Lunchtime food and drink purchasing: young people’s practices, preferences and power within and beyond the school gate

Abstract This paper highlights factors that influence young people aged 13–15 years when purchasing food or drink within or beyond the school catering service. The paper draws from a qualitative study of secondary schools in Scotland, which varied in terms of relative socio-economic deprivation and density of food and drink businesses within a 10-min walk. Analysis is situated within a children’s rights framework, underpinned by the sociology of childhood in order to make sense of the power and influence that young people have when purchasing food and drink. The data suggest that the school cafeteria is less able to meet young people’s social and food needs than the external food environment. The commercial basis of young people’s relationship with food businesses informs the way young people are treated as valued consumers. The authors consider the implications this has for the way that food and drink is sold and promoted in schools.


Introduction
Up to a third of young people's food and drink intake is consumed during the school day (Nelson et al., 2004) leading to calls for legislation to control or ban the sale of food and drink high in fat, sugar or salt in this setting (Crawley, 2005), in order to try and improve the nutritional quality of young people's lunchtime food and drink purchasing. In Scotland, where Nutrient Standards for Schools are in place for all state schools (Scottish Government, 2008), one resource (Scottish Government, 2014b) advocates a four-pronged action plan, to encourage young people to 'stay on site' and use the school catering facilities; to regulate the external food environment through licensing and planning laws; to guide caterers and retailers to provide healthier options; and to develop marketing initiatives for healthier choices within and beyond the school gate. The external, unregulated food environment in the vicinity of schools is often viewed as problematic in terms of the types of food and drink available to young people and the contribution it makes to their overall diet (Estrade ABSTRACT This paper highlights factors that influence young people aged 13-15 years when purchasing food or drink within or beyond the school catering service. The paper draws from a qualitative study of secondary schools in Scotland, which varied in terms of relative socioeconomic deprivation and density of food and drink businesses within a 10-min walk. Analysis is situated within a children's rights framework, underpinned by the sociology of childhood in order to make sense of the power and influence that young people have when purchasing food and drink. The data suggest that the school cafeteria is less able to meet young people's social and food needs than the external food environment. The commercial basis of young people's relationship with food businesses informs the way young people are treated as valued consumers. The authors consider the implications this has for the way that food and drink is sold and promoted in schools. Public Health England, 2014;Scottish Government, 2011, 2014a; this view is widespread in popular media 1 . The evidence that this perception is justified is inconclusive (Héroux, Iannotti, Currie, Pickett, & Janssen, 2012;Smith, Cummins, Clark, & Stansfeld, 2013).
Food 'choice' is a complicated process; a decision made by an individual about what food or drink to purchase is underpinned and driven by many factors, including social factors (Murcott, 1995(Murcott, , 2000. Since adolescence represents a complex period of transition within the life course, it can often have an impact on purchasing and consumption practices (Story, Neumark-Sztainer, & French, 2002;Wills, 2005). As young people move from eating mainly at home to having their own economic resources to spend on food and drink consumed with friends, and in settings outside the home, the pressure to conform and fit in, whilst simultaneously developing one's own (food and drink) identity, is strong and pervasive (Brown, McIlveen, & Strugnell, 2000;Miles, Cliff, & Burr, 1998). Young people do not operate in a social vacuum and their food and drink purchasing is subject to political, economic and legal discourses (Nestle et al., 1998;Popkin, Duffey, & Gordon-Larsen, 2005) and such discourses may be more or less keenly felt during the socially changeable period of adolescence, compared with other times across the life course (Lesko, 2012). Further, other factors, such as socio-economic deprivation (by which we mean lack of access to material and social resources within an area or neighbourhood; Townsend, 1987) is likely to influence how young people operate and are perceived in their local environment (Murcott, 2000), including by schools as well as by commercial food and drink business operators.
The purpose of this paper is to highlight some of the factors that might influence young people to purchase food or drink within or beyond the school catering service. We explore the reasons why young people might be attracted to the external food environment and compare it with the school dining experience in order to consider ways that schools can continue to enhance their on-site catering facilities. We do this because the school setting is considered within the public health paradigm as the ideal site for improving nutritional choices and providing opportunities for young people to learn about food and drink associated with better health outcomes (Crawley, 2005). We draw on the sociology of childhood 2 and a 'children's rights' perspective to further examine whether young people have power, agency or influence within school dining rooms, compared with their experiences of the external food environment, to examine whether power, agency and the ability to influence within school are factors that might be associated with where and what young people purchase to eat or drink at lunchtime.

