of education, and our paper offers a further contribution to research on student mobilities and the migration of youth across geographic space. Work on educational mobilities has unpacked the idea of mobility in a number of interesting ways. There is work on the flows of students across geographic localities and countries, both internationally and intranationally, which has picked apart the extent and nature of these movements. There is a large body of work on the international flows of students; outward from particular regions, such as Europe and East Asia; and more global in scale (Christie, 2007;Findlay, King, Smith, Geddes, & Skeldon, 2012;Holdsworth, 2009b;Prazeres, 2013). In the UK context, research on the outward international movements of UK students has evidenced links between international student mobility and the reproduction of social class advantage (Brooks & Waters, 2009;Holloway, O'hara, & Pimlott-Wilson, 2012;Waters & Brooks, 2010). Looking at the internal migration of students within the United Kingdom, work has addressed distance travelled and leaving home and regional patterns in student mobility (Duke-Williams, 2009;Singleton, 2016). Research has also dealt with a more ontological questioning of the construct mobility, exploring interesting questions around the liminality of movement and the emotive and affective dimensions to being mobile (Finn, 2017;Holton & Finn, 2018). Other work has illustrated the problematic way in which neoliberal discourses around 'aspirations' and the knowledge economy construct 'immobility', which is considered to be a 'hindrance' to life chances (Allen & Hollingworth, 2013). Elsewhere, the complexity and the gendered nature of mobility in a peripheral, Northern region of Sweden have also sought to nuance the notion that immobility automatically implies a lack of capital (Forsberg, 2019). Our work builds on these critical accounts of what it means to be 'mobile' and 'immobile', by focussing on the way uneven spatial structures are embedded within how young people look upon their choices of where, geographically, they choose to study.
Much of the work looking at the internal flows of students within the United Kingdom has been based on large quantitative, datasets, and when a qualitative approach has been adopted, it has tended to be based on single-site case study. For example, Hinton (2011), Donnelly and Evans (2016), and Ward (2015) have each explored processes of university choice and educational experience in relation to specific Welsh localities. Similarly, Holdsworth (2009a) analysed how working-class 'scouse' local identities affected students attending an academically selective university in the city of Liverpool, and Bathmaker et al. (2016) examined similar issues in Bristol. The work of Clayton, Crozier, and Reay (2009) in their studies of the experience of working-class students across universities in Britain comes closest to a multisited, relational understanding of how university choice and experience is mediated by different local and regional geographies. They explore how student mobility is rooted in particular distinctive economic and cultural geographies of place. In one instance, they describe how a British Asian student opted to attend a university in the Midlands where he would not stand out as a student of colour and which would be similar to his upbringing in a working-class part of a city in the north-west of England (Clayton et al., 2009, p. 164). More generally, the in-depth and rich qualitative work that has examined higher education choice has largely been carried out in a single region, town, or city. This makes it difficult to see wider power relations that exist, especially those that are geographic in nature, for example, contrasts between north and south, or Wales and England. The qualitative dataset used here includes data collected from 20 diverse localities stretching right across the United Kingdom and so offers a glimpse into the underlying spatial structuring of social and ethnic inequalities.
Indeed, the study began with the explicit purpose of bringing a geographic perspective to examining questions of higher education transition. We begin by theorising spatial structure, the self, and geographic mobility and outlining the theoretical ideas that frame our analysis presented later.

| THEORISING SPATIAL STRUCTURE AND THE SELF
In understanding the role played by place in shaping the geographic movements of university entrants, we combine Giddens' (1991) work on identity of the self with Massey's (2005Massey's ( , 1994 conceptualisation of space and place. Giddens' work on identity of the self helps to explain the often unconscious ways in which social practice and action is rationalised and how choices or behaviours are part of long-term trajectories and imagined futures. Massey helps us to spatially situate the self and help explain how place and the specific constellation of social relations it embodies, structures the ways in which people look upon their (im)possible geographic destinations. Furthermore, Massey helps explain how an act of mobility is in itself laden with unequal power relations. By combining Giddens' work on identity of the self with Massey's spatial perspective, we hope to provide a more holistic framework from which to make sense of the geographic destinations of university entrants.