Theoretical framework
Views concerning children and young people have shifted in recent decades to consider the rights that young people have to participate as members of a civil society (Wills, Appleton, Magnusson, & Brooks, 2008). This shift recognises that young people are not merely 'adults in the making' but social beings who absorb, reflect and also shape the contemporaneous world that they experience and inhabit. Policies and practices supporting this perspective are underpinned in most countries through the United Nations (UN) Convention on the Rights of the Child (United Nations, 1989). The UN Convention situates education as a right for all young people, along with the right to participate, to have a level of autonomy in line with their age and maturity, and to have the right to be heard.
The principle that children and young people have rights to participate and be heard sits alongside the emergence and academic development of a sociology of childhood, also referred to as the new social studies of childhood (Christensen & James, 2000;James & James, 2004). This development has been a catalyst for the way that children, young people and their childhoods are now conceptualised, at least in academic or sociological terms (Quennerstedt & Quennerstedt, 2013). Within this theoretical perspective young people are accepted as active and valued social beings with a unique view of their own practices, experiences and needs. Understanding such views is imperative for effecting change 'from within' , thus precluding the imposition of an adult-led framework that has little resonance with young people themselves (Wills et al., 2008). A theoretical consideration of the social nature of 'childhood' or, more pertinent to this paper, 'youth-hood' is critical for promoting the rights of young people to be heard and to participate (Mayall, 2002). This position is used here to situate the findings of the study presented to better understand why many young people do not purchase food and drink in schools.
Young people's own experiences and values are shaped in relation to the other people with whom they come into contact (Eckert, 2004), including peers, teachers and catering staff at school, as well as the socio-spatial, moral and political context(s) within the school itself (not to mention family and other contexts, though these are not the focus of this paper). Schools are often inherently unequal settings in terms of the relationships between adults and young people who spend their time there. Whilst the principles of involving young people in decision-making and giving them a voice are often acknowledged, the practice of ensuring this occurs in schools is somewhat more difficult (Bailey & Wills, 2010). This gives adults the balance of power (Quennerstedt & Quennerstedt, 2013) and tends then to situate young people in a passive role, as receivers of care, education and knowledge. This has been highlighted in studies of primary school dining rooms (Daniel & Gustaffson, 2010;Pike, 2010) where teachers, for example, did not perceive lunchtime as a period when children could learn about healthier eating but rather that children should put into practice over lunchtime what they had learnt in the classroom about nutrition (Pike, 2010). In terms of negotiating power relations, however, young people of secondary school age are often adept at resisting authority and at developing creative strategies to enable them to do so (Holt, 2004). In one study for example (Fletcher, Jamal, Fitzgerald-Yau, & Bonell, 2013), young people sold restricted food to each other on the 'black market' , food that they brought in from outside school in order to subvert the school and government healthy-eating agendas. Other studies of secondary schools have suggested that leaving school during the lunch break is essential for some young people, in terms of negotiating respite from aspects of the socio-spatial or socio-economic environment perceived to be oppressive or challenging, such as relations with peers and teachers or queues and arguments in the dining room (Lawton, 2008;Wills, Backett-Milburn, Gregory, & Lawton, 2005).

Selection and recruitment of case-study schools
The study presented here used a case-study design, focusing on seven state secondary schools in Scotland. Six local education authority areas were approached about their schools taking part, of which five gave consent for the research team to contact individual head teachers. The local education authority areas and the schools were selected partly on the basis of the socio-economic classification of the school postcode, to ensure schools classified as being in areas with higher and lower relative deprivation were included. In order to be eligible schools had to have no rules preventing pupils from leaving the grounds at lunchtime. Schools were also selected to try and ensure there was variation in the local food environments across the case-study sites, in terms of the density and type of food outlets within 800 metres (a 10-minute walk) of the school gate, as some studies suggest these are important factors (Crawford, Ellaway, Mackison, & Mooney, 2012;Cummins & Macintyre, 2002;Lake & Townsend, 2006).
After discussion with head teachers, seven schools were included in the study (in two council areas, two schools were included, in each of the other three council areas one school was selected), four in areas of high relative socio-economic deprivation, two in more mixed catchment areas and one with low relative socio-economic deprivation. Deprivation status was determined according to the percentage of pupils registered for means-tested free school meals in the 2013 free school meals data-set 3 and according to the Scottish Index of Multiple Deprivation ranking of postcodes 4 . In terms of the external food environment within 800 m of schools, two schools had a high number of available food outlets (100+), three had a moderate number (21-99 outlets) and two schools had a lower number of outlets (6-20). The number of available food outlets within 800 m was determined through evaluation of local authority lists of registered food businesses. This method has inherent limitations as local authority lists often inaccurately document the number of food businesses in an area (Lake, Burgoine, Stamp, & Grieve, 2012;Lake, Burgoine, Greenhalgh, Stamp, & Tyrrell, 2010), but manually adding to the list when new businesses were observed in the field helped improve the accuracy of the method overall.

Methods used and topics covered
After a pilot phase involving one school (subsequently excluded from the main study data-set) a range of methods were used to engage young people aged 13-15 years with the research, including focus groups, a semi-structured written exercise, lunchtime 'go-along' tours 5 and individual or group interviews. One focus group was conducted with 13-to 14-year-olds and one with 14-to 15-year-olds at each of the seven schools. The research team also undertook periods of observation within and outside each school at lunchtime and extensive field notes were written. Interviews were conducted with school head teachers and kitchen supervisors and with a range of local food and drink retailers in outlets where young people were seen making purchases. An online food and drink purchasing recall questionnaire was also developed and administered to young people, though the results are not included in this paper 6 . A semi-structured written exercise was developed to allow pupils who completed the online questionnaire before the end of the class period the opportunity to provide further written details of the food and drink outlets they purchased from and the reasons for this, using prompts provided by the research team on work sheets. The topics covered using each of the methods employed are described further in Table 1.
Video and audio recordings of observations, participant observation during go-along tours, interviews and focus groups were made as each situation allowed (it was not always practical, for example, to video record whilst walking with young people into local food businesses). Photographs were taken within the school and the external food environment, contributing to the data-set that was subsequently analysed.
Ethics approval was obtained prior to fieldwork commencing and the parents of young people invited to participate at each school were informed in writing that the study was taking place and given the opportunity to withdraw their child from taking part (none did so). Young people who agreed to participate in any aspect of qualitative data collection gave their written consent at the outset. Participants (young people and adults) were informed that any photographs or video footage that identified them would be seen only by the research team and would not be used when disseminating findings about the research.

Data analysis
Data analysis was inductive and iterative (Patton, 2002), building on each source of data and after regular discussion amongst the research team. Transcripts of interviews and field notes were repeatedly read and analytic notes made. These data were independently thematically coded by several members of the research team; NVivo 10™ (QSR International (UK) Ltd, Daresbury, UK) was used to manage the analysis and to develop a coding framework. Notes were made after viewing the photographs of the school and local food environment and these notes were then thematically coded in NVivo 10. Notes were also made when viewing the video footage and verbatim quotes from the video, to support the analysis, were written up; these notes were then thematically coded in NVivo 10. A summary of the analysis for each school was produced and discussed by the research team. Common themes across the case-study sites were identified and discussed throughout the analytic phase and form the basis of the findings presented in this paper.

Participants
By using a variety of qualitative methods we were able to involve more than 600 young people in this study. Whilst only 16 pupils were willing to engage in a lunchtime go-along tour with members of the research team, 50 took part in individual or small-group interviews, 155 young people participated in one of 14 focus groups and 430 young people provided their views during the semi-structured written classroom activity. Interviews were conducted with six head teachers, seven kitchen supervisors and 25 retailers (working at or managing takeaway outlets; supermarkets, grocery stores or newsagents; bakeries; ice cream vans; and leisure/community centre or other cafés).