We first describe the relevance of Giddens' work on identity, before theorising how a spatial component could be combined through incorporating Massey's ideas on space, especially 'power geometry'. We begin with Giddens and move on to Massey and provide thoughts on the ways in which both can combine to provide a complimentary theorisation. Giddens' idea of ontological security describes the self-identify each of us manufactures about who we are, where we have come from, and where we are going-our 'narrative of self' we hold and carry with us. Giddens' (1991)  and what it means to be a 'son', 'daughter' , 'wife' , or 'husband' .
The existential question of self-identity is bound up with the fragile nature of the biography which the individual "supplies" about herself. A person's identity is not to be found in behaviour, norimportant though as it isin the reactions of others, but in the capacity to keep a particular narrative going. The individual's biography, if she is to maintain regular interaction with others in the day-to-day world, cannot be wholly fictive. It must continually integrate events which occur in the external world, and sort them into the ongoing 'story' about the self. (Giddens, 1991, p. 54) Holding these 'narratives of the self' is necessary to avoid states of anxiety, and these narratives are unconsciously drawn on as we journey through life, expressing desires, preferences, and making choices.
Giddens refers to this as our 'practical consciousness'-our sense of self that guides our daily activities and ultimately the course taken in life. Anxiety is a central element here; it is an unconscious and what Giddens calls 'free-floating' kind of anxiety that is not about any specific threat or immediate external danger, but rather internalised threats to do with our self-identity-which compel us to hold on to what is familiar and known. One of the manifestations of this sense of self is the collection of habits, behaviours, and forms of conduct, which are important to maintaining this self-identity and providing ontological security of the self.
Inherent within higher education choices will be narratives of the self, as young people unconsciously navigate themselves towards paths that are in line with their 'practical consciousness' of who they are as a person, what they are about, where they belong (and do not belong), and where they are going. Belonging here is understood to have both social and inherently spatial elements; we see belonging as being shaped 'through being both at home and away and through the dialectic of roots and routes' (Urry, 2000 132-133). In terms of the focus of our analysis, the spatial preferences of young people for higher education study, and their (im)mobility intentions, will likely reflect their sense of self in terms of where they belong, who they are, and what they imagine they will be in the future. It is reflective of their perceptions about what is valuable in life and valued by them.
For example, being intimately connected with family and in close immediate contact with family members is something that is valued by many (Finn, 2017). Occupying positions in dominant and dominating institutions of society (such as elite universities) could be valued more by others. These narratives of the self are inevitably interconnected with social structure and power within society. Any understanding of an individual's sense of self that is not placed within the context of classed, racialised, and gendered power structures will likely be limited.
The narratives of the self at the point of entry to higher education are located within power structures that are interwoven with spatial inequalities and dynamics. We attempt here to spatialise Giddens' ideas around narratives of the self, by combining them with Massey's understanding of space and place, showing how power relations circulating across space shapes young people's narratives of the self. From Massey's perspective, place is conceptualised as a particular set of social relations; connections and disconnections to the multitude of identities (including cultural, social, gendered, and sexualised); and structures (including economic, political, and religious) in constant circulation and creation across space. In this sense, places are seen as unique 'pauses' within space.
Thus, the spatial is socially constituted. "Space" is created out of the vast intricacies, the incredible complexities, of the interlocking and the noninterlocking, and the networks of relations at every scale from local to global. What makes a particular view of these social relations specifically spatial is their simultaneity. … Seeing space as a moment in the intersection of configured social relations (rather than as an absolute dimension) means that it cannot be seen as static. (Massey, 1994, p. 265) Massey's work provides one way of theorising the role of place as a mediator in the higher education choices of young people. It enables an understanding of how place-interpreted here as a unique configuration of social relations circulating in space-can come to be important in shaping the migratory choices of young people for university.
Choosing the 'right' kind of geographic location may be about locating in a particular place that is (dis)connected to a particular configuration of relations circulating in space, for example, locating in a place connected to certain classed or ethnic relations. These preferences could be driven by the need to maintain an individual's narrative of the self, a desire to be connected to particular resources, and identities that will help craft an individual's idea of their future self (Giddens, 1991). In beginning to develop this theorisation, we provide some examples in this paper of how certain unique spatial constellations figure heavily in the minds of young people when they are making their university choices.