Findings
In this section direct quotes from participants are accompanied by a short description of the associated school, in terms of high/mixed/low relative deprivation and a high/moderate/ low density of food/drink outlets within 800 m of the school. Where particular differences between schools were identified in the analysis, this is highlighted in the text.

The sale of food and drink to young people
This section starts by describing some of the findings regarding the food and drink sold in schools and in the local food environment. The schools that took part differed in terms of the length of the lunch period, which varied from 40 to 60 min. Within the seven casestudy schools, young people were offered a range of hot and cold meal and snack options via a main cafeteria, in addition to other service areas such as from a 'pasta cart' , sandwich or salad bar, or 'coffee shop' as well as via vending machines. This variety of food options in some ways mirrors that available outside the school where food businesses offered different hot and cold foods (including, for example, pizza, chips, Chinese and Indian food, sausage rolls, sandwiches and filled rolls plus a wide range of crisps, confectionery and soft drinks), and some of this variety was available even in the vicinity of the schools with few retail outlets within 800 m. Some young people said they were content to purchase and eat food and drink within the school and felt it offered good value for money. These pupils also felt it was convenient to eat within school (particularly if it was raining) though it was not possible to determine whether such young people were in receipt of means-tested free school meals 7 , excluding them from purchasing food or drink outside school without spending their own money. Most schools operated a pre-loaded payment card system (as well as some taking cash payments), whereby young people not entitled to free school meals had to queue to load money onto their card at a payment point prior to purchasing food or drink.
All food and drink sold by schools has to meet the Nutrient Standards for Schools (NSS) and therefore is subject to limitations in terms of the sugar, fat and salt in a sold portion of food. Items such as chips could only be sold as part of a 'meal' and condiments were only available on request (salt and sugar were not readily available to pupils).
Schools were restricted in the ways that they could respond to young people's preferences, because of the NSS, and also the need to work within tight budgets, which perhaps limits innovation, capacity and the ready availability of popular food items. Local food businesses are not, of course, subject to restrictions regarding what they can sell and all allow young people to pay with cash, which many pupils reported that they preferred.

Meeting young people's food and drink desires and tastes
Many young people complained about the food and drink available in schools in terms of its quality, taste, availability, portion size and price (or the lack of visible price lists), reflecting other studies on young people's views regarding food and drink in schools (Crawford et al., 2012;Fletcher et al., 2013;Lawton, 2008;Wills et al., 2005;Young, Gilligan, & Bainbridge, 2014). Most young people who participated in this study said they bought small items rather than a 'meal' at lunchtime and that the external food environment was much more able to meet this preference than were schools. In schools small low-cost items such as a slice of pizza or a sausage in a bread roll were often perceived as being too small, offering poor value for money or tasting 'disgusting' (young person at Sch03, high dep/ mod food). Other small and popular items such as muffins or items from a 'deli bar' were quick to sell out therefore not available to everyone throughout the lunch period. Purchasing chips in schools meant also buying another set item such as breaded fish and many young people did not wish to eat such a set meal. One girl, for example, said: Outside school, young people could buy chips, fried rice, a sausage roll, pot-type noodles, crisps and chocolate and if they were still hungry, and had money left over, they could then decide to purchase additional food items, or a drink, often from a different retailer.
Young people were observed and said they wanted to purchase food and drink on which they could spend the least amount of money but which would satisfy their hunger and desire. Sandwiches, baguettes, panini or 'pasta pots' were popular in many schools and such items fulfilled young people's 'satiety + low cost' criteria, but pupils complained about the lack of availability of their preferred choices or the lack of opportunity for tailoring these foods, by having ham but without pickle in a sandwich, for example. Other small, low-cost items such as crisps were restricted to low-fat, baked varieties ('they're not that nice' , young person at Sch07, mixed dep/ high food) in order to meet the NSS and these were often not the varieties that young people wanted to buy. Young people also said they disliked the lack of availability of a choice of condiments and the lack of table salt was unpopular with some. Young people were observed and also reported being prepared to 'shop around' in order to find the food that they liked and they appreciated being able to purchase, for example, a roll without butter or with additional salad, if that was their preference. Such tailoring of food items or food service was often not possible in schools though some offered a sandwich bar where pupils could select what fillings they wanted; the quality of the fillings was a subject of complaint by young people.
The lack of flexibility and variety in schools was, overall, reported by young people as restricting their freedom to choose, something they rarely experienced when shopping in the external food environment. Young people were aware that it was hard for schools to have a ready availability of all their favourite foods, because of a need to minimise wastage (see quote, below) and to meet nutrient standards. However, they were also aware that retailers in the external food environment managed to consistently offer their preferred food and drinks, thereby offering them better service: If they [the school] made too much it would be a waste but…. Sometimes I've been behind someone and there's like one roll left or one bit of pizza and they take it, and what am I left with, a cake that I don't like … so, I'm stuck with what, a drink of juice and that's got to fill me up for like near 3 hours. (young person at Sch07, mixed dep/ high food) The marketing and promotion of food and drink in schools was sometimes described by young people as misleading, with advertised items described as 'amazing' (young person at Sch03, high dep/ mod food) for example, but the reality being described as disappointing. The marketing and sale of 'healthy food and drink' was resisted by some young people who said they were keen to avail themselves of food and drink higher in sugar, fat or salt than could be offered in school; some pupils discussed finding school staff who were willing to provide salt or sugar, if asked discreetly. Sometimes pupils reported that they did want to eat and drink healthier options but they did not consider the food and drink offered in school to be 'good' healthy fare, as the food and drink did not look or taste appealing. Fruit in schools was frequently complained about; grapes from a vending machine were described by one pupil as 'really watery' (Sch01; high dep/mod food), for example. Some pupils expressed cynicism about schools as, despite stating that they were trying to promote health, they made pizza and cookies more accessible in the dining area than fruit. One boy commented, for example: They've got a big tray of cookies, like, where you buy … where you buy your food and then there's a bowl of fruit and it's like away up high and I can't reach it, they just hide it like way up high. (young person at Sch06, low dep/ low food) The research team observed posters promoting 'treats' and offers for pupils to 'indulge themselves' in schools, which were antithetical to the general healthy eating ethos of schools. One school, for example, widely advertised a 'Tuesday treat' -a free caramel shortcake if pupils spent £1.75 (Sch01; high dep/mod food). The idea of a 'treat' was not raised as a reason for purchasing additional food or drink by young people in the study. There was therefore often a gap between what schools offered and the food-purchasing practices and ideals of young people.
In addition to the food and drink available to young people in different contexts, the way that young people were perceived or treated as customers further reveals ways in which young people have opportunities to exert agency, influence and autonomy at lunchtime when making purchases in the external food environment, compared with the school cafeteria.