In this way, we attempt to show how university choice may be a mix of choosing the 'right' university and course (Reay, Davies, David, & Ball, 2001) and also migrating to the places with the 'right' kinds of connections (and disconnections) to particular identities and resources. More generally, this calls into question what we mean by 'migration' and the breadth of meaning imbued with terms such as 'mobility' and 'immobility'. But it must also be recognised that these social relations conceptualised by Massey are not neutrally configured and are held together by structures of power.
Moreover, and again as a result of the fact that it is conceptualised as created out of social relations, space is by its very nature full of power and symbolism, a complex web of relations of domination and subordination, of solidarity and co-operation. (Massey, 1994, p. 265) Taking this further, her analysis of the 'power geometry' of mobility of people and flows of capital highlights the following: Using Massey's lens on the power geometries of space, we seek to rethink how power is exercised in relation to students' decisions about spatial mobility for university. As Forsberg (2019: 328) notes, it is not only Massey but also Ahmed (2014) and Cresswell (2006) who can be drawn on constructively in thinking about how power and mobility are intertwined. In both cases, mobility both reflects and creates inequities of power. Looking at broad patterns in the migration of student within and beyond the United Kingdom, it is clear that relocating to a different part of the country (or internationally) to attend university is something that is not experienced equally by all sections of societyand continues to be the preserve of largely privileged groups. In Massey's terms, it is the privileged groups who are really in control here; they are the ones who can turn migrating away to a distant university to their advantage. In one sense, their movement actively decreases the power and influence of dominated groups, who do not move, and are unable to access the sorts of institutionalised capitals and resources in locations that are either socially or spatially inaccessible (or both).
The power geometry is also significant for how places themselves come to increase or decrease the power and influence of different groups. Massey's arguments about power geometry and time-space compression originally made in relation to processes of globalisation, but we see no reason why they may not be applied to understand power imbalances on a narrower geographic scale. Indeed, in many ways, countries as spatially diverse as the United Kingdom represent their own microcosms of the social, cultural, and economic inequalities Massey was theorising about. Although it is not possible with our dataset to evidence any possible long-term impact of individuals locating in particular places, it is certainly clear that intentions to locate are caught up with a desired or imagined future self (Giddens, 1991). We show later how such narratives of the self likely have implications for the kinds of power and influence different groups may have in the long term. Elsewhere (Donnelly & Gamsu, 2018: 976), we have argued that student mobility for higher education is intrinsically political, with students' decisions embedded in narratives that are shaped by and recreate deeply uneven economic, cultural, and symbolic geographies.
In this paper, we wish to explore how students' narratives and anxieties reflect these uneven geographies of power. In the context of student mobility, there are multiple ways in which student mobility involves both control and reflects and reinforces pre-existing power relations, which we are able to examine more closely here through a rich and multisited qualitative dataset.

| METHODS AND DATA
We draw here on data from a substantial qualitative dataset collected as part of a 3-year programme of research on the spatial imaginaries and higher education mobility intentions of young people (aged 17/18) across the United Kingdom. This wider dataset includes interviews with over 200 young people and 40 teachers across 20 different UK locations. The programme of work sought to bring a geographic perspective to debates around higher education and social mobility, and so the sampling of our fieldwork sites was crucial. Locations were purposively selected to represent diversity in geographic location, including inner-city locales, rural areas, coastal towns, and postindustrial locations-and stretching across all four corners of the United Kingdom, with each reach of England represented and the three 'home' nations of Wales, Scotland, and Northern Ireland. The schools selected were themselves diverse, and included not only both private (n = 5) and state (n = 15) but also variety in social class and ethnic composition of the state schools. Our sample is socially and ethnically diverse, representing the diversity of young people in the schools (we selected based upon demographic information collected from questionnaires). Given the size and scope of this wider dataset, it would not be possible to do justice to the data in any analysis presented here, and so in what follows, we draw upon particular cases of students from five of the case study schools. These cases act as 'windows' into the phenomena in question, to observe and illustrate the deep-seated power imbalances. The full details on their social class and ethnic backgrounds are provided alongside the data presented.