Being seen and heard: young people as customers
Young people generally said that they were aware of why nutrient standards were in place in schools but few were clear about what the restrictions actually entailed in terms of the food and drink that could/could not be sold in school. Many pupils expressed dissatisfaction that changes were introduced in school without their knowledge and without consultation with them. They said they were not told or consulted about nutrient standards, price increases, menu changes or the redesign of food service areas and this lack of consultation was viewed by many as unacceptable practice. 'The school has never asked me' was a typical comment (Sch01; high dep/mod food) from a pupil. Some kitchen supervisors said they did consult with young people despite it being very difficult to do so ('it was quite difficult to make them speak' (Sch05; high dep/ high food)) though no young person participating in this study said they had been asked for their views. Whilst at Sch07, in particular, the kitchen supervisor reported 'good banter' with the young people who ate in the dining area, this cafeteria, and some of the other school canteens, was associated by young people with taking free school meals, which was given as a reason by pupils for not wanting to eat or spend time there; this makes it difficult for catering staff to engage with pupils as customers. Retailers spoke about their knowledge of young people's food and drink preferences and the way that they tailored what they offered to the young people who lived locally. This included ice-cream vans selling individual sweets at a very low price and pouring boiling water onto pot-type noodles so they were ready for pupils to eat; retailers getting orders ready for young people to collect and, for one small sandwich shop close to Sch01, offering an SMS text ordering service for young people. In the external food environment some retailers observed young people trying to make healthier choices, particularly as they became older and particularly if they were female, and some tried to build on this by offering healthier options in store (more salad choices, for example, or more sandwich fillings perceived by retailers to be healthy). Some retailers also observed what young people frequently purchased and then made or stocked similar products to tempt them to purchase alternative items. Most retailers tried hard to ensure that they had enough of the foods/drinks available that young people preferred: Yeah, so we usually know what they're after so we'll prepare it all before they come in so that we don't run out. (Bakery employee, Sch06; low dep/ low food) … we just actually listen to what they're asking for and just try to accommodate to the best of our ability, if they're saying, likes of when we first started doing the chicken tikka it was what can we bring into the mix that the young ones would like and we actually asked the kids themselves, it was a wee bit of market research, and just checked to see what flavours they would like, and we've tried out quite a few different ones…. (Leisure centre café employee, Sch02; high dep/ low food) Retailers in the vicinity of the schools we studied often discounted their prices during the lunch period, bringing prices down to levels that were more in line with the money that young people had to spend or selling cheaper options only at lunchtime (chips and beans for example, was offered at lunchtime near Sch04 for £1.50, with prices being higher in the evening). Many young people said they were not offered discounts, as these promotional lunchtime prices were considered by most to be the normal price. Within schools prices were reported by pupils to rise without notice, leaving some young people without the money to pay or requiring them to revisit the machine where money could be loaded onto their payment card before they could purchase their lunch; this was spoken about as an additional barrier to purchasing food or drink in school. Young people said that retailers had sometimes known individual pupils since they were young children ('they get to ken [know] you'; (young person at Sch01; high dep/mod food)) and also knew their parents, siblings and neighbours; they therefore felt retailers had a good understanding of what food or drink would appeal to them and several retailers felt able to tailor their services accordingly, as the following quotes illustrate: … we know them [the pupils] like, like I'm from [the same place] so I know a few of their brothers and sisters [too]…. (Sandwich shop employee, Sch02; high dep/ low food) … it's a family run business, eh, you know, so everybody knows everybody and because the shop's been there for so long everybody knows everybody, so you know, even the kids when they start school and their mother as well, come in here even in the evening. (Takeaway employee, Sch04; mixed dep/mod food) Some of the younger ones come in in the morning with the family. (Bakery employee, Sch07; mixed dep/ high food) The perceived attitude of some shopkeepers towards young people as customers, versus that of school catering staff (and teachers) was also noted by participants, in terms of the way this shaped the social environment for purchasing food and drink and eating/drinking at lunchtime. This was a significant factor in terms of where young people elected to spend their lunch break. It was notable from our analysis that retailers in areas of high relative deprivation treated young people with respect and acknowledged their custom in ways that were not evident in areas of mixed or low relative deprivation. Restrictions on young people when they were purchasing food and drink were only reported or observed in areas of relative affluence. Retailers in more affluent areas were reported to restrict the number of young people entering a shop at the same time; accused young people of shoplifting; would not allow school bags to be brought into the shop; would shout at young people if they 'messed about'; would not let young people use staffed checkouts at lunchtime and would not give young people receipts. In contrast, retailers in areas of much higher deprivation were appreciative of young people's lunchtime custom and spoke highly of young people's behaviour generally: … if adults come into our shop they'll just wait in the queue, they're all customers after all whether they're children or adults…. And I think the adults see that, the children see that and they think, 'well we're not getting shoved aside in here' , possibly that…. Aye, it does make a difference, they're not second-class citizens because they're children, eh? (Takeaway employee, Sch01; high dep/ mod food) … every customer's important to me … so if you treat them [pupils] with respect generally they're okay. (Newsagent, Sch03; high dep/ mod food) The way that young people were treated differentially as customers in the school versus the external food environment was also reflected in the way that young people could utilise these different spaces for the social element of the lunch break, which we now discuss.