Data collection involved an initial 'mapping exercise'(see Donnelly, Gamsu, & Whewall, 2019) for the purpose of selecting participants and producing a visual aid that could be used within the interview process (for participants to 'speak to'). The vast majority of young people were intending to apply to university and were on the cusp of making their application. In terms of the mapping exercise, participants were presented with a map of the United Kingdom and asked to colour- to teachers about how to administer the map, in order to ensure they did not bias the data collection process. Importantly, the map was completely blank, with no place names and county/national borders, so that participants' geographical perceptions were not framed for them; rather, they are permitted to show the researcher their subjective geographies. In applying this tool, we recognise that some may regard a map as imposing a particular frame of reference that does not permit the elicitation of subjective geographies. However, it is important to point out that the map is used as part of the interview process in our research, and acts as an orientating and starting point for discussion about young people's imagined geographies. These methodological questions are more fully addressed elsewhere.
The project amassed over 1,000 highly creative and colourful maps from participants across the 20 locations. Participants varied in how they approached the exercise, some colouring in all spaces of the maps, others only partially completing it. Some took the exercise very seriously and produced richly detailed maps, creating descriptions at a granular level of detail about their spatial preferences, aversions, and geographic biographies of themselves and their families. The maps were used in the research process to select a sample of research participants to carry out semistructured interviews with them in order to more fully understand their spatial imaginaries and mobility intentions. These interviews were orientated around their map, which was used to allow participants to speak about their geographic preferences and aversions from their own perspective. We began each interview with the simple questions: 'tell me about your map' . Follow-up questions attempted to hone in on particular aspects of their map, for example, asking them about particular areas they had shaded in certain colours. The semistructured interviews also covered a range of other topics, including family experiences of university, subject and university choices, and the school's involvement in their choices. However, it was giving participants the opportunity to 'speak to' their maps that produced the luminous and 'thick' narratives that are drawn on in this paper.
In what follows, we use these data to examine the unequal structures of power that are embedded within young people's accounts of their geographies and their immobility/mobility intentions. Giddens and Massey are used to help make sense of how young people frame where they see themselves 'fitting', in terms of their identity of the self, within the context of deep-seated spatial structures of power that pervade the UK context. The unique multisited dataset allows us to illustrate the boundaries and attachments to place that young people construct, which reflect historically rooted imbalances and inequality.
We first examine these imbalances culturally, socially, and economically, before moving on to their racial and ethnic dimension. The geographic places in which young people identified with (and presented as desirable locations for university) were often interwoven with their identity of self and imagined future self. A striking example is that of the young people in the north of England. This identification with your immediate locality, sometimes presented as being 'trapped'

| SOCIAL, CULTURAL AND ECONOMIC GEOMETRIES OF POWER
in place, has often been perceived as something associated with lower social class groups, but we found cases cutting across the social class spectrum. As in the work of Forsberg (2019) and Finn (2017), normative assumptions around immobility and a lack of power during the transition from school to work or university hide more complex realities. We explore this by drawing on two of our northern students; Alex, who attended a suburban comprehensive in the north-east, and Dan, who attended an elite independent school in the north-west.  The 'narrative of self' Dan and Alex construct here is intertwined with what they perceive to be 'northern' culture. They disassociate themselves from what is perceived (and constructed) as a cold, individualistic, and unfriendly 'south' and align their identity with the warm, communitarian and friendly 'north' . The biographies and narratives of the self these two young people exhibit, as conceptualised by Giddens (1991), are about being 'northern', which to them means being a person who is caring, has communitarian values, and is friendly to strangers. As Giddens (1991) contends, these are manufactured identities of the self created by the young people to provide them with a sense of existential security of who they are and what they are about; it provides a kind of ontological security. This security and the broader sense of spatial belonging are not necessarily based on straight-forward, historical ties to a local working-class community-Dan is a bursary student at an elite fee-paying school in the north-west, Alex's parents are Liverpudlian, and one of them is a medical professional. In each case, there is not an immediate and straight-forward relationship to historical classed identities within a particular locality or region. This suggests partly that there are more complex forms of spatial belonging closer to the 'elective belonging' described by Savage et al. (2005), but although there may be some degree of distance between each of the two young men and the local and regional identities they refer to, in both cases, they associate geographical identities of northernness with forms of sociocultural and political life associated with solidarity. In that sense, they are perhaps closer to the 'resistant aspiration' that Bright (2011) describes amongst constructed and open it up to a rigorous questioning (e.g., by experiencing places in the 'south' more and finding contradictory evidence), it could expose the fragility, complexity, and contradictions within such narratives. This is what Giddens (1991) describes in terms of the human need to maintain our ongoing narratives and biographies to protect against states of anxiety. Indeed, both Dan and Alex ultimately chose universities in the north of England, which could be to some extent linked to a need to maintain this narrative of self into the future. Of course, many other students may move away and maintain self-identity in other ways, through friendships, clubs, or societies, for example.