The social nature of lunchtime and the importance of the peer context
The external food environment was usually seen by young people as somewhere that they wanted to spend time with their friends. Retailers were rarely reported to push young people out after they made a purchase so they could 'take their time' (young person at Sch01; high dep/mod food), often with different members of a friendship group purchasing at different outlets whilst their friends waited for them. Once everyone had purchased what they wanted the group often stood around the local shops eating and drinking, or they took a slow walk back to school whilst they ate, drank and socialised. At one fish-and-chip shop close to Sch04 pupils were observed sitting in a circle on the floor of the outlet after the lunchtime rush had subsided; the owner said he was happy to accommodate this. Schools, by contrast, were sometimes seen as being anti-social, with the school dining environment seen as a barrier to spending time with friends. Staff sometimes closed the door to the dining hall to discourage pupils from coming inside to spend time there after the initial lunchtime rush and some staff started clearing the tables and cleaning floors whilst young people were chatting to their friends after eating lunch. Comments such as 'they kick us out early'; 'the café should be just for us' (Sch01; high dep/mod food) were typical from pupils at schools perceived to be unwelcoming and some catering staff were described as 'grumpy' (Sch01; high dep/mod food). Some pupils reported sitting and eating lunch in the corridors as an alternative to the main dining area though this meant some got 'screamed' at (Sch04; mixed dep/mod food). Many schools could not accommodate all the young people who elected to purchase food or drink there. One head teacher commented: … one of the things that causes a problem, we don't have enough seats, so we've only got between one hundred and two hundred dining seats but in a school of a thousand pupils, so you can see the potential difficulty there and they don't like getting up out of their seats, 'once I get my seat that's that, I'm not moving from here' . (Sch02, high dep / low food) Spending time with friends was viewed by young people as a critical element of the lunch break, prioritised over where food was purchased from. Only at Sch06, the least deprived in the study, with few food outlets nearby, did the head teacher raise the issue of the school considering how the space worked for pupils socially: … so it is a new school, so probably they try to consider the creation of social space and social area for the students. (Sch06, low dep/ low food) He also added that pupils were welcome to bring food and drink in that they had bought from retailers outside school; this was the only school where this was encouraged.

Discussion
The findings highlight the challenges that schools face in providing nutritionally balanced food and drink that young people are prepared to purchase. Nutrient standards limit the sale of food and drink that young people say they want to buy whereas the external food environment faces no such restriction. In addition, food businesses value young people as customers because they have a commercial relationship with them and this reinforces young people's position as consumers, facilitating their autonomy, agency and influence in ways that schools are unable to manage easily.
Schools and school cafeterias, unlike local food businesses, are not profit driven but have a role as social and educational enterprises. Retailers have more incentive to retain their customers by building long-term relationships (Gummesson, 1994) with young people, through providing services that are responsive to their needs, including employee friendliness and promptness of service (Kerin, Jain, & Howard, 1992). The findings suggest that some school catering staff do not view or treat young people as having a right to stay in the school dining room at all after they have finished eating, to spend time there socially. The study shows that this is in conflict with the need for young people to find ways to assert their agency and to govern what they see as their time to be autonomous or to be with friends. This highlights a conceptual difference in the way that young people are viewed, in terms of their relationship to what is essentially an educational setting (Ruckenstein, 2012) as well as their role as social actors in an education versus a commercial environment.
As schools very often cannot physically cater for every young person on the school roll because of the limited capacity of the dining room, and because they often do not provide food or drink that young people want to buy, this means young people have little power, autonomy or influence within the school setting. Such socio-spatial or socio-environmental practices act as a form of governance (Pike, 2008), pushing young people out into the external food environment. This in turn encourages young people to become customers in the commercial marketplace and to make their own food decisions (see the study by Fletcher et al. (2013) for further evidence of this). They thereby start to govern themselves, making more and more choices as consumers and thus making it increasingly difficult for schools to then attract them back to the school cafeteria. Whilst this has positive repercussions in terms of young people developing autonomy, agency and power it has more negative implications in terms of pupils being less likely to consume a healthier diet through staying on site and purchasing the more nutritious food and drink available in schools. This represents a considerable challenge to schools and to public health policy and practice.
It is notable that retailers particularly engage with young people in relatively deprived neighbourhoods; such young people are often perceived to be powerless and marginalised but this study shows that, as purchasers of food and drink at least, young people from socio-economically deprived areas can have significant social and economic capital within the local food environment. This is fuelled by neoliberalism, however, and the marketing of food and drink to young people who are all considered 'fair game' (Boyles, 2008). Retailers in socio-economically deprived areas encourage young people to spend what little money they have with them, rather than lose this business to a competitor (or a school) because they are driven by a desire to increase turnover and profit.
The educational heart of schools tends to mean they focus on future (good) outcomes for young people and this risks ignoring the temporal context of childhood and youthhood and the need to enable young people to enjoy and benefit from the time they spend at school (Quennerstedt & Quennerstedt, 2013), including during the lunch period. This future-oriented outlook sits uncomfortably with the public health goal of encouraging healthier food and drink choices during adolescence (not just in the future, in adulthood). It also sits in tension somewhat with ideas about the rights of the child advocated in the UN Convention as it does not acknowledge young people's contemporary experiences (Mayall, 2002), giving overriding authority to the adults who determine the menus and ethos for food and drink sold in schools. In the external food environment, where many retailers retain young people as customers by listening to and acknowledging their needs and preferences, this gives young people the perception that they are valued and have a certain amount of power and influence within the commercial marketplace, even though this is likely to be driven by business rather than social or moral factors.
It is relatively straightforward to consult with young people but harder to achieve young people's full participation in schools in practice (Wills et al., 2008). However, by not regularly or actively engaging young people and listening to their lunchtime needs they are too frequently denied the right to participate and to be heard (Le Borgne, 2014). Young people said that they were aware of the need to provide healthier food options and to have regulations to ensure this happens consistently across schools, but our study findings suggest that they are not active or equal partners in discussions about how this can best be achieved. This study additionally suggests that there is also a need to change the way 'healthy food' is provided and promoted to young people in schools, as pupils voice concerns about mixed messages regarding the marketing and provision of less healthy items and the often poor quality of the nutritional items on offer. Adopting a more customer/pupil-centric approach, which prioritises tasty and nutritious food that young people are willing to purchase, requires further thought. Social marketing initiatives might help support efforts to make healthier food choices more popular and accessible to pupils through a better understanding of their social needs (Evans, Christoffel, Necheles, & Becker, 2010;Stead, McDermott, MacKintosh, & Adamson, 2011). During this study young people put forward a range of ideas about the food and drink that could be served in school and they also had questions to which they required answers -regarding the criteria for eligibility for free school meals, for example. Currently these ideas and questions remain unheard, which increases the power imbalance between young people and adults and prevents young people becoming more fully responsible citizens in the school and wider environment (Le Borgne, 2014).