To understand these narratives of the self more fully, they must also be placed within the context of broader 'power geometries' and the unequal structuring of economic power in a wider sense. It is striking that both Alex and Dan only mention the 'south' (of England), and not any other part of England, or other 'home' country of the United Kingdom, when narrating the 'northern identity'. In a relational sense, the 'north' of England was defined here through its difference to the 'south'-being 'northern' is an expression of not being 'southern'. It is likely that this is suggestive of the ways in which an economically dominant south-east of England has historically dominated areas in the north, which have over a very long period been made peripheral to the institutions of cultural, economic, and social power that cluster in the south-east (Robson, 1986;Martin, 1989)

St. Alexanders Boys' School, [Independent school, London])
For Samuel and other young people at St. Alexanders Boys' School, the narrative of the self they crafted for their present and future selves was that which involved being connected to cultural, social, and political positions of power. They routinely discriminated between places where possible universities were located, discerning the way these places were connected (and disconnected) to different sets of social relations across space. When Samuel is talking about the 'feel' of the cities, Edinburgh being 'cleaner' and closer to London, he is talking indirectly about the connectedness to particular social relations that Edinburgh has over Cardiff and Swansea. He is talking here about their closeness not in a proximal distance sense, but in terms of them being close in their similar connectedness to particular social relations. It is these social relations Samuel desires and sees as reflective of his own sense of self. Progressing to university was a moment when it was necessary to maintain his present and future self and seek out these social relations that come to cluster in particular places.
The link made here by Samuel between Edinburgh as a political centre and being able to imagine himself in those environs underlines explicitly Massey's (1993)  Amongst the students that we interviewed at St. Alexanders Boys' School, considering attending an elite university and entering elite geographic spaces (as well as explicitly and implicitly rejecting dominated geographic spaces) was discussed in a tone of naturalised ease and comfort. It is also worth considering that whilst this discourse of ease in relation to moving through elite spaces expresses a position of dominance and power, it also underlines the essential insularity and highly selective nature of the geographical reach of students in the most elite British schools Wakeling & Savage, 2015). Their attitude was also present in discussions of international university choice, where (and as found by Waters and Brooks (2010)) students' discourse of how they were choosing universities was not strategic, or the object of intense effort, but was instead described as a simple range of expected outcomes: I'm still kind of half and half between the UK and US universities. The contrast in tone here is marked with these 'dream aspirations' not a realistic prospect. This is reinforced by his inaccurate estimation of the price of the UCLA course, which costs $31,949 for non-Californian residents. The detail of costs and scholarships that is a taken for granted, and absent, element in Luca's discussion of studying at an elite American university, weighs heavily in Daniel's discussion, underlining the impossibility of such a path. Instead, he described a more realistic course as attending Exeter to study film with Liverpool John Moores or Edge Hill as his back up. In the event, none of these options were possible, and he began an engineering course at a college in Liverpool. The contingent, ad hoc nature of his choices and preferences were present in his interview and are reinforced by the contrast with his final university 'choice'. The geographical setting for his decision-making further underlines how students' choices are embedded in local and regional geographies.
'Power geometries' are clear to see here, with these examples providing a window into the system of social relations that maintain uneven spatial structures and a glimpse into who really is in control of mobility (Massey, 1994). The anticipated national and international mobility of Samuel and Luca fits into a broader deep-seated geography of uneven development and their repetition of historic sociospatial trajectories of the elite. Mobility here is entirely prescribed by the power of accumulated wealth and cultural capital embedded in family backgrounds, the schools these students attend, and the broader geographical context in which they live.