Conclusions
Overall these findings provide fresh insight into some of the factors that work to give or to restrict young people's agency, power and influence in terms of the food and drink they purchase during the school lunch break. The findings suggest that schools face challenges as settings for public health improvements through the food and drink sold there because they are less able than the external food environment to meet young people's expectations regarding food, drink and the 'dining experience' . There is a need to assess whether and how school caterers can learn from retailers, to adjust their initiatives and services to attract young people to school canteens in an attempt to encourage them to purchase healthier food and drink. This is tough, as schools need to retain their status as responsible caterers offering the 'gold standard' of nutritious food and drink, and our findings show that teenagers are cynical when marketing 'gimmicks' similar to those used on the high street are employed by schools. One aspect of the retailer-pupil dynamic that may translate well would be for schools to further consider how to build better relationships between teachers, catering staff and pupils so that young people are valued and listened to, as they are by retailers. Unlike the retailer-pupil relationship, however, such improvements in schools could be driven by a genuine desire to improve young people's health and well-being. Whilst structural changes and investment, within and beyond the school gate, are perhaps inevitable if we want to enhance the 'eating experience' and healthy eating practices of young people, the findings also suggest that pupils themselves can perhaps be mobilised to push for change regarding the food and drink sold to them at school.

Notes
and direct where to go, based on their natural rhythms -not the researcher's. In addition, go-along tours are useful to investigate the context and meaning of practices that have an element of space or place -i.e. why some food outlets are frequented but not others. During go-along tours, researchers observe but also discuss with participants their food and drink purchasing and consumption practices. 6. The quantitative findings focus on food and drink purchases made outside school only and would add little to the arguments provided in the present paper. 7. An inherent challenge when using mainly group-based methods is not being able to ask young people to discuss in front of peers whether they were receiving free school meals.