| ETHNIC GEOMETRIES OF POWER
Educational research has shown how ethnicity, and the 'whiteness' of educational and geographic spaces in particular, is significant in structuring the experiences and encounters of minority ethnic groups (Bhopal, 2018;Bonilla-Silva & Forman, 2000;Dumangane, 2016;Warikoo & De Novais, 2015). Warikoo and de Novais (2015) found that white students who previously lived in white segregated neighbourhoods were likely to be influenced by the 'colour-blind frame' of their lives before university and have a lack of awareness of the ways in which white students dominate campus life. Dumangane (2016)  It was not uncommon within the narratives of our participants from lower social class groups to geographically confine and frame their mobility intentions in terms of where they knew people. This likely reflects the greater risk associated with university entry for lower social class groups, with these connections providing a sense of safety and security. In one sense, the prospect of leaving school and transitioning to university represents a potential threat to an individual's ongoing narrative and sense of self (Giddens, 1991). Choices about where to study are contingent upon the story of the self, what we tell ourselves about who we are, what is important to us, where we are going, and the place where we need to be to enable this. There is a threat to the self because certain locations could disrupt this ontological security; this ongoing story we craft about ourselves. Looking at our data, the ways in which familial connections and relationships framed where, geographically, young people imagined studying was especially acute for British Pakistani and British Bangladeshi students from lower socio-economic backgrounds. We examine this issue more closely here in relation two of our schools: Huddersley Academy (West Yorkshire) and Tower Chapel School (London).
One of the primary reasons for selecting our West Yorkshire school (Huddersley Academy) was that our analysis of HESA 1 data from 2012 entrants showed the vast majority of students from this school opted to attend the same local university. This despite there being a number of universities within commuting distance, with a mixture of pre-1992 and post-1992 provision, represents a wide realistic set of choices for students with different grades. Looking at the ultimate destinations of those we spoke to, it was clear that nobody left the area at the end of sixth form, with many being forced to repeat their A-level or attend local colleges (due to low exam achievement), and only two out of the 10 students we spoke to progressing to university-again, the same local institution. However, when we spoke to our participants during their first year of sixth form, some did mention the idea of moving away for university, as an imagined possible future option. But in their list of possible places, there was more often than not a familial association to the framing of their conceivable geographic destinations, as evident in Dunya's map (Figure 3). Both Tahir and Dunya ultimately did not progress to university at all and were forced to resit their A-levels. However, in describing their maps and where, geographically, they imagined themselves moving for university, their narratives of possible places are entirely couched in those where they have family connections. For Tahir and Dunya, these places were 'safe' spaces for university study because of these connections. Elsewhere, the students talked about close relationship with family members and their wider family network. In a Giddens sense, these relationships were spoken about as a key part of their sense of self, who they were, and what was valuable to them. Their framing of university choices around such familial connections can therefore be interpreted as maintaining this sense of self, ensuring that there ontological security is not disturbed by a move away from important ties and relationships. For the white privileged students mentioned above, familial networks were not mentioned at all in their framing of possible places. For them, their sense of self may entail quite the opposite; it may be about moving to new places, seeking out new connections, and not being grounded in any one place.
The differences evident in these narratives connect with what has been found elsewhere in relation to the 'risks' associated with university for different social class groups (Reay et al., 2001). They must also be seen in light of other research showing that South Asian (Indian, Bangladeshi and Pakistani) young people (especially girls) are more likely to stay living at home when studying at university than their white peers (Khambhaita & Bhopal, 2015). The accounts from our participants must also be taken within the context of the broader uneven geographies of race and ethnicity, which are reproduced through the higher education choices of different ethnic groups . For other ethnic minority participants, especially Muslim girls, there was sometimes an expressed desire to locate in a geographic place where they did not feel 'out of place'. One of our London state schools, Tower Chapel School, is located in an ethnically diverse neighbourhood of London. In talking about place and possible geographic locations of study, Bipasha is highly conscious of her own ethnic identity, in a way that our white British participants were not: Bipasha talks of white people where she lives 'knowing' her and not 'othering' her through everyday interactions (such as trips to hospital), in a way that she perceives white people in other geographic locations would. The fact that Bipasha alludes to her ethnic identity in the context of discussing where, geographically, it might be (im)possible to move for university, is likely a manifestation of racialised and ethnic geometries of power. As discussed earlier, university choice can be understood here as a mix of choosing the 'right' course/institution and also migrating to a place with the 'right' kind of connections and ties to wider space (Massey, 1994). For Bipasha, it is a place where there perhaps exists ties and connections to a mix of different ethnic identities, which ensure she does not feel out of place and is able to maintain her identity and sense of self (Giddens, 1991). The narratives of our white participants make no mention of their ethnic identity or how this might be perceived by others when describing possible locations for university study. This illustrates how the white ethnic group, especially white participants from more privileged backgrounds, were the group really in control of mobility over geographic space. Even in ethnically diverse localities, unequal power relations across different ethnic groups mean that white groups have agency in a way that ethnic minorities do not, reflective of the forms of 'white privilege' evidenced elsewhere. These examples underline the fact that different ethnic groups do not experience (im)mobility or perceive the range of (im)possible geographic destinations open to them, in quite the same way.

| CONCLUSION
We have explored here how social, economic, and ethnic imbalances evident across space manifest themselves in young people's imagined geographies and their consequent mobility intentions. The analysis contributes to the established body of work on the nature of place, belonging and attachment, and the burgeoning literature on geographies of education, especially student mobility research. Our work sought to bring an explicitly geographic perspective to understanding student mobility, examining how the internal geography of the United Kingdom is imagined by youth, and going beyond their immediate locales to capture the complete spatial structure wherein they are situated and embedded. To do this, the mapping method proved a useful means of accessing how young people from diverse localities understood and perceived; it elicited their geographies of the United Kingdom. We can see from the data presented here that young people's geographies of the United Kingdom vary according to where they are speaking from, in terms of their geographic location, social class, and ethnicity.
Giddens and Massey provided a theoretical lens to interpret how mobility intentions are embedded within a 'narrative of self' that must be understood within a broader set of power relations. Mobility can be reflective of different sets of historical and deeply embedded unequal power structures to do with geography, race, and class. In one sense, they are reflective of historically unequal economic geographies of power, with postindustrial areas of the north peripheral to the dominance of London and south-east. This dominance was felt and noticeable within the narratives of a number of our participants in northern localities. For such participants, a perceived 'northern' culture, defined in opposition to the 'south' , was an important part of their identity and motive for staying rooted in the region. Their narratives make even more sense when seen relationally against the narratives of the highly privileged and affluent young people we spoke to in a fee-paying London school. These young people exercised the greatest degree of spatial agency in narrating their mobility intentions and distinguished between not only types of university but also geographic places in terms of their stocks of cultural and social capital.
Their university choices overlapped with inherently place-based choices. In considering where to study, they were only constrained in terms of locating themselves in places (and positions) of economic, political, and cultural power/dominance. For example, the Edinburgh Festival was a cultural manifestation symbolic of the kind of narratives of the self these young people identified with. A fuller understanding of how historically uneven power relations across space shapes contemporary youth is evident here in examining the accounts of participants from dominated regions of the north with dominant areas of the south. Our study underlines the importance of carrying out multisited research across diverse geographic and social locations in order to capture, as advocated by Massey, the simultaneity of power relations evident across space.
Mobility intentions for those from lower social class backgrounds were also often framed in relation to places where they knew people, which was especially the case for participants from minority ethnic groups. We saw from the example provided here that in narrating their spatial imaginaries and conceivable choices, the places mentioned were all those where they had connections. To make sense of this further, spatial imaginaries of minority ethnic youth and their mobility intentions must also be seen as reflective of unequal structures of race manifest within contemporary Britain. It was also the case that white participants in the study exhibited no spatial constraints on the basis of their ethnicity, but the British Asian students we spoke to often alluded to issues of race and the racial structuring of space. As evidenced elsewhere (Ball et al., 2002), participants from ethnic minority backgrounds sometimes framed their geographies of the United Kingdom in terms of the ethnic make-up of localities and fear of being 'othered' by people in localities that were not ethnically diverse. These findings underline the importance of directly addressing ethnicity in young people's perceptions of place and mobility intentions, especially in contemporary climate of rising far right politics and the increasing marginalisation of minorities. Although it was not possible in the space available here, a fuller analysis of the wider dataset from this perspective is needed.